Filter Out The Large Numbers Of Gas Interference In Gas Monitoring And Optimization Of Adaptive Filters

Gas disaster has been a major hidden danger of coal mine safety in our country coal mine accidents frequently happen caused great loss and damage to the enterprise production and miner life safety to ensure the safety of mine production to miner life safety we must be carry out on the gas monitoring and alarm gas pulse interference in the signal transmission signal often cause false alarms. False alarm cut off the downhole power and forced to stop production affect the normal production and paralysis of the safety awareness of workers problem of large number interference has been plagued by coal mine production this article relying on jiangxi province graduate student innovation fund the problem of large number interference of gas to start a research project. This paper do the following research:(1)Due to the large number of pulse interference is a kind of nonlinear data so we selection of BP neural network for monitoring data of wave filter choice a lot of mine monitoring data as sample the BP model is established and according to the error continuously select right value and function of each layer select the optimal parameters finally has carried on the further improvement of the model and the model was tested and the training in order to highlight the filtering effect this article also take CME algorithm on the sample filter and compared two kinds of intelligent algorithm of filtering effect.The result of simulation shows that: in this paper artificial neural model is efficient to filter the gas monitoring of large number of interference while retaining the real interference pulse model is reasonable and efficient.(2)This paper attempts to use the immune theory in the filter, make improvements for the shortcomings of traditional LMS algorithm, analyzes the basic principle of linear adaptive filter and the influence of various parameters on filter performance. Immune algorithm to optimize the design parameters, and design an adaptive notch filter and an adaptive line enhancer, given the results of the algorithm optimization of key parameters ?, simulation of the filtering effect, the results show the success to filter out the noise.