When silicon pressure sensor is used in industrial environments

particularly for measuring downhole pressure in deep oil-well

the ambient temperature range is always large.Because of its unique structure

the value of pressure output appears nonlinear changes

greatly reducing the measurement accuracy of pressure sensor.This article is based on PSO-BP neural network method used in pressure sensor compensating and correcting the error when temperature changes to reach the system accuracy requirements.The intention of PSO-BP algorithm is to improve the initial weights and screen the thresholds of BP neural network

then train the samples by using BP neural network in order to improve the generalization ability and stability of system.