Nonlinear Characteristic Analysis And Study Of The Theory For Prediction And Simulation About Gas Emission In The Working Faces Of Coal Mine

The source of gas emission in the working face is complex and gas distribution in operation region is uneven. Gas easily accumulates in return airway and some certain parts (especially on top corner), so the work in the heading face is restricted. Therefore, gas emission in the heading face was selected as the research object. Focused on the complex underground structure of coal mine and the characteristics and law of gas emission under dynamic changing condition, combined with measured data of gas concentrations, the influencing factors of coal mine gas emission were identified with the theories such as systematic science, safety science, ventilation and safety of mine etc. The non-linear analysis and the theory for prediction and simulation about gas emission was processed by applying the non-linear science, modern applied mathematics, computer science etc. Then, the chaotic time series prediction model of gas emission which is based on the SVM and the CMLs model which is based on the spatiotemporal chaos behavior of gas emission were set up. The main research content and results are as follows:(1) Gas concentration series of the No.Ⅱ1024 working face is analyzed on its time structure variation coefficient characteristic. The complex change characteristic of 8 indicates that gas emission is constantly aberrated in time and space and has complicated nonlinear characteristic. It is also explained the conventional techniques is difficult to research gas emission and the characteristic and law of gas emission must be researched by the nonlinear theory;(2) Combining with on-site gas measured data, relative gas emission and absolute gas emission in the No.7126 working face were analyzed comprehensively. The results show that gas emission in this working face changes greatly and presents an obvious complexity, the relative gas emission rate and absolute gas emission rate shows more pronounced instability. It is verified the existence of complex non-linear relationship among the elements of gas emission and their interaction with external systems in coal mine working face. Systemic analysis was processed to gas emission of the working face. It was proved that chaos theory research about gas emission is feasible according to the chaos coupling characteristic of gas emission system;(3) Phase-space reconstruction is the premise of most chaotic research and calculation, and the key point of phase-space reconstruction technology is to select the embedding dimension m and delay timeτ. For the time-series data of absolute gas emission in the No.Ⅱ1024 working face, adopting the GP algorithm and mutual information function, the embedding dimension m and delay timeτcan be specified respectively, which are 10 and 15. Under this premise, non-linear characteristics of the No.Ⅱ1024 working face is identified through several chaotic characteristic analytical methods, such as power spectrum, auto-correlation function, Poincarre cross-section and return map etc. According to characteristics of the analysis graphics, it can be confirmed that gas emission in coal mine working face has chaotic characteristics. It is further proved that it is reasonable and feasible to research gas emission of the working face by chaos theory;(4) The feasibility of the prediction of absolute gas emission quantity was proved by the R/S analysis about it in the No.Ⅱ1024 working face. Through the calculation and analysis of the biggest Lyapunov index, it indicates that with the increase of step-size, the Lyapunov index value tends to be a constant which is about 0.041-0.042, and is a limited positive value. It illustrates that the attractor of gas emission system in the working face is strange attractor, the system is in chaotic sport status and the system movement presents chaos state. The determination of potential predictable scale avoids the situation of overlarge time scale when carry on the prediction research of gas emission in the working face. The contrast and analysis were processed about the prediction models of the ANN and the SVM based on the phase reconstruction method in the chaos theory, The result shows that the SVM prediction model possesses a better prediction precision, easier to build and faster training speed, and due to the consideration of time-space chaos of gas emission, it is more reliable. Also, it provides a new guideline on the predictable and the prediction model of gas emission of the working faces.(5) Gas emission system is spatio-temporal chaos system. It's not only chaotic in the time dimensionality but chaotic in the space dimensionality. Simulation research about gas emission system was processed by the CMLs model. The CMLs model about gas emission was built and the spatio-temporal development behavior of actual system was qualitatively simulated. The evolution characteristic of the CMLs model about gas emission shows that it can restrain the chaos of gas emission via appropriate technical measures and management measures to enhance the safety of the working face.(6) The nonlinear prediction model and the CMLs model about gas emission were applied to gas treatment in the No.Ⅱ1023 working face. Pertinently measures were adopted according to two models. The result shows gas pressure and gas emission all take on downtrend and gas density is not beyond the limit value. It offers biggish safe space for the production of the No.Ⅱ1023 working face. The model's validation is also verified.Nonlinear characteristic analysis and the theory of prediction and simulation about gas emission of the working face will create a new field of study to gas emission in the working face. The research result and its application have practical significance and broader application prospect for effective prevention and control of gas accidents and to improve coal mine safety, and also have greater guiding role to other types of accidents control.