The application of genetic algorithms to gain-scheduling controller analysis and design

Gray, G.J., Li, Y. , Murray-Smith, D.J., Ronco, E. and Sharman, K.C. (1997) The application of genetic algorithms to gain-scheduling controller analysis and design. In: Genetic Algorithms in Engineering Systems: Innovations and Applications (GALESIA 97), Glasgow, UK, 2-4 Sep 1997, pp. 364-368. (doi:10.1049/cp:19971207)

Full text not currently available from Enlighten.

Publisher's URL:


Gain scheduling is an acknowledged design technique for the control of nonlinear systems. It operates by switching between a network of linear controllers according to the system operating point. The genetic algorithm (GA) can be applied to the design of such a network. The GA is used to optimise the activation points of the individual controllers. GA coding for such a scheduling problem is discussed. An exhaustive search to establish the optimum number of controllers coupled with optimisation of the corresponding activation points by GA shows the relationship between controller performance and complexity. The GA designed gain scheduling network is compared with another gain scheduling design technique and the activation points identified by GA are used to determine the degree of nonlinearity of the plant throughout its operating region.

Item Type:Conference Proceedings
Additional Information:IEE Conference Publication No. 446. Published by IEE, 1997. ISBN: 0852966398.
Keywords:Control, gain-scheduling, optimisation, genetic algorithm
Glasgow Author(s) Enlighten ID:Murray-Smith, Professor David and Li, Professor Yun
Authors: Gray, G.J., Li, Y., Murray-Smith, D.J., Ronco, E., and Sharman, K.C.
Subjects:Q Science > QA Mathematics > QA76 Computer software
T Technology > TJ Mechanical engineering and machinery
T Technology > TK Electrical engineering. Electronics Nuclear engineering
College/School:College of Science and Engineering > School of Engineering > Systems Power and Energy

University Staff: Request a correction | Enlighten Editors: Update this record