P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005...
Transcript of P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005...
![Page 1: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/1.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
1
A traffic-like model for supply chains
P. Degond
MIP, CNRS and Université Paul Sabatier,
118 route de Narbonne, 31062 Toulouse cedex, France
[email protected] (see http://mip.ups-tlse.fr)
Joint work with
D. Armbruster & C. Ringhofer
Arizona State University
![Page 2: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/2.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
2Summary
1. Introduction
2. A simple discrete event simulator
3. Continuity equation
4. Constitutive relation
5. Passage to Eulerian variables
6. Kinetic model
7. Multiphase fluid model
8. Numerical results
9. Conclusion
![Page 3: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/3.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
3
1. Introduction
![Page 4: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/4.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
4A supply chain
➠ In a very simple framework:➟ A supply chain is a string (or network) of
processors (or stations) along which goods (orparts) are circulating
![Page 5: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/5.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
4A supply chain
➠ In a very simple framework:➟ A supply chain is a string (or network) of
processors (or stations) along which goods (orparts) are circulating
➠ At each processor, the goods undergo atransformation.
![Page 6: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/6.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
4A supply chain
➠ In a very simple framework:➟ A supply chain is a string (or network) of
processors (or stations) along which goods (orparts) are circulating
➠ At each processor, the goods undergo atransformation.
➠ Each processor requires a certain throughput timeT to process a given good
![Page 7: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/7.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
4A supply chain
➠ In a very simple framework:➟ A supply chain is a string (or network) of
processors (or stations) along which goods (orparts) are circulating
➠ At each processor, the goods undergo atransformation.
➠ Each processor requires a certain throughput timeT to process a given good
➠ Each processor has a limited capacityq(maximum number of goods it is able to deliverper unit of time)
![Page 8: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/8.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
5A supply chain
➠ In front of each processor, goods can be stored inbuffer queues while waiting for being treated
![Page 9: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/9.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
5A supply chain
➠ In front of each processor, goods can be stored inbuffer queues while waiting for being treated
➠ Goods are picked up in the queues according to agiven policy.➟ The simplest policy: FIFO (first in first out)➟ More complex policies (e.g. tagged ’hot lots’ to
be processed with higher priority)
![Page 10: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/10.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
6Discrete Event Simulators (DES)
➠ The simplest model to describe a supply chain
![Page 11: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/11.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
6Discrete Event Simulators (DES)
➠ The simplest model to describe a supply chain
➠ Provides a recursion formula forτ(m,n) = timeat which the partPn enters the buffer queue ofStationSm
![Page 12: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/12.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
6Discrete Event Simulators (DES)
➠ The simplest model to describe a supply chain
➠ Provides a recursion formula forτ(m,n) = timeat which the partPn enters the buffer queue ofStationSm
➠ Discrete analog of a particle model in gasdynamics
![Page 13: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/13.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
7Hierarchy of models
➠ Like in fluid dynamics, one can derive➟ fluid models➟ kinetic modelsfor supply chains
![Page 14: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/14.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
7Hierarchy of models
➠ Like in fluid dynamics, one can derive➟ fluid models➟ kinetic modelsfor supply chains
➠ The goal of this talk➟ ’Rigorously’ derive a fluid model from a large
particle limit of a DES model under FIFOpolicy
➟ Refine this model into a kinetic model able toaccount for more complex policies (hot lots)
