Dynamic prediction of gas emission based on wavelet neural network toolbox

This paper presents a method for dynamically predicting gas emission quantity based on the wavelet neural network （WNN） toolbox. Such a method is able to predict the gas emission quantity in adjacent subsequent time intervals through training the WNN with even time-interval samples. The method builds successive new model with the width of sliding window remaining invariable so as to obtain a dynamic prediction method for gas emission quantity. Furthermore, the method performs prediction by a self-developed WNN toolbox. Experiments indicate that such a model can overcome the deficiencies of the traditional static prediction model and can fully make use of the feature extraction capability of wavelet base function to reflect the geological feature of gas emission quantity dynamically. The method is characterized by simplicity, flexibility, small data scale, fast convergence rate and high prediction precision. In addition, the method is also characterized by certainty and repeatability of the predicted results. The effectiveness of this method is confirmed by simulation results. Therefore, this method will exert practical significance on promoting the application of WNN.

Cooperative prediction method of gas emission from mining face based on feature selection and machine learning

Collaborative prediction model of gas emission quantity was built by feature selection and supervised machine learning algorithm to improve the scientifc and accurate prediction of gas emission quantity in the mining face.The collaborative prediction model was screened by precision evaluation index.Samples were pretreated by data standardization,and 20 characteristic parameter combinations for gas emission quantity prediction were determined through 4 kinds of feature selection methods.A total of 160 collaborative prediction models of gas emission quantity were constructed by using 8 kinds of classical supervised machine learning algorithm and 20 characteristic parameter combinations.Determination coefcient,normalized mean square error,mean absolute percentage error range,Hill coefcient,mean absolute error,and the mean relative error indicators were used to verify and evaluate the performance of the collaborative forecasting model.As such,the high prediction accuracy of three kinds of machine learning algorithms and seven kinds of characteristic parameter combinations were screened out,and seven optimized collaborative forecasting models were fnally determined.Results show that the judgement coefcients,normalized mean square error,mean absolute percentage error,and Hill inequality coefcient of the 7 optimized collaborative prediction models are 0.969–0.999,0.001–0.050,0.004–0.057,and 0.002–0.037,respectively.The determination coefcient of the fnal prediction sequence,the normalized mean square error,the mean absolute percentage error,the Hill inequality coefcient,the absolute error,and the mean relative error are 0.998%,0.003%,0.022%,0.010%,0.080%,and 2.200%,respectively.The multi-parameter,multi-algorithm,multi-combination,and multijudgement index prediction model has high accuracy and certain universality that can provide a new idea for the accurate prediction of gas emission quantity.

Hybrid optimization model and its application in prediction of gas emission

According to the complex nonlinear relationship between gas emission and its effect factors, and the shortcomings that basic colony algorithm is slow, prone to early maturity and stagnation during the search, we introduced a hybrid optimization strategy into a max-rain ant colony algorithm, then use this improved ant colony algorithm to estimate the scope of RBF network parameters. According to the amount of pheromone of discrete points, the authors obtained from the interval of net- work parameters, ants optimize network parameters. Finally, local spatial expansion is introduced to get further optimization of the network. Therefore, we obtain a better time efficiency and solution efficiency optimization model called hybrid improved max-min ant system （H1-MMAS）. Simulation experiments, using these theory to predict the gas emission from the working face, show that the proposed method have high prediction feasibility and it is an effective method to predict gas emission.

Grey Smoothing Model for Predicting Mine Gas Emission

A grey smoothing model for predicting mine gas emission was presented by combining the grey system theory with the smoothing prediction technique. First of all, according to the variable sequence, GM(1,1) model was set up to predict the general development trend of variable as first fitted values, then the smoothing prediction technique was used to revise the fitted values so as to improve the accuracy of prediction. The results of application in the No.6 Coal Mine in Pingdingshan mining area show that the grey smoothing model has higher accuracy than that of GM(1,1) in predicting the variable sequence with strong fluctuation. The research provides a new scientific method for predicting mine gas emission.

Study on primal CO gas generation and emission of coal seam

The main method of casting coal spontaneous combustion is prediction of index gases, with carbon monoxide(CO) commonly used as an index gas. However, coal spontaneous combustion is not the sole source of CO evolution; primal CO is generated through coalification, which can lead to forecasting mistakes. Through theoretical analysis, primal CO generation and emission from coal seams was determined.In this study, six coal samples were analyzed under six different experimental conditions. The results demonstrated the change in coal seam primal gas and concentration as functions of time, different coal samples, occurrence, various gas types and composition concentration, which are in agreement with the previous study on primal CO generation. Air charging impacts on primal gas emission. Analysis of the experimental data with SPSS demonstrates that the relationship between primal CO concentration and time shows a power exponent distribution.

Spatial context in the calculation of gas emissions for underground coal mines

The prediction of gas emissions arising from underground coal mining has been the subject of extensive research for several decades, however calculation techniques remain empirically based and are hence limited to the origin of calculation in both application and resolution. Quantification and management of risk associated with sudden gas release during mining(outbursts) and accumulation of noxious or combustible gases within the mining environment is reliant on such predictions, and unexplained variation correctly requires conservative management practices in response to risk. Over 2500 gas core samples from two southern Sydney basin mines producing metallurgical coal from the Bulli seam have been analysed in various geospatial context including relationships to hydrological features and geological structures. The results suggest variability and limitations associated with the present traditional approaches to gas emission prediction and design of gas management practices may be addressed using predictions derived from improved spatial datasets, and analysis techniques incorporating fundamental physical and energy related principles.

Prediction and control of rock burst of coal seam contacting gas in deep mining

By analyzing the characteristics and the production mechanism of rock burstthat goes with abnormal gas emission in deep coal seams,the essential method of eliminatingabnormal gas emission by eliminating the occurrence of rock burst or depressingthe magnitude of rock burst was considered.The No.237 working face was selected asthe typical working face contacting gas in deep mining;aimed at this working face,a systemof rock burst prediction and control for coal seam contacting gas in deep mining wasestablished.This system includes three parts:① regional prediction of rock burst hazardbefore mining,② local prediction of rock burst hazard during mining,and ③ rock burstcontrol.