| Fully mechanized caving mining is one of the main technologies for thick coal seam mining in our country.However,because of its complex working conditions and the harsh environment,at present,top coal caving is still in the backward state of artificial coal releasing,which are prone to produce problems of "overrelease" and "underrelease".Accurately identify coal gangue and then adopt strategies to automatically control "caving"or "shutting down",is the effective way to carry out the intelligent fully mechanized caving mining safely and efficiently,and it is also a hot issue in the research of intelligent fully mechanized caving mining.On the basis of characteristics analyzing for top coal caving mining technology,this paper adopts the combination means of theoretical analysis,numerical simulation and experimental research,the impact and slip contact characteristics between coal gangue and top coal caving hydraulic support is studied,and the experimental technology research based on multi-category experimental signals and multi-information fusion is conducted.Coal gangue classification and recognition based on dynamic impact-slip contact characteristics is realized.This research is supported by the National natural science foundation of China of "Research on coal gangue recognition technology based on vibration signal of the tail beam".In this paper,research technique for coal gangue identifiability based on contact response differences is presented.Impact and contact dynamic model between rock and the metal plate at the critical compression state and the yield impact velocity model of the rock is built.Recursive solution method is proposed for the higher order nonlinear contact state equation,and the elastic impact contact dynamic mathematical model when the impact position is changed between coal gangue and the metal plate with different supporting states is obtained.Impact damage dynamic model between the quasi brittle coal gangue and the metal plate is established and the contact finite element simulation is carried out.Through the analysis of the elastic and quasi-brittle contact response,contact difference characteristics between coal gangue and the metal plate are obtained.Impact test between coal gangue and the metal plate is then carried out,and the contact vibration signal of the metal plate is extracted to identify coal gangue particles.Recognition accuracy of the test reached to 91.25%,which verified the effectiveness and feasibility of the proposed coal gangue recognition method.For the impact and slip contact behaviors between coal gangue and the hydraulic support existing in the top coal caving,the rigid-flexible coupling impact contact analysis between coal gangue and the top beam,the shield beam and the tail beam of the hydraulic support is carried out by the combination of Hypermesh and Adams.Subsequently,slipping contact simulation between different proportion coal gangue mixture and different coal drawing attitude hydraulic supports are conducted in discrete element software,and the influence of coal gangue mixed ratio and the posture of the hydraulic support on the slipping contact characteristics between coal gangue and the hydraulic support is studied.By analyzing the impact and slip contact characteristics,system contact response differences caused by the difference of coal gangue properties and mixing ratio are obtained.The structure-contact theoretical model between coal gangue particle and the tail beam structure is established,and the influence factors of system contact response are obtained through the analysis.The influence law of each influence factor on the system dynamic response is studied.Nonlinear contact behavior of the pin shaft and the system contact response characteristics induced by the coal gangue impact are studied under the multiple clearances and variable clearance conditions by introducing the nonlinear factors of clearance.By analyzing the contact response differences under multiple working conditions,it is determined that the pressure of the tail beam jack,the vibration signals of the tail beam and the pin shaft are taken as coal gangue classification and recognition parameters.Impact-slip test bed between coal gangue and the hydraulic support is built and the two dimensional multi-information acquisition system is constructed.2270 group variable coal gangue mixing ratio impact-slip tests are carried out.The orderly random classification and recognition for coal gangue "caving category" and "shut down category",the recognition reliability and "coal gangue mixture proportion range" adaptability are studied based on the single channel signal and 10 kinds of classification algorithms such as GBDT.Results show that the time domain characteristic data of the signal has stronger classification and recognition ability,the validity of the test method for coal gangue "category" recognition based on the impact-slip contact characteristics is verified.In order to further improve the classification and recognition accuracy,the time-domain characteristics of the test signals are taken as samples,coal gangue multi-information fusion classification and recognition methods based on signal serial characteristics and based on the parallel voting system are proposed.Meanwhile,coal gangue classification and recognition is studied by Kalman filtering data fusion.Recognition accuracy of each method is obtained.The study shows that coal gangue classification and recognition method based on signal serial characteristics and based on the parallel voting system can effectively improve the orderly random classification and recognition accuracy.On this basis,coal gangue completely random classification and recognition is continually conducted,the classification and recognition method of "GBDT to all channel signal serial characteristics" is determined as the best classification and recognition mode.The orderly and completely average classification and recognition accuracy all up to 99%,and this classification and recognition mode is stable and reliable. |
PDF Full Text Request