An investigation of evolutionary computing in systems identification for preliminary design

This research investigates the integration of evolutionary techniques for symbolic regression. In particular the genetic programming paradigm is used together with other evolutionary computational techniques to develop novel approaches to the improvement of areas of simple preliminary design software using empirical data sets. I t is shown that within this problem domain, conventional genetic programming suffers from several limitations, which are overcome by the introduction of an improved genetic programming strategy based on node complexity values, and utilising a steady state algorithm with subpopulations. A further extension to the new technique is introduced which incorporates a genetic algorithm to aid the search within continuous problem spaces, increasing the robustness of the new method. The work presented here represents an advance in the Geld of genetic programming for symbolic regression with significant improvements over the conventional genetic programming approach. Such improvement is illustrated by extensive experimentation utilising both simple test functions and real-world design examples. A C K N O W L E D G M E N T S Many thanks to Dr. Ian Parmee for his guidance and motivation throughout this research. 1 am also grateful to Graham Purchase from Rolls Royce pic. and the EPSRC (UK) who jointly ftinded the research. To Jo, Bryony and Beth. AUTHORS D E C L A R A T I O N At no lime during the registration for the degree of Doctor of Philosophy has the Author been registered for any other University award. This study was financed with the aid of an EPSRC/RR CASE studentship. Industrial Collaborator : Rolls Royce Pic. The research has been presented by the author at 7 relevant international technical conferences. Rolls Royce Pic. has been visited regularly for technical discussions and consultation purposes. PUBLICATIONS: Watson, A. H.; Parmee, 1. C. Systems Identification Using Genetic Programming. Proceedings of ACEDC: p248-255. University of Plymouth. 1996. Watson, A. H.; Parmee, I . C. Identification Of Ruid Systems Using Genetic Programming. Proceedings of the 2nd Online Workshop on Evolutionary Computation. p45-48. On the Inemei WWW served by Nagoya University. 1996. Watson, A. H.; Parmee, I . C. Identification Of Fluid Systems Using Genetic Programming. Proceedings EUFTT '96. Vol I , p395-399. ELITE Foundation. 1996. Watson, A.H. ; Parmee, I . C. Steady Slate Genetic Programming with Constrained Complexity Crossover. Procs. 2nd Annual Conference Genetic Programming (GP'97) MIT Press, 1997. Watson, A.H. ; Parmee, I . C. Steady State Genetic Programming with Constrained Complexity Crossover Using Species Sub-Populations. Procs. 7th International Conference on Genetic AJgorithms. (ICGA 97). Morgan Kaufmann. 1997. Watson, A. H.; Parmee, I . C. An Improved Genetic Programming Strategy For Preliminary Design Model Development. Proceedings EUFIT '91. Vol I . p682-686. ELITE Foundation. 1997. Watson, A. H.; Parmee, I . C. Improving Engineering Design Models Using An Alternative Genetic Programming Approach. Proceedings of ACDM, pl93-206. Springer. 1998. PRESENTATIONS AND C O N F E R E N C E S A T T E N D E D : ACEDC '96 2nd International Conference On Adaptive Computing In Engineering Design and Control. University of Plymouth, UK March 1996. WEC *96 2nd Online Workshop On Evolutionary Computation. WWW Served by Nagoya University, March 4-22 1996. EUFIT *96 4th European Congress on Intelligent Techniques and Soft Computing. Aachen, Germany, September 2-5 1996. GP *97 PhD Student Workshop. Stanford University, San Francisco, USA, July 12th 1997. GP *97 2nd Annual Conference on Genetic Programming. Stanford University, San Fransisco, USA, July 13th-16th 1997. ICGA *97 7th International Conference on Genetic Algorithms. Michigan State University, Michigan, USA, Julyl9th-23rd 1997. EUFIT '97 5th European Congress on Intelligent Techniques & Soft Computing. Aachen, Germany, September 8-11 1997. ACDM *98 3rd International conference on Adaptive Computing In engineering Design and Manufacture. University of Plymouth, UK April 1998. Director of Studies: Dr I . C. Parmee, Director, Pltmouth Engineering Design Centre, University of Plymouth. Industrial Supervisor: Graham Purchase Chief Design Engineer, Turbines, Rolls Royce, pic. Filton Bristol.