![Page 15: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/15.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
8References
➠ DES simulation:[Banks, Carson II, Nelson]
![Page 16: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/16.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
8References
➠ DES simulation:[Banks, Carson II, Nelson]
➠ Fluid models:[Anderson], [Billings, Hasenbein],[Newell]
➟ Recently:[Klar et al]
![Page 17: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/17.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
8References
➠ DES simulation:[Banks, Carson II, Nelson]
➠ Fluid models:[Anderson], [Billings, Hasenbein],[Newell]
➟ Recently:[Klar et al]
➠ Review: [Daganzo]
![Page 18: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/18.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
9
2. A simple discrete event simulator
![Page 19: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/19.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
10Data
➠ StationSm
➟ Capacityqm (number of parts per unit of time)➟ Throughput timeTm (time needed to process a
single part)
![Page 20: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/20.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
10Data
➠ StationSm
➟ Capacityqm (number of parts per unit of time)➟ Throughput timeTm (time needed to process a
single part)
➠ τ(m,n) = time at which the partPn enters thebuffer queue of StationSm
![Page 21: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/21.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
10Data
➠ StationSm
➟ Capacityqm (number of parts per unit of time)➟ Throughput timeTm (time needed to process a
single part)
➠ τ(m,n) = time at which the partPn enters thebuffer queue of StationSm
➠ Buffer queues are of infinite size (can be relaxed)
![Page 22: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/22.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
11Case distinction
➠ First case: buffer ofSm non-empty➟ Sm processes at full rateqm
=⇒ τ(m + 1, n) = τ(m + 1, n − 1) +1
qm
![Page 23: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/23.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
11Case distinction
➠ First case: buffer ofSm non-empty➟ Sm processes at full rateqm
=⇒ τ(m + 1, n) = τ(m + 1, n − 1) +1
qm
➠ Second case: buffer ofSm empty➟ Sm processes part when it arrives
=⇒ τ(m + 1, n) = τ(m,n) + Tm
![Page 24: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/24.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
12Recursion formula
➠ If buffer of Sm non-empty
=⇒ τ(m + 1, n) ≥ τ(m,n) + Tm
![Page 25: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/25.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
12Recursion formula
➠ If buffer of Sm non-empty
=⇒ τ(m + 1, n) ≥ τ(m,n) + Tm
➠ If buffer of Sm empty
=⇒ τ(m + 1, n) ≥ τ(m + 1, n − 1) +1
qm
![Page 26: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/26.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
12Recursion formula
➠ If buffer of Sm non-empty
=⇒ τ(m + 1, n) ≥ τ(m,n) + Tm
➠ If buffer of Sm empty
=⇒ τ(m + 1, n) ≥ τ(m + 1, n − 1) +1
qm
➠ Collect the two cases into
τ(m+1, n) = max{ τ(m+1, n−1)+1
qm
, τ(m,n)+Tm }
![Page 27: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/27.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
13The continuum limit
➠ Investigate the ’thermodynamic limit’M,N → ∞,➟ M = Number of stations➟ N = Number of partsand find a continuum model
![Page 28: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/28.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
13The continuum limit
➠ Investigate the ’thermodynamic limit’M,N → ∞,➟ M = Number of stations➟ N = Number of partsand find a continuum model
➠ Idea: find that the DES is a discrete version of aconservation law in Lagrangian variable➟ n = part index= ’mass’ variable➟ m = station index= ’space’ variable
![Page 29: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/29.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
14
3. Continuity equation
![Page 30: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/30.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
15Position of part Pn
➠ ’Position’ (or Station numberm) at which partPn
is at timet given by
µ(t, n) =1
M
M∑
m=1
H(t − τ(m,n))
whereH(x) = 1 for x > 0 and0 otherwise(Heaviside fct)
![Page 31: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/31.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
15Position of part Pn
➠ ’Position’ (or Station numberm) at which partPn
is at timet given by
µ(t, n) =1
M
M∑
m=1
H(t − τ(m,n))
whereH(x) = 1 for x > 0 and0 otherwise(Heaviside fct)
➠ Indeed, ift > τ(m,n), Pn already passedSm andwe can increment position by1, otherwise not
0 ≤ µ(t, n) ≤ 1
