Abstract:According to the nonlinear property of the output of temperature sensor,a nonlinear correction model for data acquisition of intelligent sensor is established.By using the nonlinear mapping property of the function chain neural network,the parameters of the anti-nonlinear function are quickly obtained,realizing the accurate fitting of the temperature calibration model.The methods and steps of parameter calculation using functional chain neural network are given out.The experiments on platinum resistance temperature elements are widely used in meteorological science demonstrate the effectiveness of the algorithm,and the correction accuracy reaches 0.1%.
Dempsey G L,Alig J S,Alt N L,et al.Control sensor linearization using artificial neural networks[J].Analog Integrated Circuits and Signal Processing,1997,13(3):321-332.