Approaches to realize temperature compensation of pressure sensor based on genetic wavelet neural network

The characteristics of temperature error and nonlinearity of silicon piezoresistive pressure sensor are introduced. After comparing characteristics of several neural networks, a method for compensating temperature error and non-linearity of silicon piezoresistive pressure sensor is designed using genetic wavelet neural net work which has faster speed quality convergence and higher precision than BP neural network. The experimental results show that temperature error and nonlinearity of silicon piezoresistive pressure sensor can be reduced markedly. In the range of -40~60□, temperature error can be reduced from 5.4% t o 0.2 %.