基于遗传算法的分布参数对象PID控制器设计

实际工业过程中的对象大都属于分布参数系统,控制现场主要采用比例-积分-微分(PID)控制方法。该文将基于遗传算法的PID控制器优化设计方法(简称GAOPT方法)应用到一类分布参数对象上,设计了最优控制器,并与几种基于常规整定公式的控制器进行比较。仿真结果表明,GAOPT方法设计的PID控制器,可以利用较小的控制能量,获得在超调量、调节时间和时间乘绝对误差积分(ITAE)等指标上都较优的控制效果。将GAOPT方法应用到分布参数系统上,可以提高现有工业中PID控制器的控制水平。

PID controller design for distribution parameter systems based on genetic algorithms

Many objects in practical industrial process are distributed parameter systems (DPS), with proportional-integral-derivative (PID) controllers used as the main control method. Genetic algorithms were used to optimize PID controllers (GAOPT) for distributed parameter systems. Simulations showed that controllers designed with the genetic algorithms achieve lower overshoot, shorter adjustment times and smaller integral of time-multiplied absolute value of error (ITAE) criteria even for the condition of minimizing the control energy than designs based on general tuning formulas.