![Page 32: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/32.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
15Position of part Pn
➠ ’Position’ (or Station numberm) at which partPn
is at timet given by
µ(t, n) =1
M
M∑
m=1
H(t − τ(m,n))
whereH(x) = 1 for x > 0 and0 otherwise(Heaviside fct)
➠ Indeed, ift > τ(m,n), Pn already passedSm andwe can increment position by1, otherwise not
0 ≤ µ(t, n) ≤ 1
![Page 33: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/33.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
16Velocity and specific volume
➠ ’Velocity’ of part Pn
v(t, n) =d
dtµ(t, n) =
1
M
M∑
m=1
δ(t − τ(m,n))
![Page 34: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/34.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
16Velocity and specific volume
➠ ’Velocity’ of part Pn
v(t, n) =d
dtµ(t, n) =
1
M
M∑
m=1
δ(t − τ(m,n))
➠ Spacing between the parts (= ’specific volume’)
θ(t, n) = −µ(t, n + 1) − µ(t, n)
1/N, θ(t,N) = 0
![Page 35: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/35.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
16Velocity and specific volume
➠ ’Velocity’ of part Pn
v(t, n) =d
dtµ(t, n) =
1
M
M∑
m=1
δ(t − τ(m,n))
➠ Spacing between the parts (= ’specific volume’)
θ(t, n) = −µ(t, n + 1) − µ(t, n)
1/N, θ(t,N) = 0
➠ By construction
d
dtθ(t, n) +
v(t, n + 1) − v(t, n)
1/N= 0
![Page 36: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/36.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
17Continuum limit
➠ Define➟ y = mass variable∈ [0, 1] (part number)➟ x = space variable∈ [0, 1] (station number)
n = [Ny] , m = [Mx] [·] = integer part
![Page 37: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/37.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
17Continuum limit
➠ Define➟ y = mass variable∈ [0, 1] (part number)➟ x = space variable∈ [0, 1] (station number)
n = [Ny] , m = [Mx] [·] = integer part
➠ Assumption
τM,N([Mx], [Ny]) −→ τ(x, y)
asM,N → ∞, as smoothly as we need
![Page 38: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/38.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
18Position in the continuum limit
➠ We have
µM,N(t, [Ny]) −→ X(t, y)
wheret → X(t, y) is the inverse function ofx → τ(x, y) (which is increasing)
![Page 39: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/39.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
18Position in the continuum limit
➠ We have
µM,N(t, [Ny]) −→ X(t, y)
wheret → X(t, y) is the inverse function ofx → τ(x, y) (which is increasing)
➠ Proof: just a change of variable in the integraldefiningµ
![Page 40: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/40.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
18Position in the continuum limit
➠ We have
µM,N(t, [Ny]) −→ X(t, y)
wheret → X(t, y) is the inverse function ofx → τ(x, y) (which is increasing)
➠ Proof: just a change of variable in the integraldefiningµ
➠ X(t, y) is the position in Lagrangian variables
![Page 41: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/41.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
19Velocity and Specific volume
➠ We have
vM,N(t, [Ny]) =d
dtµM,N(t, [Ny]) →
dX
dt(t, y) := v(t, y)
(velocity in Lagangian variables)
![Page 42: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/42.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
19Velocity and Specific volume
➠ We have
vM,N(t, [Ny]) =d
dtµM,N(t, [Ny]) →
dX
dt(t, y) := v(t, y)
(velocity in Lagangian variables)
➠ and
θM,N(t, [Ny]) = −µM,N(t, [Ny] + 1) − µM,N(t, [Ny])
1/N
→ −∂X
∂y(t, y) := θ(t, y)
(specific volume)
![Page 43: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/43.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
20Continuity equation
➠ Since∂
∂y
(
dX
dt
)
=d
dt
(
∂X
∂y
)
![Page 44: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/44.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
20Continuity equation
➠ Since∂
∂y
(
dX
dt
)
=d
dt
(
∂X
∂y
)
➠ The continuity equation is satisfied
∂θ
∂t+
∂v
∂y= 0
![Page 45: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/45.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
21
4. Constitutive relation
![Page 46: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/46.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
22Recursion formula
➠ was written
τ([Mx] + 1, [Ny]) = max{ τ([Mx] + 1, [Ny] − 1)
+1
q([Mx]), τ([Mx], [Ny]) + T ([Mx]) }
![Page 47: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/47.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
23Recursion formula (cont)
➠ After dividing by1/N and some rearrangement
max{
τ([Mx] + 1, [Ny] − 1) − τ([Mx] + 1, [Ny])
1/N+
N
q([Mx]),
N
M(τ([Mx], [Ny]) − τ([Mx] + 1, [Ny])
1/M+ MT ([Mx]))
} = 0
![Page 48: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/48.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
