The Study Of Forecasting On Coal And Gas Outburst Base On Fractal And Chaos

Coal and gas outburst is a very complicated dynamic phenomenon of coal rock which contains gas, it can be regarded as a catastrophe behavior of a complicated non-linear dynamic system during its evolving process in space and time, and also be one of the major disasters influencing safe production of coal mine. It has very theoretical significance and engineering applying value for effective and forecasting on coal seam fire to explore the intrinsic rule in the coal and gas outburst. Therefore, in this paper, from the view of non-linear science, some researches on dynamic gas emission has been done on chaos and fractal theory, and a new approach is proposed to predict coal and gas outburst. By reconstructing phase spaces to the representative time-series date of dynamic gas emission which through data pretreatment, and discussed its chaos parameters, consequently it is proved that time-series possess chaos and fractal characteristics, and the maximal forecasting time can be got by the method based on the maximal Lyapunov exponent. Researching on the time series of coal autoxidation indicates that time series can be qualitication described with fractal dimension under no-scaled zone. And it obeys the fractal distributions, although it is disorderly and unsystematic on the surface. Finally, the time series have been forecasted by the method based on the maximal Lyapunov exponent and an Artificial Neural Network model is set up for outburst prediction of coal seam tunneling face based on the chaos and fractal characteristics of time-series.