Research On Intelligent Control Strategy Of Advance Support Equipment On Fully Mechanized Face

The frequent occurrence of roof collapse accidents in fully mechanized mining faces,poor safety conditions,low efficiency of supporting equipment,and high labor intensity are major problems that need to be solved in safe and efficient production of coal mines.With the development of science and technology,in the face of complex working conditions in the underground,coal mining should vigorously develop intelligent control,and realize the exploitation of few people and even no one in the underground.As an important roadway support equipment for underground mines,the advanced support equipment provides guarantee for the safety of underground workers.However,China's current comprehensive mining support equipment is developing slowly,focusing on the structural performance of the support,and intelligent control.There are few studies on strategy.Aiming at this problem,this paper will study the intelligent control of the advanced support of the fully mechanized mining face,realize its autonomous prediction of the initial support force setting and the moving step,and reduce the overall equipment in the process of self-completed roadway support.Move and lift,and solve the fluctuation of support force during the process of equipment shifting.The specific work contents are as follows:Based on the structure and working principle of the advanced support equipment,the coupling dynamics model of the advance support and the roadway roof system is established,and the contact force of the full contact state is studied.The selection of the characteristic parameters of the deep learning prediction model and the electro-hydraulic coupling are equipped.Systematic research provides the basis.It is proposed to use the deep belief network to complete the prediction of the initial support force setting and the moving step.At the same time,the original network is improved by introducing Bayesian regularization and improved particle swarm optimization algorithm,and the training error is improved under the premise of ensuring the training error is small.The generalization ability of the overall network.Through the collection of 23 coal mine information,the sample space is established and the establishment of the network model is completed.When the test samples were used for simulation verification,the experiment was compared with the BPNN network to verify the superiority of the method.In order to effectively solve the fluctuation of support force during the transition process,the novel firefly algorithm neural network is used to optimize the PID controller.On the basis of the firefly algorithm,by improving the decision domain update coefficient and introducing the step size reduction formula,the convergence speed of the algorithm is accelerated and localoptimization is avoided.At the same time,when the PID parameter tuning optimization is carried out,the energy index and the overshoot indicator are introduced to solve the problem of slow response speed and many oscillation times.Finally,the optimized control system and the electro-hydraulic coupling system of the advanced support equipment are simulated and analyzed to verify its superiority.The optimized PID controller is simulated and analyzed based on Matlab analysis software and compared with the original PID.Based on AMESim software,the simulation model of the electro-hydraulic coupling system of the pillar support cylinder and the shift jack of the advanced support equipment is established respectively.The simulation parameters and control parameters in the simulation model are set to simulate the descending and lifting process of the pillar cylinder of the advanced support equipment.,simulate the process of moving the jack.The paper has 28 pictures,30 tables,and 73 references.