Design Of Coal And Gas Outburst Acoustic Emission Monitoring Based On BP Neural Network

As the first energy source in China, the demand of coal growing in every day with the developing of economy, accompanied by the coal mine safety should not be overlooked. Coal and gas outburst is a serious natural disaster in coal mine, it is unexpected, but were some precursors before an outburst, one of them is the acoustic emission. To the current situation of our country’s coal mining, the thesis designed a coal and gas outburst of the acoustic emission monitoring based on the theory of acoustic emission, and simulating, debugging, finally achieve the expected effect. The thesis designed the monitor from the following three aspects.In the coal mine, the acoustic emission signal is much many but no obvious pattern, and artificial neural network has a strong non-linear processing, which can not study the complex relationship between a large number of monitoring data, only the memory of learning through the input and output can find the corresponding nonlinear relationship, which is suitable for the dynamic prediction of the acoustic emission data. This paper systematically studied the principle of BP neural network prediction of coal and gas outburst acoustic emission, and demonstrates the feasibility of the design.Based on the theory of the acoustic emission and the BP neural network, the thesis designed an overall framework of the acoustic emission detector. According to the demand of rapid, timely, accurate prediction of coal and gas outburst, designed a DSP+MCU monitoring system of dual CPU architecture, clever use the DSP data processing capabilities of the acoustic emission signals and the microcontroller interactive features.Departure from the hardware architecture, the thesis designed of the monitoring requirements based on the prominent acoustic emission data in various functional modules in the system. Subsequently, the software programming and debugging based on the hardware structure of each module according to their function.Finally, it is that build a model of the acoustic emission signals BP neural network processing by the neural network toolbox in MATLAB soft ware, which corresponding the data by learning and training, and finished the simulation, and improved the BP neural network by using of genetic algorithm, too. The results show that the method can be better prediction of coal and gas outburst.