基于混沌粒子群算法的水轮机调速系统参数辨识及建模试验

针对粒子群算法存在的后期收敛速度慢和易陷入局部最优等缺点,引入收缩因子和混沌优化思想对其改进,并将其应用于调速系统被控对象有关参数辨识问题上。提出一种水轮机调速系统参数辨识满意度函数设计的新方法,该方法直接计算系统响应的上升时间、调节时间、反调峰值功率、反调峰值时间等品质参数,并以系统总体满意度作为满意度函数。对某混流式水轮机调速器控制参数进行实测并对机组引水道参数进行辨识,试验结果表明:仿真数据能够准确模拟机组负荷的频率阶跃扰动响应,可以满足电网稳定性计算要求;在系统受到较大干扰时,该算法仍具有精确的参数辨识能力和很高的收敛效率。

Parameters Identification of Hydro-turbine Governing System Based on Chaos Particle Swarm Optimization and Modeling Experiment

To overcome the shortcomings of standard Particle Swarm Optimization (PSO), for example, prone to lo-cal optimum and slow later convergence and so on , shrinkage factor and chaos idea were adopted to improve stand-ard PSO in the study .A novel design method for satisfactory function of hydro-turbine governing system was put for-ward.Chaos PSO was applied to parameters identification of controlled object for governing system .Quality parame-ters, such as rise time, settling time, hydro-turbine’ s reverse peak power and reverse peak time , were directly measured , and the overall satisfaction level of system was taken as fitness function .On the basis of the new method , the control parameters of a hydro-turbine governor were measured in association with parameter identification of hy-droelectric turbine-conduit system .Test results show that the simulated data correctly reflect the response character-istics of cascade frequency disturbance for the unit load , and meet the requirements of power grid stability calcula-tion.Furthermore, under large interference , the algorithm still has accurate parameter identification and high con-vergence efficiency .