Abstract
A novel approach based on genetic algorithms (GA) is developed to find a global minimum-jerk trajectory of a space robotic manipulator in joint space. The jerk, the third derivative of position of desired joint trajectory, adversely affects the efficiency of the control algorithms and stabilization of whole space robot system and therefore should be minimized. On the other hand, the importance of minimizing the jerk is to reduce the vibrations of manipulator. In this formulation, a global genetic-approach determines the trajectory by minimizing the maximum jerk in joint space. The planning procedure is performed with respect to all constraints, such as joint angle constraints, joint velocity constraints, joint angular acceleration and torque constraints, and so on. We use an genetic algorithm to search the optimal joint inter-knot parameters in order to realize the minimum jerk. These joint inter-knot parameters mainly include joint angle and joint angular velocities. The simulation result shows that GA-based minimum-jerk trajectory planning method has satisfactory performance and real significance in engineering.