混沌变异进化规划在水轮机调速器PID参数优化中的应用

A CHAOS MUTATION EVOLUTIONARY PROGRAMMING FOR OPTIMIZATION OF PID PARAMETERS OF HYDRO-TURBINE GOVERNING SYSTEMS

针对水轮发电机组非线性时变系统导致调速器在线最优参数难以整定的特点,将混沌动力学引入到进化规划的变异算子中,提出了一种混沌变异进化规划方法。该方法突破了传统的用固定分布的随机数实现变异的思想,采用类随机的确定性方法来构造变异算子,通过对实际系统的仿真,该算法能有效实现PID参数最优整定,系统具有稳定、超调量小、响应快等特点。

The mutation factors of some improved evolutionary programming (EP) methods are usually realized with Guassian or Cauchy random series. In consideration of the difficulties in setting the online optimal parameters of the governing system due to the nonlinear time-variant hydro-turbine governing system, this paper introduces chaos kinetics into the mutation factor of EP and gives a chaos mutation evolutionary programming (CMEP) method. The method breaks through the traditional idea of realizing mutation with a fixed random distribution and it uses a quasi-random finite way to formulate the mutation factor. In the simulation of a real system, the method can optimize PID parameters efficiently. The system has the characteristics of stability, low overshoot, and fast response.