基于自适应遗传算法的多目标PID优化设计

Multi-objective optimization of PID regulators based on adaptive genetic algorithms

提出一种基于自适应遗传算法的多目标PID优化设计方法。采用染色体实数编码和具有自适应交叉概率和变异概率的遗传算法对PID参数寻优,有效地提高了遗传算法的全局搜索能力和收敛速度。通过在适应度函数中引入表示超调量、上升时间和稳态误差的指标项,并对指标项适当加权,可使优化后的PID调节器的综合性能达到满意程度。仿真结果表明,该PID调节器的性能优于常规方法获得的PID调节器。

A novel multi-objective optimization method for parameter tuning of PID regulator based on adaptive genetic algorithms is proposed. By using real value encoding and adaptive crossover and mutation probabilities, the global searching ability and the convergence speed of the genetic algorithms are significantly improved. With the introduction of the terms which represent overshoot, rise time and steady error of the system in the fitness function and properly weighting these terms, the overall performance of the PID regulator is optimized to satisfaction. Simulation results show that the performance of the optimized PID regulator is superior to that of the conventional PID regulator.