基于小波神经网络的燃气轮机后备电源运行故障诊断方法  

Failure Diagnosis Method of Gas Turbine Backup Power Supply Operation Based on Wavelet Neural Network

作　　者：李元俊 LI Yuanjun(Tianjin Chentang Thermal Power Co.,Ltd.,Tianjin 300000,China)

机构地区：[1]天津陈塘热电有限公司,天津300000

出　　处：《通信电源技术》2024年第18期118-120,共3页Telecom Power Technology

摘　　要：为提升燃气轮机后备电源运行故障诊断的准确性,提出基于小波神经网络的燃气轮机后备电源运行故障诊断方法研究。首先通过传感器与数据采集卡采集电源运行参数;其次利用小波变换方法提取故障特征;最后构建小波神经网络模型,实现对燃气轮机后备电源运行故障的全方位诊断。实验结果表明,提出方法应用后,表现出了较高的故障诊断准确率和较低的识别误差,有效提高了诊断的精确度。To improve the accuracy of fault diagnosis for gas turbine backup power supply operation,this study proposes a wavelet neural network-based fault diagnosis method for gas turbine backup power supply operation.First,collect power operation parameters through sensors and data acquisition cards.Then use wavelet transform method to extract fault features.Finally,a wavelet neural network model is constructed to achieve comprehensive diagnosis of operational faults in the backup power supply of gas turbines.The experimental results show that the proposed method exhibits high fault diagnosis accuracy and low recognition error after application,effectively improving the accuracy of diagnosis.

关 键 词：小波神经网络 燃气轮机 后备电源 

分 类 号：TN86[电子电信—信息与通信工程]