Navigation Design Of Mobile Robot Based On ROS

With the rapid development of science and technology,the mobile robot is playing an increasingly important role in human's daily life,and has been widely concerned by people.The mobile robot navigation design involves research on environment perception,Simultaneous Localization and Mapping(SLAM)and path planning.In this thesis,Robot Operating System(ROS)is used as a development platform to study mobile robot navigation from the aspects of SLAM and path planning.In terms of SLAM,an improved traditional Rao-Blackwellised Particle Filter(RBPF)algorithm is used to improve the accuracy of mobile robot pose estimation and environment map establishment in this thesis.In terms of path planning,based on improved ant colony algorithm and dynamic window method,a fusion algorithm named Ant Colony Optimization Dynamic Window Approach(ACO-DWA)is proposed,which enables mobile robots to achieve autonomous obstacle avoidance in a dynamic environment and to select the best path to reach the specified target.The main research contents of this thesis are as follows.(1)Optimal design of SLAM algorithm.In order to reduce the number of sampling particles and the complexity of the calculation process,this thesis optimizes the traditional RBPF-SLAM algorithm,adds sample observations to the odometer model,improves important density distribution functions,sets the fusion ratio of the odometer and radar observation model,reduces the gap between the particle importance density distribution and the target distribution,and successfully reduces the number of sampled particles.In order to ensure the diversity of sample particles,this thesis judges the weights of effective particles,retains samples with medium weight,and implements a resampling strategy for samples with high and low weights to ensure the diversity of particles.Finally,the effectiveness is verified by comparing the simulation comparison results before and after the improved SLAM algorithm.(2)Optimal design of path planning algorithm.Path planning and design include global path planning and local path planning.First of all,in the global path planning,this thesis proposes an improved ant colony algorithm,which optimizes the initialization part of the ant colony algorithm and the pheromone update mechanism to obtain the optimized path,then performs the second path planning on this basis,and finally gets the optimal path.The improved ant colony algorithm improves the convergence performance of the algorithm and shortens the global path length.Secondly,the Dynamic Window Approach(DWA)is used in local path planning,and different obstacle avoidance strategies are designed.When a mobile robot encounters a dynamic obstacle with an unknown motion law,a second-level safety distance determination rule is proposed to improve the obstacle avoidance function of DWA in complex environments,which realizes partial obstacle avoidance and obtains the optimal path.Finally,this thesis combines these two algorithms together to propose an ACO-DWA fusion algorithm based on improved ant colony algorithm and dynamic window method,which enables mobile robots to achieve autonomous obstacle avoidance in a dynamic environment to obtain global information.The simulation verifies the effectiveness of the ACO-DWA algorithm.(3)System design.Mobile robot navigation design includes four parts: overall system architecture design,simulation model building,system hardware platform and software platform design.The overall system architecture design includes the host computer interaction layer,function realization layer,and hardware execution layer.The simulation model is built to provide the model parameters of the mobile robot for later SLAM and path planning research,which is convenient for the physical design and testing of the mobile robot body.The hardware platform design mainly includes the selection and design of the main control board,motor and drive circuit,lidar sensor,etc.,and completes the assembly and construction of the hardware platform.The software platform design mainly includes communication mechanism and TF transformation to complete the tasks of SLAM and path planning.(4)System performance test and experimental analysis.In this thesis,tests and analyses are carried out in a virtual environment and a real environment respectively.The GAZEBO platform is used to create a three-dimensional space scene,and a virtual model is built to perform three-dimensional simulation experiments with improved RBPF-SLAM algorithm and ACO-DWA fusion algorithm and further verify the SLAM and path planning functions of the mobile robot in the real environment.Experimental results show that the improved RBPF-SLAM algorithm is more accurate for pose estimation and trajectory estimation,and improves SLAM performance.The effectiveness of ACO-DWA fusion algorithm in obstacle avoidance performance,convergence speed and path planning in complex environment is verified.