Controllability analysis of multi objective control systems

Stewart, Paul, Jewell, G. W., Clarke, R. E. and Fleming, P. J. (2002) Controllability analysis of multi objective control systems. In: Proceedings of the 2002 IEEE International Symposium on Computer Aided Control System Design, 18-20 September 2002, Glasgow, Scotland, UK.

Documents
Controllability analysis of multi objective control systems
The performance requirements stated in project specifications often comprise conflicting objectives. These objectives may further be a complex mix of steady state and dynamic performance. Control devices such as solenoid actuators are often chosen purely on steady state force characteristics, due to the difficulty of appraising the conflicting and generally non-linear nature of the performance objectives. This can have ramifications in terms not only of the actuator performance, but also in the overall controllability of the system when closed-loop control is implemented. An example automotive application examining the multi objective controllability of electronically actuated valves is presented. Multi objective evolutionary techniques are utilised to derive the optimal force-displacement characteristics and also dynamic characteristics of the desired actuator under the constraint of design performance criteria. The selected actuator is then assessed for its controllability and dynamic performance.
[img]
[Download]
[img]
Preview
PDF
ELVAS_upload.pdf - Whole Document

4MB
Item Type:Conference or Workshop contribution (Paper)
Item Status:Live Archive

Abstract

The performance requirements stated in project specifications often comprise conflicting objectives. These objectives may further be a complex mix of steady state and dynamic performance. Control devices such as solenoid actuators are often chosen purely on steady state force characteristics, due to the difficulty of appraising the conflicting and generally non-linear nature of the performance objectives. This can have ramifications in terms not only of the actuator performance, but also in the overall controllability of the system when closed-loop control is implemented. An example automotive application examining the multi objective controllability of electronically actuated valves is presented. Multi objective evolutionary techniques are utilised to derive the optimal force-displacement characteristics and also dynamic characteristics of the desired actuator under the constraint of design performance criteria. The selected actuator is then assessed for its controllability and dynamic performance.

Additional Information:The performance requirements stated in project specifications often comprise conflicting objectives. These objectives may further be a complex mix of steady state and dynamic performance. Control devices such as solenoid actuators are often chosen purely on steady state force characteristics, due to the difficulty of appraising the conflicting and generally non-linear nature of the performance objectives. This can have ramifications in terms not only of the actuator performance, but also in the overall controllability of the system when closed-loop control is implemented. An example automotive application examining the multi objective controllability of electronically actuated valves is presented. Multi objective evolutionary techniques are utilised to derive the optimal force-displacement characteristics and also dynamic characteristics of the desired actuator under the constraint of design performance criteria. The selected actuator is then assessed for its controllability and dynamic performance.
Keywords:Multiobjective evolutionary algorithm, control systems analysis and design, Robust design
Subjects:G Mathematical and Computer Sciences > G700 Artificial Intelligence
H Engineering > H330 Automotive Engineering
H Engineering > H360 Electromechanical Engineering
G Mathematical and Computer Sciences > G530 Systems Analysis and Design
Divisions:College of Science > School of Engineering
ID Code:2199
Deposited On:02 Mar 2010 23:51

Repository Staff Only: item control page