Robot Localization In Intelligent Warehouse Management System

Mobile robots have been widely used in industry, agriculture, military, service, hospital and many other fields, to expand human's ability and release them from heavy and dirty work. As the fast improvement of robot technology, it's very important to localize the robot rapidly and accurately. In Intelligent Warehouse Management System (IWMS), some common sensors such as infrared range, GPS can not reach the requirement of localization for the restriction of its environment. So, new localization technologies for IWMS are needed to obtain robot's position. QR Code (Quickly response Code) is a new matrix 2-D bar code with high quality. For its huge capacity, high reliability, low cost, easy manufacture and ability to encode Chinese characters and even figures, it can be used to localize robots in IWMS properly.In this dissertation, according to the environment of IWMS, we present a mobile robot localization algorithm by taking use of QR Codes. The algorithm also has been realized and tested in the plant of P3-DX. Firstly, the related research knowledge such as IWMS, robot localization, and QR Code is introduced. Secondly, mathematic model of P3-DX, which adopts two differential drives, is build, and its error model is analyzed. Thirdly, a case using QR Code to localize robots in IWMS is proposed based on the environment and the simple motion of robot in IWMS. Based on the case, the encoding, decoding and abstraction process of QR Code is introduced to obtain the robot's position. Fourthly, to get the related position of robot with QR code, the 3D pin-hole model of camera is analyzed, and by using Zhang's algorithm, the parameters of camera are calibrated. From the calibrating equation, the related position of robot with QR Code is gained. Finally, a localization algorithm combined Perception Filter (PF) with Extended Kalman Filter (EKF) is proposed by fusing the robot's encoder and camera's information, which is precise and effective. Moreover, the availability and reliability of the algorithm is simulated and tested.