24Scaling hypotheses
➠ N → ∞, qM,N → ∞ and
N
qM,N([Mx])→
1
q(x)
![Page 49: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/49.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
24Scaling hypotheses
➠ N → ∞, qM,N → ∞ and
N
qM,N([Mx])→
1
q(x)
➠ M → ∞, TM,N → 0 and
MTM,N([Mx]) → T q(x)
![Page 50: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/50.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
24Scaling hypotheses
➠ N → ∞, qM,N → ∞ and
N
qM,N([Mx])→
1
q(x)
➠ M → ∞, TM,N → 0 and
MTM,N([Mx]) → T q(x)
➠N
M→ 1 Number of parts and stations are of the
same order of magnitude
![Page 51: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/51.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
25Recursion: continuum limit
➠ Under scaling assumptions, asN,M → ∞:
max{−∂τ
∂y+
1
q, −
∂τ
∂x+ T} = 0
![Page 52: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/52.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
25Recursion: continuum limit
➠ Under scaling assumptions, asN,M → ∞:
max{−∂τ
∂y+
1
q, −
∂τ
∂x+ T} = 0
➠ Using thatt → X(t, y) is the inverse fct ofx → τ(x, y)
∂τ
∂x(X(t, y), y) =
1
v(t, y),
∂τ
∂y(X(t, y), y) =
θ(t, y)
v(t, y)
![Page 53: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/53.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
25Recursion: continuum limit
➠ Under scaling assumptions, asN,M → ∞:
max{−∂τ
∂y+
1
q, −
∂τ
∂x+ T} = 0
➠ Using thatt → X(t, y) is the inverse fct ofx → τ(x, y)
∂τ
∂x(X(t, y), y) =
1
v(t, y),
∂τ
∂y(X(t, y), y) =
θ(t, y)
v(t, y)
➠ Gives
max{−θ(t, y)
v(t, y)+
1
q(X(t, y)), −
1
v(t, y)+T (X(t, y))} = 0
![Page 54: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/54.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
26Resolution of the max
➠ Either max is attained for 1st argument and
v(t, y) = q(X(t, y))θ(t, y) ≤1
T (X(t, y))
![Page 55: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/55.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
26Resolution of the max
➠ Either max is attained for 1st argument and
v(t, y) = q(X(t, y))θ(t, y) ≤1
T (X(t, y))
➠ or max is attained for 2nd argument and
v(t, y) =1
T (X(t, y))≤ q(X(t, y))θ(t, y)
![Page 56: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/56.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
26Resolution of the max
➠ Either max is attained for 1st argument and
v(t, y) = q(X(t, y))θ(t, y) ≤1
T (X(t, y))
➠ or max is attained for 2nd argument and
v(t, y) =1
T (X(t, y))≤ q(X(t, y))θ(t, y)
➠ Thus
v(t, y) = min{1
T (X(t, y)), q(X(t, y))θ(t, y)}
![Page 57: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/57.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
27Supply chain model (in Lagrang. coord.)
∂θ
∂t+
∂v
∂y= 0
v(t, y) = min{1
T (X(t, y)), q(X(t, y))θ(t, y)}
−∂X
∂y(t, y) = θ(t, y)
![Page 58: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/58.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
27Supply chain model (in Lagrang. coord.)
∂θ
∂t+
∂v
∂y= 0
v(t, y) = min{1
T (X(t, y)), q(X(t, y))θ(t, y)}
−∂X
∂y(t, y) = θ(t, y)
➠ Last eq. equivalent to:
X(t, y) =
∫ 1
y
θ(t, z)dz
![Page 59: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/59.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
28
5. Passage to Eulerian variables
![Page 60: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/60.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
29Goal
➠ Obtain a model in(t, x) rather than(t, y)
![Page 61: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/61.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
29Goal
➠ Obtain a model in(t, x) rather than(t, y)
➠ Classical procedure in gas dynamics➟ Coordinate change:
x = X(t, y) =
∫ 1
y
θ(t, z)dz strictly ց of y
y = Y (t, x) inverse fct
![Page 62: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/62.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
30Eulerian unknowns
➠ Using thaty → Y (t, x) is the inverse fct ofx → X(t, y)
∂Y
∂x(t, x) = −
1
θ(t, Y (t, x)):= ρ(t, x)
Number density of parts atx at timet
![Page 63: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/63.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
30Eulerian unknowns
➠ Using thaty → Y (t, x) is the inverse fct ofx → X(t, y)
∂Y
∂x(t, x) = −
1
θ(t, Y (t, x)):= ρ(t, x)
Number density of parts atx at timet
➠ and
∂Y
∂t(t, x) = −
v(t, Y (t, x))
θ(t, Y (t, x)):= ρ(t, x)u(t, x)
u(t, x) = v(t, Y (t, x)) velocity in Eulerian coord.
![Page 64: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/64.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
31Continuity and constitutive eqs.
➠ Since∂
∂t
(
∂Y
∂x
)
=∂
∂x
(
∂Y
∂t
)
![Page 65: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/65.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
31Continuity and constitutive eqs.
