The Parameter Optimization of Human-Simulated Intelligent Controller Based on Improved Genetic Algorithm
Guiqiang Chen, Linjian Tang, Yan Tang · 2009
The human-simulated intelligent controller, using a hierarchical and multi-mode control structure, effectively solves the control difficult problems of many complex systems. But there are many parameters to be determined. In order to solve them, the paper introduces an improved genetic algorithm (IGA) which is combined with orthogonal design method. It is served as the simulation experiment platform by the cart-double pendulum. Through selected the reasonable fitness function and determined the appropriate parameters range, using the IGA, the parameters of the human-simulated intelligent controller are optimized. The simulation results show the IGA is better than the manual adjustment method. So it is proved that the IGA is more effective in solving the parameter optimization of human-simulated intelligent controller, and also promotes the more extended application space of intelligent control theory.