研究了空调制冷系统制冷剂泄漏的故障特征 ,分析了制冷剂泄漏引发的流量和压力参数变化情况 ,提出了以制冷机正常运行时与泄漏运行时的压力参数检测值作为分类人工神经网络的输入矩阵 ,得到了能对制冷系统运行状态进行分类的神经网络模型 ,它能诊断和预报泄漏故障的发生 The characteristics of the refrigerant leakage in the air conditioning and refrigeration system are studied. The variation of the flow rate and the pressure parameter caused by the refrigerant leakage are analyzed. The measured value of the pressure parameter in the normal operation and the leakage operation of the refrigerator is proposed as the input of the classified artificial neural network Matrix, a neural network model that can classify the operating conditions of the refrigeration system is obtained, which can diagnose and predict the occurrence of leakage faults