Research On Sorting Technology Of Coal And Gangue In Mine Based On Machine Vision

Coal is an important basic energy and raw material in China,which has an important strategic position in the national economy.As one of the important links of coal production,the separation of coal and gangue is of great significance to improve the efficiency of coal use and the protection of ecological environment in mining area,which promoting the implementation of green energy development in China.Coalgangue recognition is the precondition of coal-gangue separation.In recent years,with the rapid development of machine vision technology,coal-mine equipment based on machine vision technology has been widely used in all walks of life,including coal industry.This thesis takes the sorting technology of coal and fangue in mine based on machine vision as the research objective,builds a complete system framework of recognition and location of coal and gangue based on machine vision,and finally realizes the recognition and accurate location of coal and gangue based on machine vision.The main work and research results of this thesis are as follows:(1)Combined with the requirements of coal-gangue recognition and positioning system based on machine vision,a coal gangue recognition and positioning technology based on visible-light machine vision is proposed,which balances the contradiction between the demand of coal-gangue separation equipment and high layout cost in mine,and puts forward a complete system architecture and a set of coal-gangue recognition and positioning algorithm including image denoising,enhancement,segmentation and positioning.(2)The noise characteristics of coal-gangue image,which collected by coalgangue recognition and positioning system based on machine vision,are analyzed.Aiming at the complex noise contained in coal-gangue image,a coal-gangue image denoising algorithm based on wavelet threshold theory and chaotic firefly algorithm is proposed.By introducing chaotic random field theory,the accuracy and robustness of firefly algorithm to realize iterative optimization are improved.The optimized chaotic firefly algorithm is used to iteratively optimize the threshold in the wavelet threshold image denoising algorithm,whiche can improve the quality of coal-gangue image effectively.(3)A coal-gangue image segmentation algorithm based on intuitionistic fuzzy theory is proposed.Combined with the chaotic firefly algorithm proposed in this thesis,two key parameters of intuition-fuzzy C-means clustering algorithm are optimized.The coal-angue image segmentation and positioning model based on intuitionistic fuzzy theory and chaotic firefly algorithm improved the applicability and robustness of the image segmentation algorithm.On this basis,the unconnected sub-region in the coalgangue image is divided again,and then the accurate positioning of the coal gangue sub-region is realized.(4)A coal-gangue recognition algorithm based on SVM algorithm is proposed,and a coal-gangue image feature model based on directional gradient histogram and gray-level co-occurrence matrix is established.Based on the proposed chaotic firefly algorithm,the penalty factor and the parameter of adial basis kernel functionradial basis function in the support vector machine are further optimized.By constructing the improved support vector machine model with higher classification accuracy,the accuracy and robustness of the proposed coal-gangue identification algorithm are improved.Based on the research of the sorting technology of coal and fangue in mine based on machine vision,an on-ground simulation experiment platform and an underground industrial experiment platform were built in the collaborative innovation center of the province and the ministry of mining intelligent mining equipment and Xinzhuang coal industry of Henan Shenhuo Group Co.,Ltd.The experimental results show that the recognition accuracy of the coal-gangue recognition and positioning technology based on machine vision proposed in this thesis is more than 90%,and the statistical error of coal-gangue positioning is about 2.51%,which meets the accuracy requirements of coal-gangue recognition and positioning in practical engineering.The research results is of great significance for further promoting the practical application of coal gangue separation equipment.In this dissertation,there are 26 figures,16 tables and 103 references.