风电机组发电机前轴承健康度预测方法及实现  

Prediction Method for Health Degree of Front Bearing of Wind Turbine Generator and Implementation

作　　者：尹诗[1,2] 侯国莲 迟岩[2] 弓林娟 胡晓东 Yin Shi;Hou Guolian;Chi Yan;Gong Linjuan;Hu Xiaodong(School of Control and Computer Engineering,North China Electric Power University,Beijing 102206,China;Zhong Neng Power-Tech Development Co.,LTD,Beijing 100034,China)

机构地区：[1]华北电力大学控制与计算机工程学院,北京102206 [2]中能电力科技开发有限公司,北京100034

出　　处：《系统仿真学报》2021年第6期1323-1333,共11页Journal of System Simulation

基　　金：国家重点研发计划(2019YFB1505402);国家自然科学基金(61973116)。

摘　　要：针对双馈式风电机组发电机前轴承劣化趋势问题,提出了一种新的组合建模方法对发电机前轴承健康度进行趋势预测。采用高斯混合模型(Gaussian Mixture Model, GMM)对机组运行工况进行辨识,并在各个子工况内分别建立基于极限学习机(Extreme Learning Machine, ELM)的发电机前轴承温度模型,将温度残差特征与前轴承振动信号时频域特征相融合,并计算前轴承健康度,提出基于注意力机制的双向长短期记忆(Bi-directional Long Short Term Memory, Bi-LSTM)神经网络对前轴承健康度进行建模并预测其趋势。实验结果表明:该组合建模方法具有较高的准确度和泛化能力。Aiming at the deterioration trend of front bearing of doubly-fed wind turbine generator, a new combined modeling method is proposed to predict health degree of front bearing of generator. The GMM is used to identify operating conditions of wind turbines. The temperature model of front bearing based on ELM is established respectively in each sub-condition. Combining with temperature residual characteristics and time-frequency characteristics of vibration signal, the health degree of front bearing is calculated. Based on attention mechanism, the Bi-LSTM neural network is proposed to model and predict health degree of front bearing. The result shows that the combined modeling method has high accuracy and generalization ability.

关 键 词：风电机组 发电机前轴承 注意力机制 双向长短期记忆 健康度 

分 类 号：TK83[动力工程及工程热物理—流体机械及工程] TP391.9[自动化与计算机技术—计算机应用技术]