Research And Application Of Robot Technology Based On Multi-Sensor Information Fusion

The packaging and printing industry,serving various fields,is playing an increasingly important role in the continuous development of the social economy.However,the traditional packaging and printing industry in China faces significant problems such as low efficiency,high energy consumption,and lack of market competitiveness.Therefore,it is crucial for these traditional factories to transform and upgrade towards intelligent packaging and printing factories.Autonomous mobile robots have important application value as the"bridge" for intelligent material scheduling in intelligent packaging and printing factories.This paper studies the technology of autonomous mobile robots based on multi-sensor information fusion,aiming to achieve simultaneous localization and mapping,autonomous navigation,and automatic obstacle avoidance for robots in intelligent packaging and printing factories.The main research contents are as follows:(1)Research on SLAM based on multi-sensor information fusion.This paper design a SLAM algorithm that utilizes multi-sensor information fusion to address the insufficient perception of autonomous mobile robots by integrating laser radar,depth camera,IMU and odometer data.The algorithm’s mapping performance was evaluated using structural similarity index,mean square error and side length sum error metrics through SLAM mapping simulation experiments conducted in a constructed physical simulation environment for intelligent packaging and printing factories,alongside Hector,Gmapping,and Cartographer algorithms.Results showed that the proposed SLAM algorithm achieved a superior mapping effect.(2)Research on robot navigation and obstacle avoidance.To achieve the goals of autonomous navigation and automatic obstacle avoidance of robots,depth camera information was added to the original overall framework of robot navigation and obstacle avoidance,and a combination path planning algorithm combining improved A~*algorithm and DWA algorithm was designed to obtain a navigation solution that combines multi-sensor information fusion and combination path planning.Then,simulation experiments were conducted in the simulation environment to achieve the goal of autonomous navigation and automatic obstacle avoidance for robots.(3)This paper presents experimental verification and result analysis of the proposed SLAM algorithm and navigation solution.SLAM mapping experiments were conducted in two scenarios after configuring the experimental platform’s communication and calibrating the depth camera.Results demonstrate that the proposed SLAM algorithm overcomes the impact of factors such as long corridors,large loops,and multiple similar structures on mapping performance and achieves mapping of obstacles with heights below the two-dimensional laser scanning plane.Robot navigation and obstacle avoidance experiments were conducted in scenario 2,verifying that the navigation solution utilizing multi-sensor information fusion and combination path planning can achieve the functional requirements of autonomous navigation and automatic obstacle avoidance for robots.Localization experiments were conducted in a known environment,achieving a mean squared error of 0.0525m for robot localization,satisfying the requirements for robot localization in intelligent packaging and printing factories.This paper utilizes multi-sensor information fusion to enhance robots’ perception capability for themselves and their surrounding environment,addressing the deficiencies in robot SLAM mapping,navigation,and obstacle avoidance processes,and achieving the application requirements of autonomous mobile robots in intelligent packaging and printing factories.