| With the development of industrial robot technology,intelligent mobile robot went into the vigorous development stage,in which positioning technology is the key technology to solve the mobile robot navigation and to achieve the automation of mobile robot.Compared to the mobile robot positioning in the outdoor environment,the posit ioning accuracy in the indoor environment is higher and more difficult to achieve.In this paper,a two-wheel differential indoor mobile robot is used as a platform to explore the indoor mobile robot positioning scheme theoretically and practically in a known structured environment with LIDAR,micro inertial measurement unit and encoder as the main sensors.Firstly,the kinematic model of two-wheel differential indoor mobile robot is established based on odometer and micro-inertia unit.An alignment algorithm for calculating the initial attitude of a robot is proposed in the case of uneven or inclined ground.The Kalman filter is used to fuse the data of the odometer and the inertial unit,and the roll angle and pitch angle information in the motion process are introduced into the location updating algorithm to help reduce the positioning errors caused by wheel slips or potholes and bumps of ground.Fina lly,simulation is carried out based on the odometer,micro inertia unit and odome ter/micro inertia combination,which verifies the correctness of the proposed location update algorithm.Secondly,a segmentation strategy based on Split-Merge is proposed for the problem of over-segmentation and over-integration in LIDAR data segmentation.The Hough transform method is used to fit the arc characteristic.The line segment and arc feature in the environme nt are extracted successfully.Thirdly,on the basis of the extraction of environmental characteristics,the initia l posit ioning scheme and the racking posit ioning scheme in moving process for indoor mobile robot are developed.The initia l posit ioning algo rithm is characterized by a complete line segment to match map and the micro-inertia unit is used to assist the LIDAR to achieve positioning.The optimized positioning algorithm has achieved better positioning in the non-similar and symmetrical similar environment.The motion tracking and posit ioning scheme adopts the tight ly coupled integrated method with using the unscented K alman filter as the integrated posit ioning filter,which reduces the computational complexity and improves the posit ioning accuracy.The simulation results show the effectiveness of the integrated positioning scheme proposed in this paper.Finally,the experimental platform of indoor mobile robot is constructed according to the positioning method in this paper.The experimental study is carried out in two non-similar environments and symmetric similar environments.The results show that the indoor mobile robot has achieved good posit ioning accuracy in both environments,which means,the feature extraction algorithm can fit the line segment and arc characteristics correctly,and the addition of the init ial alignment algorithm enables the LIDAR to identify the heading angle correctly in a symmetric similar environment.In addit ion,the integrated posit ioning scheme can estimate the initial position of the robot and the trajectory during the movement correctly. |
PDF Full Text Request