车载空调制冷系统故障诊断研究

为实现车载空调制冷系统故障诊断功能,快速判断空调制冷系统可能出现的故障类型,文章建立车载空调制冷系统一维仿真模型,并以压缩机进出口温度、压力等参数为特征参数,冷凝器风量降低、制冷剂泄漏等故障为输出目标结果,构建车载空调制冷系统的反向传播(back-propagation,BP)神经网络故障诊断模型和决策树故障诊断模型。研究结果表明:当冷凝器风量降低时,压缩机排气温度与排气压力上升,空调系统的制冷量和性能系数(coefficient of performance,COP)下降。通过对比2种不同诊断策略的仿真测试结果发现,采用BP神经网络进行车载空调制冷系统故障诊断的准确率可以达到92.5%。

In order to realize the fault diagnosis function of on-board air conditioning refrigeration system and judge the possible fault types quickly, a one-dimensional simulation model of on-board air conditioning refrigeration system was established. The back-propagation (BP) neural network fault diagnosis model and decision tree fault diagnosis model of on-board air conditioning refrigeration system were constructed by taking the temperature and pressure of compressor inlet and outlet as the characteristic parameters, the condenser air volume reduction and refrigerant leakage as the output target results. The results show that when the condenser air volume decreases, the exhaust temperature and pressure of the compressor increase, and the cooling capacity and coefficient of performance (COP) of the air conditioning system decrease. By comparing the simulation test results of two different diagnosis strategies, it is found that the accuracy of fault diagnosis of on-board air conditioning refrigeration system using BP neural network can reach 92.5%.