基于自适应遗传算法的风电机组变桨控制参数自寻优方法  

Self-optimization of Wind Turbine Variable-pitch Control Parameters Based on Adaptive Genetic Algorithm

作　　者：常晟 万宇宾 蒋韬 CHANG Sheng;WAN Yubin;JIANG Tao(CRRC Zhuzhou Institute Co.,Ltd.,Zhuzhou,Hunan 412001,China)

机构地区：[1]中车株洲电力机车研究所有限公司,湖南株洲412001

出　　处：《控制与信息技术》2024年第2期40-48,共9页CONTROL AND INFORMATION TECHNOLOGY

基　　金：国家重点研发计划项目(2022YFB4201600)。

摘　　要：当前,风力发电机组变桨系统采用传统PID参数整定方法,整定效率低、精确度和适应性差。为此,文章提出一种基于自适应遗传算法的风力发电机组变桨控制参数自寻优方法,包括变增益变桨PID参数自寻优和塔架加阻参数自寻优。其首先阐述了变桨控制系统基本原理;接着基于遗传算法根据风力发电机组不同机型对约束条件进行规划设计,涉及初始种群选择、自寻优算法约束及适应度函数设计;最后对传统参数整定方法与自寻优参数整定方法的响应曲线、仿真及现场运行结果进行对比分析。结果表明,相较传统方法整定的PID参数,自寻优变桨参数下发电机转速波动约降低10%,塔底My疲劳载荷约降低2%,对整定效率、精确度及适应性具有较好的改善效果。For the low parameter tuning efficiency,accuracy and adaptability of the traditional PID parameter tuning method currently adopted in the wind turbine variable-pitch system,this paper presents a method for self-optimization of wind turbine variable-pitch control parameters based on adaptive genetic algorithm,including variable-gain variable-pitch PID parameter selfoptimization and tower damping parameter self-optimization.The author first described the basic principle of variable-pitch control system;then based on the genetic algorithm,planned and designed constraint conditions for different wind turbine models,covering initial population selection,self-optimization algorithm constraints and adaptability function design;and finally compared and analyzed the response curve,simulation and field operation results of the traditional parameter tuning method and self-optimization parameter tuning method.The results show that,compared with the traditional parameter tuning method,the self-optimization parameter tuning method reduces the motor speed fluctuation by about 10%,and the fatigue My load of the tower bottom by about 2%.This shows the self-optimization parameter tuning method can improve the tuning efficiency,accuracy and adaptability.

关 键 词：风力发电机组 变增益变桨PID控制 遗传算法 自动寻优 转速波动 塔底My疲劳载荷 

分 类 号：TM315[电气工程—电机]