【摘要】 针对遗传算法(GA)收敛速度慢,不利于在实时控制中应用这一问题,构造出一种快速收敛的混合遗传算法(HGA),该算法利用遗传算法的全局搜索能力,并采用N e lder-M ead单纯形法来加强算法的局部搜索能力,加快了算法的收敛效率.将基于该混合遗传算法的模型参考自适应控制方法引入连续搅拌反应釜(CSTR)这一复杂的非线性系统,根据参考模型的输出,通过混合遗传算法对控制系统的P ID参数进行在线寻优和在线调整,以达到参考模型所要求的控制效果,仿真结果表明了该方法的良好控制性能.更多还原

【Abstract】 Premature convergence and low converging speed are the distinct weaknesses of genetic algorithms(GA),which result in their poor performance in real-time controls system.Based on the combination of Nelder-Mead Simplex and GA,a hybrid genetic algorithm(HGA) that can quickly converge to the optimal set is proposed.The global search capability of GA is utilized to explore a wide search space and detect a promising area,and the Nelder-Mead Simplex is used to strengthen the local search and fast convergence of the HGA.A new model reference adaptive control(MRAC) method based on HGA is presented,and an application of this method to the continuous stirred tank reactor(CSTR) system which is a complex nonlinear system,is studied.Using the novel hybrid genetic algorithm to optimize the on-line PID′s control parameters,desired control effect is obtained.Simulation result shows the validity of this method.更多还原