基于改进遗传算法的时滞系统辨识与控制

Smith predictor is a common method to control the system with pure delay, but it requires high accurancy of the model. In this paper a fuzzy adaptive genetic algorithm is proposed to identify the dynamic model parameters especially the pure delay accurately. In order to speed up the algorithm and avoid the precocity, it adopts float number coding and proper selecting, crossover and mutation operation, and auto-adjusts crossover and mutation probability with the evolutionary process and the diversity of the population based on fuzzy rules. Simulation shows the Smith predictor using the model identified by the advanced genetic algorithm can overcome the effect of the pure delay effectively and get good control performance.