Research And Application Of Key Technologies Of Intelligent Dredging Robot For Underground Sump In Coal Mine

Underground sump stores large amounts of coal slurry and wastewater produced during coal mining and is an important facility for preventing mine flooding and a production system that every mine must be equipped with.However,underground sump dredging is carried out under harsh environmental conditions,the water content in the coal slurry is large,and it often becomes semi-fluid or fluid after churning,so periodic dredging requires a lot of manual labor and cannot guarantee the life safety of workers.With the construction of intelligent mines and the increased awareness of mine flood safety,autonomous clearance has become the trend.According to the actual production demand of underground sump,an intelligent dredging robot that can continuously and autonomously clean the sump is designed to provide a reference for the development of various types of autonomous robots for underground sump slime cleaning and to promote the intelligent and unmanned process of underground sump cleaning.Key technologies of intelligent dredging robot for coal mine underground sump are fuzzy PID control and Simultaneous Localization and Mapping(SLAM)technology for slime cleaning density in the dredging process,and the specific research includes the following aspects:Firstly,the physicochemical characteristics of coal slurry in water bunker and the parameters of pumping system are analyzed and the data conditions are provided for establishing the step fuzzy PID control model of coal slurry density step in dredging process,and the effectiveness of this model is verified by simulation.Secondly,the algorithm principles of Kalman filter algorithm and standard particle filter algorithm are analyzed,the mathematical model of SLAM problem and underground sump dredging robot model are established,and the map construction of SLAM problem is analyzed and selected,so that the SLAM problem of underground sump dredging robot is decomposed into dredging work positioning problem and map building problem.Furthermore,the FastSLAM algorithm is introduced and its principle is analyzed.The binary tree optimization is used to reduce the time overhead and computational complexity of the original algorithm,and the particle swarm optimization algorithm is used to improve the poor global convergence and depletion of the particles of the FastSLAM algorithm so that each particle integrates the joint influence of individual particles and group particles,and the position and weight values of the particles are continuously optimized and updated.Without increasing the number of particles,the real posterior probability distribution of the system is approximated,which in turn makes the dredging robot closer to the real system state distribution,and simulation experiments are conducted for the improved algorithm under the barrier-free area and underground sump raster model to analyze the effects of different numbers of wayfinding points and odometer parameters on the differences in algorithm performance and to verify the effectiveness of the improved algorithm.Finally,the design of the mechanical system,hydraulic system and intelligent control system of the dredging robot is introduced,and the application debugging of the built dredging robot in the underground sump is carried out for one dredging cycle,and the debugging data are analyzed to further verify the effectiveness of step fuzzy PID-based slime density control and FastSLAM based on particle swarm optimization algorithm in the dredging process.