基于机器学习的风力发电变桨系统自适应容错控制方法

The adaptive fault tolerant control method for wind power pitch system based on machine learning

针对风力发电变桨系统自动控制误差计算错误、不能自动容错而导致停止运行的问题,研究基于机器学习的风力发电变桨系统自适应容错控制方法。简化向量矩阵的计算过程,保证计算控制器跟踪误差的计算准确率,设计控制器补偿信号的模型结构,定义风力发电系统的真实控制率,基于机器学习优化容错算法,建立风力发电变桨系统容错控制自适应逼近模型。实验结果表明,在15组重复性实验中,实验组功率超出临界值的次数共有74次,远小于对照组,表明该方法能够更好地保证机组的正常运行,从而提高机组的稳定性和可靠性。

Aiming at the issue of automatic control error calculation errors and inability to automatically tolerate faults in wind power pitch systems,a machine learning based adaptive fault-tolerant control method for wind power pitch systems is studied.By simplifing the design process of vector matrix and ensuring the calculation accuracy of controller tracking error,it designs the model structure of controller compensation signal,defines the true control rate of wind power generation system,and establishes an adaptive approximation model for fault-tolerant control of wind power generation pitch system based on machine learning optimization fault-tolerant algorithm.The experimental results show that in 15 repetitive experiments,the number of times that the experimental group exceeds the critical power threshold is 74,which is much lower than the control group.This indicates that the proposed method can better ensure the normal operation of the wind turbine system,thereby improv the stability and reliability of the system.