基于振动传感器的风力发电机故障检测算法

Fault Detection Algorithm of Wind Turbine Based on Vibration Sensor

风力发电机长时间处于高应力状态下,易产生设备细小故障,若不能及时找出,会造成风电系统严重受损或瘫痪,为此提出了基于振动传感器的风力发电机故障检测算法。利用向量索引值集合及广义逆向矩阵,在压缩感知下匹配追踪信号,恢复缺失数值,经过经验模态分解后,加入白噪声求解信号分量平均值,识别信号的振动类型和能量变化,判断信号故障类型,利用逻辑回归模型分类信号是否属于故障信号,实现风力发电机故障信号分类及判断。仿真结果表明,所提算法可以准确区分正常工作状态和故障状态,迭代30次后可以将故障检测误差率保持在0.035。由此可知,所提算法能准确判断故障及所属类型,且精准度高。

When the wind turbine is under high stress for a long time,it is easy to produce small equipment faults.If it can not be found out in time,it will cause serious damage or paralysis of the wind power system.Therefore,a wind turbine fault detection algorithm based on vibration sensor is proposed.The vector index value set and generalized inverse matrix are used to match the tracking signal under compressed sensing and recover the missing value.After empirical mode decomposition,white noise is added to solve the average value of signal components,identify the vibration type and energy change of signal,judge the signal fault type,classify whether the signal belongs to fault signal by logistic regression model,and finally realize the classification and judgment of wind turbine fault signal.The simulation results show that the proposed algorithm can accurately distinguish between normal working state and fault state,and the error rate of fault detection can be maintained at 0.035 after 30 iterations.Therefore,the proposed algorithm can accurately judge the fault and its type,and has high accuracy of detection.