Design Of Autonomous Navigation Mobile Robot

Robotics is a highly integrated discipline, and the mobile robot technology is one of its branches. Positioning is the key issues of robot navigation. This paper studies localization method of a sensor module which is three encoders connecting with omni-directional wheels. This method meet certain accuracy requirements of robot positioning, and a short time autonomous navigation of the robot can be finished. If combined with the external environment sensor, completely autonomous navigation function can be realized. A reliable, modular electronic hardware platform of mobile robot was designed, which had a strong versatility and scalability. A uncoupled software architecture of robot was built, its structure is clear, portability and functionality scalability. The main jobs of this paper are following here.1. Finished the analysis of the kinematic model of the mobile robot and methods of navigation. The kinematics model of two differential mobile robot and three omni-directional wheels mobile robot in Geometric method was derived. Based on the kinematics model, this paper analyzed the localization method of three encoders which connect with omni-directional wheels. The positioning parameter calibration method to correct the geometric deviation was proposed. Based on Robot positioning, the tracking of a fixed-point, a straight-line trajectory and a fixed arc were analyzed, and realized the navigation, In the process of the tracking of a curve consisting of discrete points, the algorithm of perspective line was used to adjust the velocity of robot motion. Finally, based on the theoretical analysis, a simulation model of robot autonomous movement navigation was build.2. The platform of three omni-directional wheel robot was build. The positioning sensors consisted of three encoders which connect with omni-directional and electronic compass, gyroscope. The robot’s electronic control system was modular designed and used STM32F103ZE6as its main chip.3. The acquisition and processing the positioning sensor data was finished. The collection method of the data of quadrature incremental photoelectric encoder was introduced. Data acquisition of the ADXRS453which is a single-axis MEMS gyroscopes and electronic compass was completed. The low-pass filter and Kalman filter was used to process the original signal of gyroscope. Angle data of gyroscopes and electronic compass was fused by Kalman filter which played the respective advantages of each other.4. Planning and design of the robot control software system was finished. The Robot control software based on standardized library of STM32F10x and the μC/OS-Ⅱ real-time embedded system. Various functional modules of the robot is divided into the Nodes and Topics with which Nodes exchange data between each other. This method reduced the coupling of the various parts of software.This subject finally realized the autonomous navigation of three omni-directional wheel robot, with the positioning accuracy of3cm-5cm. It has laid a solid foundation for the upgrade research work of the mobile robot autonomous navigation.