| With the gradual development and maturity of automation,intelligent technology and equipment,more and more fully mechanized coal mining faces in China have promoted the application of automation and intelligent technology,which has brought broad prospects for the safe production of coal mines.In the production process of automatic and intelligent working face,automatic follow-up control of hydraulic support is one of the key technologies.At present,the technology has made some progress in theoretical research and engineering application,but there are still many problems.Firstly,it is not suitable for the complex and changeable mining geological conditions of the working face,especially for the working face with large dip angle,unstable coal seam and geological structure development,which has poor adaptability.Secondly,the existing automatic follow-up control of hydraulic support is carried out under the mode of program customization,lacking the adaptive ability to the changes of external environment,and the level of intelligence is low;thirdly,the automatic follow-up technology is only based on the hydraulic support's own control,lacking of coordination with shearers,scraper conveyors and Emulsion Pumping stations,which makes the supply of Emulsion Pumping Stations insufficient.As a result,the hydraulic support moves slowly,the straightness of scraper conveyor can not be perceived,and it is difficult to advance continuously and automatically.The shearer,hydraulic support,scraper conveyor and emulsion pump station of fully mechanized mining face constitute the key equipment operation system of fully mechanized mining face.The highly adaptive control system of each equipment and the cooperative operation of each equipment are the key to realize the intelligent working face.In order to solve the above problems,this paper takes the following control of hydraulic support as the research object,and makes a deep research on the key technologies of self-adapting working condition of "three machines and one pump" and on-demand liquid supply of emulsion pump under the following mode of hydraulic support,straightness adjustment of scraper conveyor,position and attitude adjustment of hydraulic support,and traction speed control of shearer under the memory cutting mode of shearer.The main achievements are as follows:(1)Aiming at the problem that the actual time of hydraulic support moving with machine is short of data support by manual test,the working space of shearer and hydraulic support and the movement law of hydraulic support are analyzed and studied.Based on the principle of mathematical statistics,the calculation model of the actual time of hydraulic support moving with machine is established,and the dynamic calculation of the actual time of hydraulic support moving with machine is put forward.Method of calculation.The results of data test show that a reasonable prediction value of the time of frame removal can be found for each primary mining face.(2)Aiming at the technical problems such as the lack of timely and inadequate lateral and longitudinal deflection adjustment,the existing adjusting mechanism is not perfect,does not adapt to the changes of geological conditions,and lacks the basis of adaptive control,the theoretical model of hydraulic support deflection adjustment is established based on the basic mechanics theory,and the self-adjusting deflection adjustment and centralized control suitable for hydraulic support in fully mechanized mining are put forward.The shelving technology.According to the principle of adaptive control technology,the self-adjusting deviation device of hydraulic support is designed and developed.(3)In order to solve the problem that the emulsified pump station can not fully realize the automatic on-demand liquid supply in the process of moving the hydraulic support with the machine,it needs to adjust the cooperation of several emulsified pumps manually,a decision-making model of starting and stopping emulsified pump based on the input/output data of the controlled system is proposed.Aiming at the problem of data imbalance affecting model misjudgment rate,cost-sensitive learning method is used to allocate data weights.Classification tree is trained with 26950 data in actual production process.The misjudgment rate of decision-making is less than 9%.The results show that the model can realize the self-adaptive start-up and shutdown of multiple emulsifying pumps,and complete the on-demand liquid supply under the mode of hydraulic support following machine.(4)Aiming at the problem of adaptive straightness control of scraper conveyor,a rectangular calculation model of the chute in the lateral moving bend section of the scraper conveyor in working face is established,and the functional relationship among the horizontal angle,the number of chutes N and the propellance B of the chute of scraper conveyor is revealed.At the same time,the straight-line positioning and moving method of scraper conveyor is put forward,and the linear positioning and moving model of scraper conveyor is established.The condition that the scraper conveyor moves one step B at a time and the oblique length Lx and the horizontal section Lw formed should satisfy is that the horizontal angle between the chutes should not be more than 3 degrees.The horizontal direction of chutes in the initial stage of straight-line positioning and moving in the normal cycle stage is determined.Amount of movement.Practice shows that the scraper conveyor can realize self-straightening in the course of moving by adopting linear positioning technology.(5)In view of the fact that the existing memory cutting technology of shearer is not suitable for the change of coal seam thickness and dip angle,a method of obtaining continuous mining height data along the inclination direction by Kriging interpolation estimation algorithm is proposed to guide shearer memory cutting self-adapting height adjustment.At the same time,based on the theory of model predictive control(MPC)algorithm,the speed of shearer is taken as the control target and the hydraulic support is used.The multi-output regression tree algorithm consisting of multiple trees is designed to realize rolling optimization function,and the predictive regression tree control model of adaptive speed regulation of shearer is established.The underground industrial test results show that the control method can effectively avoid the bracket movement delay caused by insufficient liquid supply,solve the problem that the scraper conveyor straightness is not balanced,shorten the cycle time of automatic coal cutting,reduce the number of working face,and improve the production efficiency of working face.The technology has stronger adaptability to complex mining geological conditions and wider application scope.The research results of this paper are of great significance to realize mechanization,automation,intellectualization and informationization,reduce casualties and promote the development of "unmanned and less manned" in fully mechanized mining face. |
PDF Full Text Request