Research And Application Of The Path-planning For Robot Arm Collision-avoidance

In the past thirty years, the research of robotics has developed rapidly, and the imagination and creativity from the public make robotics research seem to be more endless and diversely. In this period, the robot with the technology gradually mature, develops from the simple selection, painting and welding robot to a more complex functional robot, Such as, insertion of the IC chip robot, printed circuit board robot, mobile car parts handling robots, robotic delivery, and other minimally invasive surgical robot Robotics etc. at the same time, the development of robotics research in universities, which has become a major research direction in many universities.The research direction of this paper is to study the problem of autonomous path planning in the work space with an arbitrary structure and arbitrary degree of freedom mechanical arm. In this problem, robot path planning is transformed into the inverse kinematics problem, and then the inverse kinematics problem is transformed into the mathematical problem of the objective function, and then the robot path planning problem is solved.The problem of robot kinematics is divided into forward kinematics problem and inverse kinematics problem. At present, the solution to forward kinematics problem has general method, D-H Notation method and screw theory can be used to analyze and solve, using to the classic PUMA and SCARA robot structure to describe the two algorithms in detail. However, the inverse kinematics problem does not exist a general solution method, this article first introduces several commonly used methods, such as, geometry, matrix inversion method, screw theory and Lie algebra and genetic algorithms and neural network algorithm,then analyzes the limitations of the commonly used algorithms above, then an algorithm to inverse kinematics problem is proposed,which transforms inverse kinematics problem into the mathematical maximum problem of the multidimensional variable function, for solving the mathematical problem, the conventional ant colony algorithm is improved,and using the advantages of ant colony algorithm heuristic and optimization, coupled with the fuzzy control updating pheromone, makes the improved ant colony algorithm convergence more quickly than before, and solving more fast and stability better than unrevised ant colony algorithm. Then fuzzy ant colony algorithm proposed is applied to the experimental robotic arm autonomous path planning in workspace with obstacles,the problem is solved ultimately, and this proposed method is a general method for robotic arm in any structure, any degree of freedom.In order to verify the feasibility of the algorithm, this paper firstly uses PUMA robot to make MATLAB simulation for verifying the algorithm feasiblity, and then apply experimental robot to make actual control, driver board is STM32, core controller is PC motherboard for directly controlling, MATLAB software is directly used via USB serial port connected to the driver board, the control result is better, achieving the goal standard, proves that the algorithm is feasible.