基于SCADA数据的风电机组电控系统故障诊断研究

发布时间：2018-08-01 12:45

【摘要】：风电机组的电控系统实现了对风电机组的控制,如自启动/停机、平稳并网、手动和自动的无扰动切换、变速变频控制、自动偏航控制、解缆控制、变桨距控制、数据监测和处理、自动保护以及故障记录等。随着当今风电产业快速发展,针对风电机组的电控系统故障频发的现象,本文对风电机组电控系统故障检测进行讨论和研究。论文通过采集和分析风速、风电机组运行参数等数据,针对电控系统中的变桨故障和主控PLC故障进行讨论。论文首先汇集了某风电场一年的完整数据,其次结合运行数据和维护数据整理出风电机组电控系统中的常见故障类别,再次分别运用时域分析方法、时频域分析中的小波分析方法和自适应Kurtosis分析三种方法进行电控系统故障的诊断分析,最后通过风场实际数据对上述方法进行仿真验证。运用时域分析方法、时频域分析和自适应Kurtosis分析三种方法进行电控系统故障的诊断,时域分析只能粗略的诊断设备是否异常,无法诊断故障部位。风电机组的电控系统中存在着许多的旋转部件,此类部件会产生一定的振动信号。这些振动信号具有周期变化但是不平稳的特性,本文采用的时频域分析法和自适应Kurtosis方法分析风电机组电控系统中的非平稳信号,效果较为明显,适用于电控系统故障诊断。在MATLAB平台下,结合以上的分析方法,通过分析功率比值等数据,并对比均值、方差、均方幅值和峰值等指标,得到了一定的故障检测规律。
[Abstract]:The electric control system of wind turbine has realized the control of wind turbine, such as self-start / shutdown, steady grid-connected, manual and automatic no-disturbance switching, variable speed frequency conversion control, automatic yaw control, decoupling control, variable pitch control, etc. Data monitoring and processing, automatic protection and fault recording, etc. With the rapid development of wind power industry, the trouble detection of wind turbine electronic control system is discussed and studied in this paper. By collecting and analyzing wind speed and operating parameters of wind turbine, the paper discusses the faults of variable propeller and main control PLC in electronic control system. Firstly, the paper collects the complete data of a wind farm for one year, then combines the operation data and maintenance data to sort out the common types of faults in the wind turbine electronic control system, and then uses the time domain analysis method, respectively. The wavelet analysis method and adaptive Kurtosis analysis method in time-frequency domain analysis are used to diagnose and analyze the fault of electronic control system. Finally, the simulation results of the above methods are carried out through the actual wind field data. Time domain analysis, time and frequency domain analysis and adaptive Kurtosis analysis are used to diagnose the fault of electronic control system. The time domain analysis can only roughly diagnose whether the equipment is abnormal or not. There are many rotating parts in the electric control system of wind turbine, which will produce certain vibration signals. These vibration signals have the characteristics of periodic variation but non-stationary. The time-frequency domain analysis method and adaptive Kurtosis method are used to analyze the non-stationary signals in the wind turbine electrical control system. The results are obvious and suitable for the fault diagnosis of the electronic control system. On the MATLAB platform, by analyzing the power ratio data and comparing the mean value, variance, mean square amplitude and peak value, a certain rule of fault detection is obtained.
【学位授予单位】：河北工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TM315

【参考文献】