Research On Location Methods Of Storage Robot Driven By Mecanum Wheels

With the rapid development of e-commerce and express industry and the continuous expansion of online retail market,the social demand for logistics services is increasing.Intelligent storage mobile robots have been widely used in logistics industry due to several disadvantages of traditional storage systems,such as low efficiency and huge expense on manpower and building a large transmission system.On the one hand,the intelligent storage mobile robots are mostly based on two-wheels differential driving at present,which require turning radius when changing directions.As the channels between some shelves in the warehouse are very narrow,the movement of differential driven robots in the warehouse is not flexible.On the other hand,positioning technology is one of the key technologies of intelligent storage mobile robots.Only after the accurate positioning can robots achieve the next navigation.Therefore,this dissertation takes the warehouse environment as the application background to study the positioning method of mobile robots driven by mecanum wheels.In order to solve the problem that the differential driven robots are not flexible in the warehouse with narrow aisle,the dissertation proposes a scheme that the storage mobile robots use the mecanum wheels to drive,and builds the hardware platform of the storage mobile robot driven by the mecanum wheels.In this dissertation,the kinematics equation and inverse kinematics equation of the robot are derived,and the dead reckoning positioning model is established.Aiming at the problem of large positioning error in this method,a positioning method using monocular camera to identify two-dimensional code is proposed to obtain the absolute pose information of robot with high accuracy.In order to solve the problem of large accumulated error in encoder-based positioning method and the problem that the positioning method based on twodimensional code can not meet the real-time positioning requirements,this dissertation proposes a positioning method based on EKF(Extended Kalman Filter)fusing encoder and IMU(Internal Measurement Unit),with two-dimensional code information for robot pose correction.The state equation is set based on the encoder information while the observation equation is set based on the IMU information.When the camera cannot recognize the two-dimensional code,the robot uses the encoder information and IMU information to locate,when the camera detects the two-dimensional code,the robot uses the tag information to locate.The validity of the localization method proposed in this thesis is verified on robot in the real world.In order to further improve the positioning accuracy,this dissertation uses a multisensor fusion positioning method based on EH_?F(Extended H Infinity Filter).Encoder information is used to set state equation while IMU and two-dimensional code information is used to set observation equation.The parameters of EH_?F algorithm are analyzed and selected,and experiments in real environment are carried out to verify that the method used in this dissertation can further improve the positioning accuracy of robot.