The Design Of Intelligent Gas Sensor

This paper firstly introduces the situation of coal mine safety, the importance of mining, and the problems exsisting in gas monitoring system of coal mine in china, pointing out the real-time gas concentration detection does have important prerequisites in guaranteeing coal mine produces safely. And then introduces the theory and method of gas concentration detection.Thermo catalysis sensors take on strong property of nonlinearity when gas concentration is high. The neural network has good performance of nonlinear approximation and self-learning, adaptive; In order to enhance the stability, reliability, and usability of Methane sensor, this paper studies the BP Neural Network application in Intelligent Sensing Instrumentation System. The research indicates that using BP neural network algorithm can effectively improve the nonlinearity performance of sensor.This paper introduces a new kind of design scheme of sensing system for thermo catalysis sensors. The 16-bit dsPIC33FJ64GP206 digital signal controllers (DSC) is chosen as the microprocessor. The hardware design of this control system contains circuit of power supply, main chip, LED display module design, the storage module design, alarm circuit and communication circuit, and so on. Gas safety concentration, nonlinear coefficient and zero drift coefficient are stored in an external memory EEPROM, achieved power-down protection. In communications, the data transmits between the sensors and the master station through RS-485 bus.In order to improve the sensor stability, accuracy and real-time, the modular programming approach is adopted on the software design. Code is wrote in C language and assembly language. Using DSC s DSP engine and DMA function, through the BP neural network nonlinear correction to process the collected data. Software programming has been done to realize these functions of program in detail. For more visually the technological flowsheets are drawn. The simulated test is held.