基于核主元分析的风电机组变桨距系统故障诊断研究

发布时间：2019-01-28 22:41

【摘要】：变桨距系统是风电机组控制系统的重要组成部分,其运行状态直接关系到风电机组是否安全可靠运行。风电机组Supervisory Control and Data Acquisition(SCADA)系统具有数据采集与状态监测功能,能够实时监控风电机组变桨距系统的运行状态。然而,由于变桨距系统具有复杂的机电结构,导致变桨距系统的故障之间可能存在连锁反应或者相互影响,所以很多时候维修人员无法通过风电机组SCADA系统及时准确地判断出引发变桨距系统故障的故障源。因此,开展变桨距系统故障诊断研究具有重要的学术意义和应用价值。本文针对变桨距系统的相关参数多、非线性及精确建模困难等问题,从风电机组SCADA系统监测的风电机组运行数据出发,利用基于核主元分析(Kernel Principal Component Analysis,KPCA)的数据挖掘方法,进行风电机组变桨距系统故障检测与辨识的研究,实现了基于KPCA的风电机组变桨距系统故障诊断,仿真结果验证了该方法的有效性。本文的主要研究工作如下:(1)本文对风电机组变桨距系统典型故障的故障模式、故障原因以及故障间的关系进行了分析。由分析结果得出,变桨距系统是一个故障率高、相关运行参数多且相互耦合以及故障形式复杂的非线性系统,为开展变桨距系统故障诊断研究奠定了理论基础。(2)为了提高基于KPCA的变桨距系统故障诊断方法的快速性和准确性,本文分析了风电机组SCADA系统监测到的与变桨距系统相关的运行参数,应用Relief算法选择出一些最具代表性、分类性能最好的变桨距系统故障特征变量,构建了变桨距系统观测向量。(3)核函数参数的最优化对基于KPCA的变桨距系统故障诊断方法至关重要,本文应用了基于粒子群优化(Particle Swarm Optimization,PSO)的核函数参数寻优方法,获得了最优核函数参数;以风电机组SCADA系统监测的变桨距系统观测向量运行数据为基础,提出了基于KPCA的变桨距系统故障诊断方法,并进行变桨距系统的故障检测与辨识,实现了变桨距系统的故障诊断;应用风电机组SCADA系统监测到的故障信息开展了仿真研究,验证了基于KPCA的风电机组变桨距系统故障诊断方法的有效性。
[Abstract]:Variable pitch system is an important part of wind turbine control system, and its running state is directly related to the safe and reliable operation of wind turbine. The Supervisory Control and Data Acquisition (SCADA) system of wind turbine has the functions of data acquisition and condition monitoring, and can monitor the running state of variable pitch system of wind turbine in real time. However, because of the complex electromechanical structure of the variable pitch system, there may be a chain reaction or interaction between the faults of the variable pitch system. So many times the maintainers can not accurately determine the fault source of the variable pitch system through the wind turbine SCADA system. Therefore, the research on fault diagnosis of variable pitch system has important academic significance and application value. In this paper, aiming at the problems of variable pitch system, such as many related parameters, nonlinear and accurate modeling, the data mining method based on kernel principal component analysis (Kernel Principal Component Analysis,KPCA) is used to solve the problem of wind turbine operation data monitored by SCADA system of wind turbine. The fault detection and identification of variable pitch system of wind turbine is studied, and the fault diagnosis of variable pitch system of wind turbine based on KPCA is realized. The simulation results verify the effectiveness of the method. The main work of this paper is as follows: (1) this paper analyzes the typical fault modes, fault causes and the relationship between the faults of wind turbine variable pitch system. From the analysis results, it is concluded that the variable pitch system is a nonlinear system with high failure rate, multiple related operating parameters, mutual coupling and complex fault forms. It lays a theoretical foundation for fault diagnosis of variable pitch system. (2) in order to improve the speed and accuracy of fault diagnosis method of variable pitch system based on KPCA, In this paper, the operating parameters related to the variable pitch system monitored by SCADA system of wind turbine are analyzed, and some of the most representative and best classification characteristic variables of variable pitch system are selected by using Relief algorithm. The observation vector of variable pitch system is constructed. (3) the optimization of kernel function parameters is very important to the fault diagnosis method of variable pitch system based on KPCA. The kernel function parameter optimization method based on particle swarm optimization (Particle Swarm Optimization,PSO) is applied in this paper. The optimal kernel function parameters are obtained. Based on the observation vector running data of variable pitch system monitored by SCADA system of wind turbine, the fault diagnosis method of variable pitch system based on KPCA is put forward, and the fault detection and identification of variable pitch system are carried out. The fault diagnosis of variable pitch system is realized. The fault information monitored by wind turbine SCADA system is simulated and studied, which verifies the effectiveness of the fault diagnosis method of wind turbine variable pitch system based on KPCA.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM315

【参考文献】