➠ Since∂
∂t
(
∂Y
∂x
)
=∂
∂x
(
∂Y
∂t
)
➠ The continuity equation is satisfied
∂ρ
∂t+
∂ρu
∂x= 0
![Page 66: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/66.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
31Continuity and constitutive eqs.
➠ Since∂
∂t
(
∂Y
∂x
)
=∂
∂x
(
∂Y
∂t
)
➠ The continuity equation is satisfied
∂ρ
∂t+
∂ρu
∂x= 0
➠ From the constitutive relation in Lagrangianvariable, we get
ρu(t, x) = min{1
T (x)ρ(t, x) , q(x)}
![Page 67: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/67.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
32Continuum supply chain model
∂ρ
∂t+
∂ρu
∂x= 0
ρu(t, x) = min{1
T (x)ρ(t, x) , q(x)}
![Page 68: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/68.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
32Continuum supply chain model
∂ρ
∂t+
∂ρu
∂x= 0
ρu(t, x) = min{1
T (x)ρ(t, x) , q(x)}
➠ Hyperbolic model with flux constraint
![Page 69: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/69.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
33
6. Kinetic model
![Page 70: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/70.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
34Particle interpretation of fluid model
➠ Fluid model
∂ρ
∂t+
∂ρu
∂x= 0
ρu = min{ρV0 , q} , V0 =1
T
![Page 71: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/71.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
34Particle interpretation of fluid model
➠ Fluid model
∂ρ
∂t+
∂ρu
∂x= 0
ρu = min{ρV0 , q} , V0 =1
T
➠ Particle interpretation
X =
{
V0
0ρ =
{
−ρ∂xV0 if ρV0 ≤ q
−∂xq if ρV0 > q
![Page 72: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/72.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
35Particle motion
➠ Either particle moves or is blocked according towhether the ’free’ fluxρV0 is below or exceeds thethresholdq.
![Page 73: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/73.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
35Particle motion
➠ Either particle moves or is blocked according towhether the ’free’ fluxρV0 is below or exceeds thethresholdq.
➠ Kinetic model: ’regularization’ of this singulardynamics
![Page 74: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/74.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
35Particle motion
➠ Either particle moves or is blocked according towhether the ’free’ fluxρV0 is below or exceeds thethresholdq.
➠ Kinetic model: ’regularization’ of this singulardynamics
➠ Introduce an attribute variableξ to each particle➟ Particles move with actual velocity
V (t, x, ξ) ≤ V0
➟ s.t. total flux≤ q
![Page 75: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/75.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
36Distribution function
➠ f(x, ξ, t) density of parts at timet, positionx,with attributeξ
![Page 76: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/76.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
36Distribution function
➠ f(x, ξ, t) density of parts at timet, positionx,with attributeξ
➠ Densityρ and fluxρu
ρ =
∫
f(x, ξ, t) dξ , ρu =
∫
f(x, ξ, t)V (x, ξ, t) dξ
![Page 77: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/77.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
36Distribution function
➠ f(x, ξ, t) density of parts at timet, positionx,with attributeξ
➠ Densityρ and fluxρu
ρ =
∫
f(x, ξ, t) dξ , ρu =
∫
f(x, ξ, t)V (x, ξ, t) dξ
➠ Maximal possible fluxQ:
Q =
∫
f(x, ξ, t)V0(x) dξ = ρV0
Flux if there would be no capacity limitation
![Page 78: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/78.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
37Policy
➠ Higher priority to parts with lower attribute values➟ Note: same methodology would apply for other
policies
![Page 79: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/79.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
37Policy
➠ Higher priority to parts with lower attribute values➟ Note: same methodology would apply for other
policies
➠ Procedure: move parts by increasing attributenumber with maximum allowed speedV0 untilprocessor capacity is reached
![Page 80: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/80.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
37Policy
➠ Higher priority to parts with lower attribute values➟ Note: same methodology would apply for other
policies
➠ Procedure: move parts by increasing attributenumber with maximum allowed speedV0 untilprocessor capacity is reached
