基于神经网络的供暖散热器热红外模型

The Infrared Mode of Heating-Radiator Based on BP Neural Network

红外测取供暖散热器故障表面温度往往会产生误差,影响红外故障检测和诊断的准确性。应用BP人工神经网络,建立以供暖散热器的进、出水口温度等6个量作为输入值和热电偶测定的散热器表面温度作为单个输出值的热红外模型,利用280组数据进行训练和验证,对红外测取的故障散热器表面温度进行了修正并取得了较好的效果,可以改善散热器故障红外检测的准确性。

There have some error in detecting the surface temperature of heating-radiator leading to the inaccuracy of failure detection and diagnosis. With the BP neural network, an infrared mode of heating-radiator is building based on six input values and one output value. After 280 groups of data training and verification, the surface temperature of heating-radiator from the infrared thermal imager is modified and a good result is obtained. It lays a foundation for the quantitative analysis of the infrared imaging technology detecting heating-radiator fault.