Research On Identification And Statistical Analysis Of Coal Macerals Based On The Idea Of Peak Splitting

As a comprehensive subject with strong cross application,Statistics covers almost all fields of real life with the integration of mathematical knowledge and professional knowledge of other disciplines.In the current situation of global energy scarcity,coal is an indispensable non-renewable energy for human survival and social development.How to use it rationally and take into account green development is a worldwide concern.As one of the important indexes to evaluate coal properties,coal macerals are often used to guide the coal blending and coking work in actual production practice.In addition,vitrinite reflectance is the main basis to characterize the degree of coalification and identify mixed coal.Therefore,the determination of vitrinite reflectance and the quantitative statistics of the content of coal macerals have important research value and application significance.Based on the statistical characteristics of coal and rock and the peak splitting rules of distribution function,this paper studies the identification and statistics of coal macerals.The main work is as follows:Design a preprocessing scheme for coal and rock images.First,experts in the field complete the acquisition of coal and rock images,and formulate a four-step preprocessing scheme according to expert experience and image characteristics.Secondly,the standard BBPSO is improved by adding the particle adaptive perturbation term,and the convergence of the ABPSO algorithm is proved theoretically by the stochastic analysis method.The BP neural network model is trained by the ABPSO optimization algorithm to complete the denoising process of coal and rock images.Then,the grayscale is converted into reflectance by Least square method,the reflectance distribution map is drawn,and the texture statistical characteristics of coal and rock particles are extracted.Finally,based on the prior knowledge that vitrinite reflectance of single coal is gaussian distribution,the peak offset range of vitrinite of each single coal is determined according to the empirical rule of Gaussian distribution.Constructing an adaptive peak-finding method for reflectivity distribution.According to the mathematical characteristics of the peak points,an adaptive peakfinding algorithm based on fuzzy entropy weight iteration improved S-G method is developed.First,under the premise of maximum information entropy,the optimization method is used to prove that the entropy weighted fuzzy iterative method satisfies local convergence.Secondly,the optimal parameters of S-G filter are found by multi-index comprehensive evaluation method of entropy weight fuzzy iterative method,and the smooth processing of reflectivity distribution function is completed to effectively remove burrs and ensure data integrity.Then,an adaptive effective maximum peak point and effective inflection point peak point screening algorithm is designed to filter out false peak points and redundant peak points.Finally,the designed adaptive peakfinding method of reflectivity distribution is used for experimental verification,and the results show that the peak search result of the proposed method is more effective than the traditional peak-finding algorithm.Formulate the identification strategy of macerals of single coal.In view of the algorithm of coal macerals identification based on the idea of peak separation for single coal,firstly,based on the analysis of the results of adaptive peak separation,a multistrategy peak position identification algorithm is designed to classify the coal and rock particles into active-inert particles requiring peak clustering and pure vitrinite particles,inertinite particles and exinite particles without peak clustering,and the peak positions of coal and rock particles requiring peak clustering are selected.Secondly,combined with the peak separation rule,the statistical characteristics of coal and rock particles and the Gaussian fitting method,the thresholds of exinite,vitrinite and inertite were determined respectively,and the clustering analysis of coal and rock particles was completed.Finally,using the designed single coal maceral identification method for experimental verification,the experimental results show that the method designed in this paper can effectively identify single coal particles and realize the quantitative statistics of maceral content.The accuracy is 96.85% and the minimum entropy is as low as 0.6153.The effect is obviously better than the traditional method and has good practical significance.Formulate the identification strategy of maceral of mixed coal.First of all,referring to the preprocessing scheme of single coal and rock images,and based on the characteristics of the mixed coal and rock images,the preprocessing scheme of mixed coal and rock images are designed.Secondly,the standard deviation screening threshold0.625 and the bimodal discriminant model are added on the basis of the multi-strategy peak position identification algorithm of single coal,and the multi-peak discrimination model is optimized based on the logic rule of reference peak position,a coal maceral identification method suitable for mixed coal is designed.Finally,using the designed mixed coal macerals identifying method for experiment,the experimental results show that the method designed in this paper is highly consistent with the automatic identification and determination of macerals of mixed coal and the manual marking by experts.The method designed in this paper has a certain guiding significance for the ratio of active-inert and the analysis of coal and rock particles in coal refining operation.