➠ Number of parts with attribute≤ α is∫ α
−∞
f(x, ξ, t) dξ
![Page 81: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/81.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
38Implementing policy I
➠ If parts with attribute≤ α all move with maximalspeedV0, the associated flux is
β(x, α, t) = V0(x)
∫ α
−∞
f(x, ξ, t) dξ
![Page 82: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/82.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
38Implementing policy I
➠ If parts with attribute≤ α all move with maximalspeedV0, the associated flux is
β(x, α, t) = V0(x)
∫ α
−∞
f(x, ξ, t) dξ
➠ The fonction
α ∈ R → β(x, α, t) ∈ [0, Q]
is increasing
![Page 83: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/83.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
38Implementing policy I
➠ If parts with attribute≤ α all move with maximalspeedV0, the associated flux is
β(x, α, t) = V0(x)
∫ α
−∞
f(x, ξ, t) dξ
➠ The fonction
α ∈ R → β(x, α, t) ∈ [0, Q]
is increasing
➠ Denoteβ−1(x, ·, t) its inverse
![Page 84: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/84.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
39Implementing policy II
➠ Processors process all parts with attribute≤ α
➟ whereα s.t. associated flux= processorcapacity
β(x, α, t) = q (⇔ α = β−1(x, q, t) )
![Page 85: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/85.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
39Implementing policy II
➠ Processors process all parts with attribute≤ α
➟ whereα s.t. associated flux= processorcapacity
β(x, α, t) = q (⇔ α = β−1(x, q, t) )
➠ except if maximal possible fluxQ lower thanprocessor capacityq➟ in which caseα = ∞
q > Q =⇒ α = ∞
![Page 86: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/86.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
39Implementing policy II
➠ Processors process all parts with attribute≤ α
➟ whereα s.t. associated flux= processorcapacity
β(x, α, t) = q (⇔ α = β−1(x, q, t) )
➠ except if maximal possible fluxQ lower thanprocessor capacityq➟ in which caseα = ∞
q > Q =⇒ α = ∞
➠ Thenβ(x, α, t) = min{q,Q}
![Page 87: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/87.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
40Actual velocity V (x, ξ, t)
➠ Note: processor velocityV0 can be attributedependent
V0 = V0(x, ξ)
![Page 88: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/88.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
40Actual velocity V (x, ξ, t)
➠ Note: processor velocityV0 can be attributedependent
V0 = V0(x, ξ)
➠ Actual velocity
V (x, ξ, t) =
{
V0(x, ξ) if ξ ≤ α(x, t)
0 if ξ > α(x, t)
![Page 89: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/89.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
40Actual velocity V (x, ξ, t)
➠ Note: processor velocityV0 can be attributedependent
V0 = V0(x, ξ)
➠ Actual velocity
V (x, ξ, t) =
{
V0(x, ξ) if ξ ≤ α(x, t)
0 if ξ > α(x, t)
➠ or
V (x, ξ, t) = V0(x, ξ)H(α(x, t)−ξ) H = Heaviside
![Page 90: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/90.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
41Actual velocity II
➠ Sinceβ ր fonction ofα, we have
H(α(x, t) − ξ) = H(β(x, α(x, t), t) − β(x, ξ, t))
= H(min{q,Q} − β(x, ξ, t))
= H(q − β(x, ξ, t)) (sinceβ ≤ Q)
![Page 91: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/91.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
41Actual velocity II
➠ Sinceβ ր fonction ofα, we have
H(α(x, t) − ξ) = H(β(x, α(x, t), t) − β(x, ξ, t))
= H(min{q,Q} − β(x, ξ, t))
= H(q − β(x, ξ, t)) (sinceβ ≤ Q)
➠ and
β(x, ξ, t) =
∫ ξ
−∞
V0(x, ξ′) f(x, ξ′, t) dξ′
=
∫
R
V0(x, ξ′) f(x, ξ′, t)H(ξ − ξ′) dξ′
![Page 92: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/92.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
42Actual velocity III
➠ Finally
V (x, ξ, t) = V0(x, ξ)H(q − β(x, ξ, t))
β(x, ξ, t) =
∫
R
V0(x, ξ′) f(x, ξ′, t)H(ξ − ξ′) dξ′
![Page 93: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/93.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
43Part dynamics
➠ By analogy with fluid model
X = V (X,Ξ, t) = V0(X,Ξ)H(q − β(X,Ξ, t))
f = −f(∂xV )|(X,Ξ,t)
Ξ = 0
![Page 94: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/94.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
43Part dynamics
➠ By analogy with fluid model
X = V (X,Ξ, t) = V0(X,Ξ)H(q − β(X,Ξ, t))
f = −f(∂xV )|(X,Ξ,t)
Ξ = 0
➠ Characteristics of the first order kinetic eq.
