Fault Detection And Diagnosis Of Circulating Pump In Heat Exchange Unit

Centralized heating has become the main heating method in winter in northern China,and the problem of heating system failure has been a key issue affecting the quality of centralized heating.The failure of three-phase asynchronous motor is the main factor of heat exchanger unit failure which may cause the heat exchanger station to stop operation.That may cause uneven heating and even affect the subsequent heating operation in severe cases,bringing great inconvenience to production and life.Therefore,it is very important to study the fault prediction and diagnosis of circulating pump motor of heat exchanger unit.In this paper,we focus on the mechanism analysis,mathematical modeling and fault diagnosis of stator turn-to-turn short circuit and rotor broken strip fault of circulating pump motor of heat exchanger unit of central heating system.The following main masks are accomplished in this paper:First of all,the current status of research of the heat exchange unit at home and abroad is described in depth.The basic working principle and common faults of the heat exchange unit are introduced,and the working characteristics and mechanism of the three-phase asynchronous motor under the condition of driving the circulating pump load frequency conversion and speed regulation are analyzed in detail.The difficulty of fault diagnosis of pump motor in heat exchanger unit is analyzed and the mathematical model of fault of three-phase induction motor is established.Secondly,this dissertation has studied the coordinate transformation model in depth and combines it with the multi-loop analysis method to establish mathematical models of induction motor stator turn-to-turn short circuit and rotor broken bar fault based on phase coordinate transformation and multi-loop analysis method for two types of faults of motor turn-to-turn short circuit and rotor broken bar respectively.Simulations are carried out by MATLAB/Simulink to validate the stator current waveforms with different types and degrees of faults by which the fault characteristic signals are analyzed.The paper also use Maxwell finite element simulation software to establish a physical model of the motor fault,and derive the stator current waveforms at the time of the fault through simulation tests and analyze the current waveform characteristics,which verifies the rationality and accuracy of the mathematical model of the motor fault again.Finally,this dissertation uses the EMD to decompose the motor fault current signal and derive the IMF components of different types and degrees of faults.The EMD program is written in MATLAB to extract the energy entropy values for each fault’s IMF component.The paper adds the energy entropy data into SVM to extract the feature signals of energy entropy and finds that the optimized SVM penalty factor and kernel function parameters by combining the whale optimization algorithm and the SVM conduce to achieving accuracy of up to 98% in the process of fault diagnosis.The accuracy of the support vector machine based on whale optimization algorithm for motor fault diagnosis method is verified.