Research On Path Planning For Mobile Robot On The Platform Of Arm

Path planning of robot is a kernel content of research in intelligent robot area, which started in twentieth Century 70’s. So far, there have had a lot of research results been reported. Artificial intelligence technology applied to obtain optimal paths is to study mobile robot path planning with artificial intelligence technology. For example, swarm intelligence computation, fuzzy logic and information fusion et al. It is the relatively recent progress over path planning technologies.The development and research situation at home and abroad of mobile robot were presented in this thesis, and the main aspects of research on mobile robot path planning technology at present were described. The improved ant colony algorithm and its simulation analysis were studied in this paper, and also experimental procedure of mobile robot path planning based on a chaotic ant colony algorithm, analysis of the main parameters’ influence on the performance of the algorithm by simulation and the research of algorithm convergence. We took ARM as the core processor of the robot, then autonomously designed mobile robot hardware platform. At the same time, the obstacle avoidance of mobile robot on this platform is studied.Firstly, this paper introduced some file system types on the embedded Linux and how to transplant root file system. We researched on Linux device driver functions and structure, and mainly discussed the kernel mechanism for driver. We introduced the configuration of S3C6410 MPU timer, and also PWM function settings. Secondly, the ant colony algorithm was deeply studied. Improved the traditional ant colony algorithm, the effectiveness of our algorithm was compared by simulation with the existing algorithms. Simulation results showed that comparing with traditional ant colony algorithm, ant colony hierarchical search algorithm had stronger global search capability and faster convergence speed. That the algorithm cost function value was a sequence of submartingale process was proved with martingale theory. A stopping time of the submartingale was given. Moreover, the algorithm converged with probability 1 was proved.Thirdly, mobile robot path planning based on ant colony algorithm was intensively studied. We focused on the establishment of simulation environment and the model, the simulation effect and determination of the parameters of ant colony algorithm. Joined the chaos operator, the performance of ACS algorithm is improved. Simulation results show that chaotic ant colony algorithm not only has fast convergence speed, but also has high precision even on complex map.At last, two aspects of the hardware platform and software environment of embedded system were studied. The output PWM of the S3C6410 accessed stepping motor driver chip, to driver stepper motor for the control of robot moving direction. At the same time, we transplanted Linux operating system and designed Linux device driver based on the ARM platform.