基于数据驱动的空压机温度过高故障诊断仿真

Fault Diagnosis and Simulation of Air Compressor Over Temperature Based on Data Drive

随着空压机在工业生产中的广泛应用,对空压机运行过程中的温度进行准确地监测,已成为整个化工行业重点研究的课题,为此提出基于数据驱动的空压机温度过高故障诊断系统。首先,设计了由远程终端单元、服务器和远程监测终端三部分模块组成的故障诊断结构,为了对空压机的性能进行预测与优化,设计了全面的智能诊断系统。然后采用皮尔森相关系数对空压机内部运行参数的相关性进行分析,构建空压机的温度监测模型,并采用健康指数对空压机的健康情况进行评估,以数值形式反映出空压机的运行状态。最后建立以温度故障向量作为输入、故障类型向量作为输出的BP神经网络模型,通过不断的训练,求解出最优的目标函数。实验结果表明,采用本文方法对空压机的温度故障进行诊断,具有较好的监测效果,而且随着模拟故障程度的逐渐增加,能够更早的判断出空压机的温度阈值,对故障判断的准确率较高。

With the wide application of air compressors in industrial production,accurately monitoring the temperature during the operation of air compressors has become a key research topic in the whole chemical industry.Therefore,a data-driven fault diagnosis system for excessive temperature of air compressors is proposed.Firstly,the fault diagnosis structure composed of a remote terminal unit,server and remote monitoring terminal is designed.In order to predict and optimize the performance of the air compressor,a comprehensive intelligent diagnosis system is designed.Then the Pearson correlation coefficient is used to analyze the correlation of the internal operating parameters of the air compressor,the temperature monitoring model of the air compressor is constructed,and the health index is used to evaluate the health of the air compressor,so as to reflect the operating state of the air compressor in numerical form.Finally,a BP neural network model with temperature fault vector as input and fault type vector as output is established.Through continuous training,the optimal objective function is solved.The experimental results show that using this method to diagnose the temperature fault of the air compressor has a good monitoring effect,and with the gradual increase of the degree of simulated fault,the temperature threshold of the air compressor can be judged earlier,and the accuracy of fault judgment is higher.