| Industry 5.0,as a new industrial production mode where people,machines,objects,and the environment interact,collaborate,and integrate with each other,puts forward higher requirements for the automation,collaboration,and accuracy of mobile robots due to its people-oriented characteristics.In this context,starting from the overall design of the system,it is an important problem to explore the high-precision positioning of mobile robots based on multi-source information fusion,and develop its motion planning and control based on this.The proper resolution of this issue is of great significance for enhancing the working ability of mobile robots and enhancing the market competitiveness of enterprises.By consulting the relevant literature,this paper first studied and analyzed the research status of mobile robot positioning,motion planning and control at home and abroad;Secondly,with the construction of a navigation control system for a multi-source fusion SLAM mobile robot as the core,relevant technical research has been carried out from three aspects: system overall design,multi-source sensor fusion positioning,mobile robot path planning,and motion;Finally,based on the above research,a mobile robot positioning and navigation experiment was conducted to verify the accuracy and effectiveness of the proposed method.The main research content of this article is as follows:Firstly,based on the analysis of the advantages and disadvantages of different gear train mobile robot layouts,the experimental research platform for a rear wheel drive front wheel steering mobile robot was determined by combining practical needs;Secondly,the kinematics model and simplified model of the mobile robot are established based on the kinematics of the robot.Thirdly,the communication system of the mobile robot is designed;Finally,the communication requirements between the control system and external modules of the mobile robot were analyzed,and the corresponding communication system was designed.A localization method for mobile robots based on multi-source sensor fusion has been proposed.Firstly,the positioning principle of Li DAR sensors was analyzed;On this basis,by filtering the laser point cloud,the computational complexity during the use of the laser point cloud is reduced and its accuracy is improved;Secondly,based on the factor graph model and its algorithm theory,a localization framework for sensor factors and multi-source fusion was established,transforming the state estimation problem of mobile robot localization and navigation into a maximum posterior estimation problem,and solving it through least squares;Finally,a multi-source fusion positioning framework scheme based on factor graphs was adopted,greatly improving the accuracy and robustness of mobile robot positioning.The proposal of this method effectively solves the problems of missing inter frame data in low-frequency Li DAR and cumulative error in IMU inertial navigation.At the same time,adding GNSS positioning information further improves the positioning effect,providing positioning evaluation indicators for the navigation control of mobile robots.A path planning scheme for mobile robots based on high-precision semantic maps has been proposed.Firstly,using Li DAR point clouds for Euclidean clustering analysis;On this basis,the K-Dtree method is adopted to accelerate the retrieval speed of 3D LiDAR point clouds and reduce clustering time;Secondly,based on the Li DAR point cloud map as the base map and semantic annotation,a high-precision map of the mobile robot was constructed,providing a corresponding decision-making basis for its global path planning;Finally,aiming at the tracking problem of fixed low-speed mobile robots in straight segments and small curvature paths,an adaptive preview distance path tracing method is proposed,which uses the least square method to fit the path and then calculate the curvature.By dynamically adjusting the preview distance of the tracking algorithm,the lateral error and variance of the mobile robot tracking are reduced,and the stability and accuracy of the tracking are greatly improved.Based on the above content,a mobile robot verification platform has been established,and the navigation control of mobile robots based on multi-source fusion SLAM has been successfully implemented.The effectiveness and accuracy of the research method in this paper have been verified through examples. |
PDF Full Text Request