制冷剂泄漏的诊断与预报

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。

Diagnosis and forecast of refrigerant leakage in the refrigeration system

Analyses the fault characteristics of refrigerant leakage in the refrigeration system for air conditioning and the pressure and flow rate of refrigerant. Describes a series of characteristic indexes of pressure signals to indicate the leakage fault. Proposes a back propagation neural network model with the input matrix composed of the normal and abnormal pressure indexes to classify the running state of the refrigeration system to diagnose and forecast the leakage fault.