ApplicationParallel Processing in

Distributed Real Time Control of Aircraft Testing

Prepared by :Rohayu Mohamed

Siti Nor Fatin GhazmiSiti Fatnin Mohd Na’izam

Introduction

• An innovative multi-channel digital control system for durability testing of aircraft has been developed at the Defence Science and Technology Organisation, Australia.

• It is currently in use on a full-.wale durability test of the FIA-18 aircraft for the Royal Australian Air Force.

• The system using parallel processing and inter-processor communication.

• This system is based on a network of controller processors physically distributed over the test rig

• One processor at each actuator station with a single system management processor

• All linked by fast inter-processor communications for system state and control state-variable data.

• Parallel processing is used to implement system management. Multiple Input Multiple Output (MIMO) control, data acquisition and communications.

• Full-scale aircraft durability tests are one-off tests whose control requirements typically change as the test rig is developed. Once commissioned, tests often run for several years.

• The system described here was motivated by a need to reduce the set-up and running time of tests by providing flexibility during test rig development and maximising testing speed

Parallel architecture used

• This system is based on a network of controller processors physically distributed over the test rig, with one processor at each actuator station. and a single system management processor; all linked by fast inter-processor communications for system state and control state-variable data

• It provide high-speed synchronous MIMO control at update rates to 1OOOHz. Up to 511 separate controllers. each using one Inmos T805 Transputer as a processor, are connected in a binary tree to make a single distributed control system.

• A special hardware broadcast link spans the whole network tree and provides critical real-time communications and synchronisation.

• Controllers are located adjacent to actuators. with up to 20 metres between any pair of directly connected controllers. Pneumatic, hydraulic and electromagnetic actuators are supported, with appropriate control laws.

The advantages of using the proposed architecture

• To provide high-speed synchronous MIMO control at update rates to 1OOOHz.

• To control systems for aircraft testing consists of a centralised m y of independent PID style controllers with command inputs provided by a central computer.

• Parallel processing is used to implement system management. Multiple Input Multiple Output (MIMO) control, data acquisition and communications

Conclusion

• The exercise of desiping, implementing, and applying this system to a number of small tests and one large test has confirmed the suitability of the distributed parallel processing model described above. Performance targets have been met and, as intended. the system has proved to be easy to install and reconfigure. Enhancements envisaged for the current system include more built-in support for system identification and improvements in the way new control law software is generated. with direct links to a control system modelling and analysis software package such as Matlab or MATRIX,. While design goals have been met with the cumnt hardware, implementation of the software would have ben easier with a faster processor and faster communications.