热泵空调换热器水侧污垢故障的诊断

Artificial Neural Network Fault Diagnosis on the Heat Exchanger Fouling of Heat Pump Unit

提出了换热器水侧污垢的理论模型;利用减小水流量的方法来模拟污垢故障,当故障 发生时,吸排气压力会上升,这相当于热泵的制热系数降低了;神经网络由于具有模拟任何连续非 线性函数的能力和利用样本学习的能力,已被用于本系统的故障诊断中;采用感知器学习算法对 热泵空调换热器水侧污垢故障进行诊断。

Fault detection and diagnosis is an important method of improving safety and reliability of the system. Modeling of the heat exchanger is put forward in the paper. There are some pressure variations of the heat pump with the fouling. The heat change efficiency of the unit becomes lower. Fault data are gotten by the experiment. Using the strong abilities of artificial neural networks (ANN) in self-learning and mode distinguishing, the fault of heat exchanger fouling is identified through the perceptron ANN model which is suitable to the simple pattern classification.