A temperature compensation system for silicon pressure sensor based on neural networks

Abstract

We present a temperature compensation system for silicon piezoresistive pressure sensor based on neural network. This system can be used for measuring the pressure of various media. And the design can simplify the implementing hardware of pressure measurement system. Compared with traditional design, it can output three signals: current signal, digital signal based on RS485 and Zigbee wireless signal, which make the system more practical to use. Due to temperature drift of silicon piezoresistive sensors, a program developed on LabVIEW in PC is used for temperature compensation using neural networks. The efficacy of neural networks has been verified by linearity, zero temperature drift and sensitivity temperature drift of pressure sensor after temperature compensation. After being tested over temperature range and pressure range, the accuracy of pressure measurement system from 0.7%FS (full scale) has been promoted up to 0.2%FS.