基于NSET的风力发电机后轴承温度预警方法

采用非线性状态估计(NSET)方法对风力发电机的后轴承温度进行建模与分析;并采用Relief方法对特征参数进行选择,减少了NSET建模的复杂性。通过所建模型预测发电机后轴承温度,将其与正常运行温度进行比较;并利用滑动窗口对残差均值进行比较。若超出阈值范围,则发电机出现故障。验证结果表明,非线性状态估计(NSET)方法可以准确预测发电机后轴承温度,实现发电机运行状态的预警监测。

The rear bearing temperature of early warning method in wind turbine based on NSET

Nonlinear state estimation (NSET) method is used for wind turbine rear bearing temperature on modeling and analysis of characteristic parameters are chosen and the Relief method, reduces the complexity of NSET modeling. Through the model prediction generator bearing temperature, compared with the normal operating temperature, and by using the sliding window residual mean comparison, if beyond the threshold, the generator fault occurs. Verification results show that nonlinear state estimation (NSET) method can accurately predict generator rear bearing temperature, realize the generator running state of early warning and monitoring.