基于IPSO优化的MPC轨迹跟踪控制研究
Research on MPC Trajectory Tracking Control Based on IPSO Optimization

针对机械臂轨迹跟踪过程中因外界干扰和固定控制器参数难以应对动态环境变化而导致的跟踪精度不足问题，本文提出了一种基于改进粒子群算法的机械臂自适应模型预测轨迹跟踪控制策略。通过机械臂动力学方程建立系统模型，并设计用于轨迹跟踪的模型预测控制器，引入粒子群优化算法对控制器中的权重系数进行在线调整。采用线性惯性权重下降策略，克服了传统粒子群算法易陷入局部最优的问题，同时提高了收敛速度。仿真实验结果表明，所提出的改进粒子群优化模型预测控制器具有更稳定、精准的跟踪性能，相较于原版控制器和传统粒子群优化控制器，关节角均方误差分别降低了33.85%和23.08%。
To address the issue of insufficient tracking accuracy in robotic arm trajectory tracking due to external disturbances and the inability of fixed controller parameters to adapt to dynamic environments, this paper proposes an adaptive model predictive trajectory tracking control strategy based on an improved particle swarm optimization (PSO) algorithm. The system model is established using the dynamic equations of the robotic arm, and a model predictive controller (MPC) is designed for trajectory tracking. The PSO algorithm is introduced to adjust the weight coefficients of the controller online. A linear inertia weight decay strategy is employed to mitigate the problem of traditional PSO algorithms easily falling into local optima, while also improving convergence speed. Simulation experiments demonstrate that the proposed improved PSO-based MPC achieves more stable and accurate tracking performance, with the mean squared error of joint angles reduced by 33.85% and 23.08% compared to the original controller and the traditional PSO-improved controller, respectively.