Extensive Experiments on Genetic Algorithms for the Optimization of Piezoelectric Actuator Locations

摘要: For shape control of smart structures, we have developed a series micro-genetic-algorithms for optimal placement large number piezoelectric actuators. Here, investigate the effect on solution quality 1) different random seed generators; 2) restart criteria in micro-genetic-algorithms, specifically two parameters, generations used inner loop and level diversity population; 3) numbers We also report comparison our genetic algorithms with heuristic integer programming algorithms: worst-out-best-in exhaustive single point substitution proposed by Haftka Adelman ("Selection Actuator Locations Static Shape Control Large Space Structures Heuristic Integer Programming," Computers Structures, Vol. 20, 1985, pp. 572-582). Using current algorithms, not only get better layouts actuators than reported previous publications, but find that most distinct nature is randomness robustness. The best parameter setting dependent both evaluations termination generator used. Moreover, varies as changes. To highest solutions, multiple runs using generators are necessary. time investigation can be significantly reduced coarse grain parallel computing. Comparison shows group usually converge faster first few thousand more likely to solutions algorithms.