Impact-slip experiments and systematic study of coal gangue “category” recognition technology Part I: Impact-slip experiments between coal gangue mixture and top coal caving hydraulic support and the study of coal gangue “category” recognition technology

Abstract

Advancing of the intelligent fully mechanized caving mining technology (IFMCMT) is the fundamental method to improve the output of coal resources. Accurate identification of coal gangue is the necessary condition for the realization of IFMCMT. To achieve coal gangue recognition, Internet of Things (IoT) technology and impact-slip contact characteristics based method for coal gangue “category” recognition is proposed in this paper. An IoT recognition system is constructed, then impact-slip test bed between coal gangue mixture and the hydraulic support is designed. Secondly, coal gangue “category” recognition parameters are discussed, hereby multi-dimensional contact characteristic information acquisition system is constructed. After that, impact-slip tests were carried out to extract and process the multi-source and multi-azimuth signals, and three kinds of data samples were obtained. Finally, 10 kinds of classification algorithms were selected for random classification recognition of coal gangue “caving category” and “shut down category”. Coal gangue “category” recognition ability by algorithms, signal types and data types is compared, which provides the theoretical basis for further classification recognition effect improving research. Research show that coal gangue “category” random recognition accuracy can reach to 0.935 (Define the range of recognition accuracy as 1), which proves the effectiveness of “category” by IoT system and impact-slip contact characteristics.