In view of the effects of temperature on silicon piezoresistive pressure sensor output

this paper proposes a novel method for pressure sensor temperature compensation by combining Principal Component Analysis（PCA） with improved BP neural networks.In this method

the prime information in temperature compensation is extracted by using PCA so that the multi-dimensional problem is simplified and the noise error in data eliminated and the neural network improved so as to bring its strong generalization and fault-tolerance capabilities into full play.The research results show that this method can restrain the effects of temperature on pressure sensors effectively and enhance their stability and accuracy.