【摘要】 NTC热敏电阻具有高非线性特性,BP神经网络具有较强的非线性映射的能力,试验应用BP神经网络对NTC热敏电阻传感器进行非线性校正,将电阻值在输入训练网络前进行归一化处理,并利用网络对实际的温度采集系统进行校正,得到了理想的结果。此方法实现简单,精度高,在实际应用中有较高的应用价值。更多还原

【Abstract】 The characteristic of negative temperature coefficient(NTC) thermistor is high non-linear;BP neural network has good ability of nonlinear mapping.The test used BP neural network to correct the non-linear characteristic of NTC thermistor,the input coefficient were normalized before the network was trained,and used this neural network to calibrate a real temperature system,obtained the ideal result.The result proves that the method can be realized very easily and with high precision,and it has high application value.更多还原