Division of IT Convergence Engineering Optimal Demand-Side Energy Management Under Real-time...
-
Upload
stephen-todd -
Category
Documents
-
view
213 -
download
0
Transcript of Division of IT Convergence Engineering Optimal Demand-Side Energy Management Under Real-time...
Division of IT Convergence Engineering
Optimal Demand-Side Energy Management UnderReal-time Demand-Response Pricing
Jin Xiao1, Jae Yoon Chung2, Jian Li2, Raouf Boutaba1,3, and James Won-Ki Hong1,2
1 IT Convergence Engineering, POSTECH, Pohang, Korea2 Dept. of Computer Science and Engineering, POSTECH, Pohang, Korea
3 School of Computer Science, University of Waterloo, Canada
The ability to reduce electricity usage and wastage through better demand-side management and control is considered a key solution ingredient to the global en-ergy crisis. One effective measure that has been put into place in many countries around the globe is the Demand Response (DR) program.
Request d with a starting time s and a ending time f
Schedule d to some-where within the speci-fied time frame
Assign d to the time-slots with the lowest cost among the candidate timeslots
Input: the schedule produced by the minMax algo-rithm
Find the peak cumulative cost
Shifts part of its demand forward in time filling in the time slots that are under-utilized
Repeat operation 2) and 3) until no shifting can be performed.
Server Stub: Web service interface to the client ap-plications and power utility
Services: a collection of GHS services such as me-tering, decision engine, security service, etc
Repository: manages metering data as well as de-vice specific information such as the adaptor-to-ap-pliance mapping.
Adapter: generates communication messages de-pending on the manufacturer’s message format and data model. Multitude of communication technolo-gies are enabled through the use of appliance spe-cific adapters.
Designed and implemented GHS Modeled the demand-side energy management prob-
lem (NP-hard) Designed a scheduling algorithm for demand side
energy management Showed that our algorithm can find near-optimal Showed the effect of battery on demand smoothing
Integrate the minMax algorithm in the Green-Home Service implementation
Conduct field-test experiment in real home and large enterprise settings.
Sources: Fortis Investments, www.urbanecoist.com
Introduction and Motivation minMax Algorithm
GHS Implementation
Simulation Study
Daily Energy Price from Utility
Variables Defined for Simulation
Variable Value Description
Energy Demand 1 ~ 5Energy that is required by each task per hour
Task Length 1 ~ 5The contiguous running time of each task
Shift Time 1 ~ 5Shift time range of each task
Daily Energy Price
3 ~ 10Energy price per one power unit in each hour
Scheduling Result with 10 Tasks
Scheduling Result with 1000 Tasks
Conclusion & Future Work
Conclusion
Future Work
…
Repository
Home Server
Operating System
ServerStub
Request Handler Error Handler Result HandlerInterfacePublisher
Services Control Service
Metering Service
Decision Engine
SecurityService
Hardware Comm.Module
PLC Zigbee WiFiCPU Memory
HDD
Config.Repository
DataRepository
App-Dev Mapping Metering Data
Adapter
Ethernet
Device 1Adapter
Device 2Adapter
Device 3Adapter
Adapter Interface
minMax Algorithm Overview
Algorithm Improvement using Battery
1 2 3 4 5 6 70
2
4
6
8
Time slot [t]
Cost
r2
r3
s1s2 s3 f2 f1,f3
l1 = 4
l2 = 3
l3 = 2
1 2 3 4 5 6 70
2
4
6
8
Time slot [t]
Cost
Demand-side Energy Management
Key Deterring Factors of Current Grid
Research Goals Design and Implement Green-Home Service (GHS)
architecture to provide advanced metering and con-trol
Design scheduling algorithm to provide decision making capabilities
Lack of information Lack of smart planning Customers are risk-averse
GHS Components