Design of downhole CO gas monitoring system based on GA-PSO-BP hybrid optimization algorithm

Based on the complex and changeable temperature and humidity environment of underground working face and the negative influence of suspended coal dust on the monitoring accuracy of CO gas in mine, a scheme of underground CO gas monitoring system based on GA-PSO-BP hybrid optimization algorithm was proposed, this system was designed with 9 sample sampling nodes, the sensing end of each node used infrared gas sensors. The BP Neural Network based on Genetic Algorithms and Particle Swarm Optimization was used in the decision-making end to compensate the temperature and humidity of the sensing segment under complex and variable conditions. Multiple experimental results show that, compared with the widely used BP neural network algorithm and the particle swarm optimization BP neural network algorithm, the maximum error of CO gas concentration returned to the monitoring center by 9 test sample nodes after using the hybrid optimized BP neural network (GA-PSO-BP) algorithm is no more than 1.30%, which meets the requirements of downhole gas monitoring accuracy.