Research on compensation for pressure sensor thermal zero drift based on back propagation neural network implemented by hardware circuits

This paper presents a mixed simulation of a back propagation (BP) neural network algorithm with a digital and analog circuit. The analog circuit is designed to realize the main part of the neural network to speed the operation rate, which may solve one difficulty in neural network applications. Meanwhile, the digital circuit is employed for its advantages on storage and control ability to constitute the control circuit. This paper applies the research result to the compensation for thermal zero drift of pressure sensors. The experimental result proves validity of the design scheme, which provides a feasible solution for the designing of intelligent pressure sensors.