The Optimal B-spline Based On Genetic Algorithm And Its Use In The Robot Trajectory Planning

As the science and technology growing, the demand of the robot's performance become higher, robots should not only be capable of moving in conventional PTP manner and simple CP manner, but also be capable of moving along a specified path of arbitrary shape at high velocity, in order to realize the less time in the motion. To satisfy such demand, a profound research about the time optimum trajectory planning of robots along a specified path is carried out in this paper. The specified path is assumed to be composed of a series of discrete path points, and the whole time optimum trajectory planning is decomposed into two steps: in the beginning, cube-spline is employed to approach these discrete path points, then genetic algorithm is employed to generate a time optimum trajectory along the B-spline.With the algorithm, the robot is guaranteed to moved in an optimum manner and not deviate from its original path.Firstly, the current situation and development prospect of the technology and applications for the robot was introduced in this paper. The research achievements of time-optimal trajectory planning of robot at home and abroad were summarized. At the same time, the drawbacks of the current trajectory planning were present. According to principle of time-optimal trajectory planning, the optimization of B-spline based on genetic algorithm was proposed. The optimal B-spline was applied in the time-optimal trajectory planning of robot. In this way, the work efficiency was significantly improved, and then general problems in robot trajectory planning were analyzed. In joint space, cubic interpolation, high-order polynomial interpolation and linear interpolation which was joined by the parabola were detailed accounted separately at different situations. In Cartesian space, linear interpolation and arc interpolation are discussed deeply.Secondly, the basic principle of robotic time-optimal trajectory planning was studied. The paper discussed the theory of genetic algorithm and the calculation method of B-spline for constructing trajectory of robot. A method of time-optimal trajectory planning based on genetic algorithm was proposed because of the property of B-spline.In this way, the total time which used by the robot sporting along the whole B-spline could be translated into the sum of part time of the robot sporting along each part of B-spline. The constraint conditions of joint angle velocity and joint angle acceleration were discussed. The wrist joint, manipulator joint, and waist joint of KLD-600 robot were optimally calculated. The algorithm of trajectory planning simulations implemented by the Matlab, its result showed that the total time of each joint arriving the target point reduced, and the trajectory of every joint is smooth, so it demonstrated the efficiency of the proposed algorithm.Finally, the hardware construction of KLD-600 robot was introduced. The communication principle between PMAC motion controller and master was studied. The dialog box of user and machine was built with VC. The real-time communication between main-machine and submitted-machine was realized by calling the Pcomm32. The robot kinematics model and inverse kinematics model of KLD-600 were set up through D-H method. The optimal path points in Cartesian space in teaching operation were obtained by genetic algorithm, in this way, the time-optimal trajectory planning was implemented.