∂tf + ∂x(V f) = 0
![Page 95: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/95.jpg)
![Page 96: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/96.jpg)
![Page 97: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/97.jpg)
![Page 98: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/98.jpg)
![Page 99: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/99.jpg)
![Page 100: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/100.jpg)
![Page 101: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/101.jpg)
![Page 102: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/102.jpg)
![Page 103: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/103.jpg)
![Page 104: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/104.jpg)
![Page 105: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/105.jpg)
![Page 106: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/106.jpg)
![Page 107: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/107.jpg)
![Page 108: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/108.jpg)
![Page 109: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/109.jpg)
![Page 110: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/110.jpg)
![Page 111: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/111.jpg)
![Page 112: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/112.jpg)
![Page 113: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/113.jpg)
![Page 114: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/114.jpg)
![Page 115: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/115.jpg)
![Page 116: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/116.jpg)
![Page 117: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/117.jpg)
![Page 118: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/118.jpg)
![Page 119: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/119.jpg)
![Page 120: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/120.jpg)
![Page 121: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/121.jpg)
![Page 122: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/122.jpg)
![Page 123: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/123.jpg)
![Page 124: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/124.jpg)
![Page 125: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/125.jpg)
![Page 126: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/126.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
61Case 2: Multiphase and kinetic models
0 5 10 15 20−100
0
100
200
300
time
to d
ue d
ate
PARTICLES
0 5 10 15 20−100
0
100
200
300
t=50
2 PHASE MODEL
0 5 10 15 20−100
0
100
200
300
time
to d
ue d
ate
0 5 10 15 20−100
0
100
200
300
t=70
0 5 10 15 20−100
0
100
200
300
station
time
to d
ue d
ate
0 5 10 15 20−100
0
100
200
300
station
t=90
Phases as a function of ’space’ (or DOC)at various times (top to down)
Left: kinetic model Right: two-phase model
![Page 127: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/127.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
62Case 2: Multiphase and kinetic models
0 5 10 15 20120
140
160
180
t=33
.75
PHASE
0 5 10 15 2010
0
102
104
WEIGHT
0 5 10 15 2050
100
150
200t=
67.5
0 5 10 15 2010
2
104
106
0 5 10 15 200
100
200
300
t=10
1.25
0 5 10 15 2010
2
104
106
0 5 10 15 20150
200
250
300
station
t=13
5
0 5 10 15 2010
0
105
station
Phases (left) and densities (right) as a function of ’space’
at various times (top to down)Kinetic (×, ∆) and two-phase (—, –· –) models
![Page 128: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/128.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
63Case 2: Multiphase and kinetic models
20 40 60 80 100 120 140−20
0
20
40
60
80
100
120
time
aver
age
time
to d
ue d
ate
on o
utpu
t
Expected time to due-datem1/m0 in the last cellas a function of time
Kinetic (· · · ) and two-phase (—) models
![Page 129: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/129.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
64
9. Conclusion
![Page 130: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/130.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
65Summary
➠ (Rigorous) derivation of continuum model fromDiscrete Event Simulator➟ Nonlinear hyperbolic model with saturated flux➟ Reproduces DES model satisfactorily even in
situations where capacity has large variations
![Page 131: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/131.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
65Summary
➠ (Rigorous) derivation of continuum model fromDiscrete Event Simulator➟ Nonlinear hyperbolic model with saturated flux➟ Reproduces DES model satisfactorily even in
situations where capacity has large variations
➠ Kinetic model with an internal variable (policyattribute)➟ Simple closure recovers the fluid model➟ Multiphase closure allows to implement
policies➟ Correct agreement between two-phase and
kinetic models
![Page 132: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/132.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
66Work in progress
➠ Fluid model➟ Networks
![Page 133: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/133.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
66Work in progress
➠ Fluid model➟ Networks
➠ Kinetic model➟ Randomness (random failures)
![Page 134: P. Degond - unice.fr · (Summary) Pierre Degond - Traffic-like models for supply chains - Nov 2005 (Conclusion) Discrete Event Simulators (DES) 6 The simplest model to describe a](https://reader035.fdocuments.net/reader035/viewer/2022071021/5fd51084e3c75b159b0542e6/html5/thumbnails/134.jpg)
(Summary) (Conclusion)Pierre Degond - Traffic-like models for supply chains - Nov 2005
66Work in progress
➠ Fluid model➟ Networks
➠ Kinetic model➟ Randomness (random failures)
➠ More complex models➟ Orders, payments, etc.