Research On Technologies Of Passenger Train Compartment Cleaning Robot

With the continuous development of modern science and technology,service robots are gradually entering the market and people’s homes.As a widely used service device,cleaning robots liberate people from heavy cleaning labor and improve people’s quality of life and work efficiency.China is a big country with a huge number of rail trains.However,at present,the cleaning of the carriage of passenger trains is completely responsible by human.The area of train carriage is large,and the efficiency of manual cleaning is low at the area below the seat.At the same time,the carriage environment is complex,it is difficult to apply the traditional small cleaning robot.Aiming at the above problems,this paper studies the technologies of cleaning robots used in the carriage of passenger trains,and proposes an intelligent cleaning robot design that can proceed complete coverage cleaning in the environment of carriage.The main research contents of this paper are listed below:Aiming at the limited cleaning ability of traditional mobile robots,a folding cleaning mechanism is designed based on the characteristics of application environment,and combined with the mobile robot platform,so the coverage space of mobile robot in the carriage environment is expanded.In this paper,a planar two-degree-of-freedom serial manipulator with simple,stable and practical structure is used as the extension cleaning mechanism,and two driving modes of the manipulator with stepper motors and steering engines are designed.After the scheme design is completed,the kinematics of the manipulator and the Mecanum wheel mobile robot platform are analyzed,which lays a foundation for the follow-up research of complete coverage path planning.Aiming at the problems of low coverage capability and high path repetition ratio of traditional complete coverage path planning algorithm,this paper combined with the complete coverage path planning algorithm based on biologically inspired neural network,has proposed connectivity detection algorithm and path backtracking algorithm.The algorithm can automatically generate a complete coverage path of environment in both known and unknown environment with low path repetition ratio.The dynamic deadlock escape algorithm combined with neural network is proposed to improve the path planning ability of the algorithm in complex environment.The proposed algorithm does not need motion template and energy loss function and has higher feasibility and significant advantages compared with traditional complete coverage path planning algorithm.The cooperative complete coverage strategy of mobile robot and folding manipulator based on regional decomposition method is proposed.The grid types are redefined according to the different working spaces of the folding cleaning mechanism and the mobile robot,and the map is decomposed into the same sub-region grid maps according to the characteristics of the environmental map and the distribution of obstacles using the improved rectangular regional decomposition method.In the subregions,an improved biologically inspired neural network algorithm is used for mobile robot platform.For the folding manipulator,different covering strategies are formed according to the different coverable grids of the folding manipulator when the position of the mobile robot is fixed.A value function based on time cost is proposed,and the dynamic programming algorithm is used to evaluate each strategy.The optimal covering strategy is found when the total number of grids is fixed.All regions are connected by biologically inspired neural network point-to-point path planning algorithm to realize efficient cooperative cleaning of folding cleaning mechanism and mobile robot in the environment map of the carriage.Finally,experiments and simulations are made to verify each research content.Two kinds of planar two-degree-of-freedom folding manipulators with different driving modes are processed,and the motion experiment are carried out.The experimental results meet the requirements of the cooperative complete coverage path planning algorithm.The grid modeling of the real carriage environment map and complete coverage path planning simulation are carried out.The complete coverage path planning of the whole environment is completed by defining the grid attributes,decomposing region,planning complete coverage path of mobile robot and folding manipulator in sub-region,and planning point-to-point path between different regions.The simulation results verify the feasibility of the proposed algorithm and have the possibility to be extended to the environment with similar characteristics.