小波包和RBF神经网络的燃气调压器故障诊断  

Fault Diagnosis of Gas Pressure Regulator Based on Wavelet Packet and RBF Neural Network

作　　者：刘力宁 郝学军 刘旭海 LIU Lining;HAO Xuejun;LIU Xuhai

机构地区：[1]北京建筑大学环境与能源工程学院,北京100044

出　　处：《煤气与热力》2018年第7期36-39,43,共5页Gas & Heat

摘　　要：提出一种基于小波包和径向基函数神经网络(RBF神经网络)的高-中压燃气调压器故障诊断智能化的方法。调压器出口压力信号通过小波包分解为能量信号,将能量信号和燃气调压器的稳压精度相结合构建RBF神经网络的输入特征向量,通过分析RBF神经网络的输出向量实现高-中压燃气调压器故障类型的智能诊断。An intelligent fault diagnosis method for high-medium pressure gas regulators based on wavelet packet and radial basis function （RBF） neural network is proposed. The outlet pressure signal of the regulator is decomposed into energy signals by the wavelet packet, and the input feature vector of the RBF neural network is constructed by combining the energy signal with the pressure stabilizing precision of the gas regulator. The intelligent diagnosis of fault types of high-medium pressure gas regulators is realized by analyzing the output vector of the RBF neural network.

关 键 词：小波包 RBF神经网络 燃气调压器 稳压精度 故障智能诊断 

分 类 号：TU996.8[建筑科学—供热、供燃气、通风及空调工程]