演化非线性参数估计在PID控制器设计中的应用研究  

Application of Nonlinear Regression Based on Evolution Computation in the Design of PID Controllers

作　　者：胡健闻[1] 曾益[1] 柴苍修[1] 

机构地区：[1]武汉理工大学机电工程学院,湖北武汉430070

出　　处：《计算机仿真》2004年第5期86-88,共3页Computer Simulation

摘　　要：在目前的工业过程控制中,采用最多的控制策略依然是PID方式。这是由于PID控制器具有简单而固定的形式,较好的鲁棒性。PID参数复杂的整定问题一直是这一领域中的一个重要的技术问题。该文将由阶跃响应求取受控对象模型的过程看作是一个非线性回归问题,讨论了用演化算法来进行参数估计的方法和步骤。数值算例表明:该方法具有良好的计算效率和精度。这种方法也适用于一般的传递函数模型。In current industrial process control, PID method is still the most widely used control strategy, since it has a simple, stable form as well as good robustness. However, its complex parameters tuning is an important technical problem in this area. This paper deems it a nonlinear regression problem to acquire the model of controlled object through its step respond and then, discusses the approach and step for parameters estimation based on Evolution Algorithm. A numerical instance shows the above-mentioned method has good computational efficiency and precision. Also, it is applicable to more general models of transfer function.

关 键 词：PID控制器 非线性回归 演化算法 参数估计 传递函数模型 

分 类 号：TM571[电气工程—电器]