摘要
主轴轴承是风电机组的重要部件之一。通常主轴轴承故障诊断方法主要是基于振动信号和温度信号以及润滑油成分分析等。这里利用支持向量机建立了风电机组发电机输出功率模型,输入量为风速、变桨角度、风向角与机舱角偏差;输出量为发电机输出有功功率。在相同输入条件下,当主轴轴承存在磨损等故障时,发电机输出有功功率将随故障的逐步加重而逐渐减小,发电机输出有功功率实际值与预测值之间的残差将超出正常的阈值。这里以某风电场风机主轴轴承实际故障进行了仿真验证。
Spindle bearing is one of the main failure parts in wind generator. At present, mostly fault diagnosis methods of spindle bearing are based on vibration signal, temperature signal and composition analysis of lubricating oil, etc. In this paper, using svm to establish model. Its input variables include wind speed, blade angle and the deviation of wind angle and plane angle. The output variable is the power of generator. Under the condition of the same inputs, there is bearing failure such as wear and tear of main spindle, along with the gradually increasing of the fault, the output power of generator will fall. Residuals between power of generator actual value and predictive value will be beyond the normal threshold. In this paper, there will be a simulation and verification based on real fault of spindle bearing in a wind turbine.
出处
《仪器仪表用户》
2016年第11期88-92,共5页
Instrumentation
