基于粒子群优化神经网络自适应控制算法的并联机器人仿真研究

摘要

针对传统控制算法对并联机器人轨迹跟踪精度控制效果不好的问题，提出了一种并联机器人的改进粒子群优化神经网络自适应控制算法，首先对粒子群优化算法进行惯性权重的优化和变异操作的改进，然后用改进的PSO算法优化神经网络的初始权值并进行在线调节PID参数。最后以六自由度并联机器人为研究对象，将传统PID控制与基于改进PSO优化的神经网络自适应控制算法分别进行了仿真实验。仿真结果表明，在快速性和稳定性能上，基于改进PSO优化的神经网络自适应控制算法比单纯的PID控制更加优越，在一定程度上减小了轨迹输出的误差并且提高了轨迹跟踪精度。%Traditional control algorithm to control the problem of bad effects on trajectory tracking precision of parallel robot ,this paper proposes a improved parallel particle swarm optimization neural network robot a-daptive control algorithm based on the first PSO algorithm to optimize inertia weight and improved contrac-tion factor, the improved PSO algorithm is then applied to optimize the initial weights of neural network and adjust PID parameters on-line. Finally,a 6-DOF parallel robot is chosen as the study object. The traditional PID control and improved PSO optimize neural network adaptive control algorithm are simulated. Simulation results prove that the performance of improved PSO optimize neural network adaptive control algorithm is far superior to the traditional PID control on the rapidity and stability, and reduce the output trajectory error and improved the trajectory tracking precision to a certain extent.