The Research Of Coal-rock Recognition Method Based On Current Signal

The automatic and unmanned of fully mechanized coal face is one of the developmentdirection of mining automation.Its development bottleneck of technology is the issue ofcoal-rock interface recognition(CIR).Aiming at the shortcoming that existing coal-rock recognition is susceptible to geologicalconditions,roof condition and so on,the new method of coal-rock recognition based upon thecutting current of shears was proposed.Around the problem,this paper mainly do the work asfollows:Aiming at the method of cutting current,the essence of coal-rock recognition is extractingprinciple that is load characteristic of cutting coal-rock.Characteristics of the drum cuttingprocess was researched,and according to the fractal characteristics of coal crushingprocess,the drum cutting resistance torque model based on fractal theory and EMDdecomposition was structured to overcome shortcomings in the tradition of stationary randommodel,which provides fundamental basis for research of coal-rock recognition characteristicsbased on the current method.Aiming at the situation that current signal of cutting motor containing feature informationof transmission dynamics in cutting department,the model named‘cutting resistancetorque-cutting motor current’ was established for removing the influence of cutting departmenton component of cutting coal-rock characteristics and providing the model support for thesimulation of cutting motor current under different working conditions.Aiming at the current signal of cutting motor having the characteristics of nonstationaryand nonlinear,this paper used the technology of wavelet packet to do the feature extraction ofcurrent signal.Extraction method based on the energy distribution of wavelet packetdecomposition is able to determine the sensitive frequency bands of current signal.By thisway,we can proving the effectiveness of coal-rock recognition based on current signal in theory.