Direct Trajectory Planning Method Based on IEPSO and Fuzzy Rewards and Punishment Theory for Multi-Degree-of Freedom Manipulators

Manipulator is a kind of commonly used multi-degree-of-freedom nonlinear system. There is a strong coupling between the links, and the movement of each link constraints and affects each other, which increases the difficulty of motion analysis of the system and reduces its trajectory planning efficiency under specific task targets. To solve this problem, a direct trajectory planning method based on an improved particle swarm optimization (PSO) algorithm, called IEPSO, and the fuzzy rewards and punishment theory is proposed in this paper. First, on the basis of preserving the local search ability of PSO, the global search ability of the population is improved by increasing a population exchange item. At the same time, in order to avoid the population falling into the local optimal value, the last elimination principle is incorporated into the standard PSO algorithm. Second, the fuzzy rewards and punishment theory is introduced to reduce the redundant decoupling operation, which can not only ensure the accuracy of manipulator trajectory planning but also effectively reduce the calculation amount of the trajectory planning for the multi-degree-of-freedom manipulator, to improve the optimization efficiency. Finally, the direct trajectory planning method of the multi-degree-of-freedom manipulator is compared and tested. It can be seen that the efficiency scalar and accuracy of the proposed direct trajectory planning method are significantly higher than those of other optimization methods.