Experimental Study On Thermal Fault Of Inverter Air Conditioner System

With the rapid development of society and people’s growing need for a better living environment,inverter air conditioner system occupies a very high proportion in the current air conditioner market because of its good comfort and control flexibility.However,inverter air-conditioning system may have various faults,increasing system’s energy consumption.Therefore,it is necessary to research on the faults of inverter air conditioner system to explore the characteristics of different types of faults,and then develop the detection and diagnosis strategies of the common faults so that they can accurately detect the occurrence of faults and identify the types of faults and take corresponding measures,improving the system’s operating efficiency.In this paper,a simulation experiment platform for thermal faults of inverter air conditioner system was built,and four common thermal faults such as compressor valve leakage,condenser fouling,evaporator air volume insufficiency and liquid line resistance increasing were simulated.And under different operating conditions,the characteristics of the 4 types of thermal faults in inverter air conditioner system were analyzed.Experimental results show that all the four kinds of thermal faults will reduce the performance of inverter air conditioner system.And the characteristics of the faults are influenced by not only indoor and outdoor temperature but also the compressor operating frequency.By analyzing the experimental results under different working conditions and different frequencies,the characteristics of four types of thermal faults in inverter air conditioner system are summarized.In this paper,Principal Component Analysis(PCA)was used to test the four kinds of thermal faults under different operating conditions The influence of statistic types on the test results was also discussed.An adaptive PCA fault detection model based on the decision tree was proposed and its validity was verified.The results show that the PCA model can detect the occurrence of the four types of thermal faults in the air conditioning system under the fixed frequency operation of the system.And its detection effect is affected by the types of model statistics.The PCA model can not detect the faults under the converting frequency operation of system.However,the PCA based on the decision tree can effectively detect the four types of thermal faults with the detection rate above 97%.Principal component analysis and its improved method can effectively detect the fault occurred,but do not recognize the occurrence of fault location and can’t provide effective guidance for the follow-up maintenance.Therefore,this paper uses the method of Mind Evolutionary Algorithm(MEA)to optimize the neural network to diagnose the four kinds of thermal faults.The results show that the optimization model can effectively separate four kinds of thermal faults,which can achieve 90.5% diagnostic accuracy,higher than the neural network model.