The application of improved genetic algorithm in nonlinear PID control

A method that integrates nonlinear PID control and genetic algorithm is introduced for the drawback of traditional PID control. It combines the global convergence of genetic algorithm and self-adaptive capacity of nonlinear PID and enhances the accuracy and robustness of the control. In the genetic operations, a crossover program related to the individual fitness is designed based on random non-uniform linear cross to prevent the loss of good genes after cross. A scheme that presents a crossover rate and mutation rate can adjust adaptively responds to the evolutionary process is provided, which ensures the global search capability during prophase and population diversity during later period. The scheme was applied to the control of one order time-delay system and simulation results show that the speed, stability and accuracy of the controlled system is improved.