Aimed at the low efficiency of fully mechanized coal mining of shearer,a centralized control method of fully mecha-nized coal mining face based on deep learning is proposed.By long short-term memory(LSTM)neural network and residual network(ResNet),this paper extracts the time-space information characteristics of the shearer,a prediction model for shearer cutting drum trajectory based on LSTM-ResNet is constructed to predict the future trajectory of the cutting drum.The motion control model based on linear quadratic regulator is used to solve the swing angle of the rocker arm of the shearer cutting drum and the traveling speed of the shearer,which can track the motion trajectory of the cutting drum,and realize the adaptive oper-ation control of the shearer cutting drum.The simulation results show that the proposed LSTM-ResNet prediction model can more accurately predict the cutting trajectories of the supper and lower drums,and the root mean square errors are 0.022 m and 0.011 m,respectively,which provides a reference for intelligent control of the fully mechanized mining face.
维普
万方数据