Wang Chengjun, Yao Xingyu, Ding Fan, Yu Zhipeng
School of Artificial Intelligence, Anhui University of Science and Technology, Huainan 232001, China.
School of Mechanical Engineering, Anhui University of Science and Technology, Huainan 232001, China.
Math Biosci Eng. 2024 Feb 4;21(2):3364-3390. doi: 10.3934/mbe.2024149.
In order to meet the efficiency and smooth trajectory requirements of the casting sorting robotic arm, we propose a time-optimal trajectory planning method that combines a heuristic algorithm inspired by the behavior of the Genghis Khan shark (GKS) and segmented interpolation polynomials. First, the basic model of the robotic arm was constructed based on the arm parameters, and the workspace is analyzed. A matrix was formed by combining cubic and quintic polynomials using a segmented approach to solve for 14 unknown parameters and plan the trajectory. To enhance the smoothness and efficiency of the trajectory in the joint space, a dynamic nonlinear learning factor was introduced based on the traditional Particle Swarm Optimization (PSO) algorithm. Four different biological behaviors, inspired by GKS, were simulated. Within the premise of time optimality, a target function was set to effectively optimize within the feasible space. Simulation and verification were performed after determining the working tasks of the casting sorting robotic arm. The results demonstrated that the optimized robotic arm achieved a smooth and continuous trajectory velocity, while also optimizing the overall runtime within the given constraints. A comparison was made between the traditional PSO algorithm and an improved PSO algorithm, revealing that the improved algorithm exhibited better convergence. Moreover, the planning approach based on GKS behavior showed a decreased likelihood of getting trapped in local optima, thereby confirming the effectiveness of the proposed algorithm.
为了满足铸件分拣机器人手臂的效率和平滑轨迹要求，我们提出了一种时间最优轨迹规划方法，该方法结合了受成吉思汗鲨（GKS）行为启发的启发式算法和分段插值多项式。首先，根据手臂参数构建了机器人手臂的基本模型，并对工作空间进行了分析。采用分段方法将三次多项式和五次多项式相结合形成一个矩阵，以求解14个未知参数并规划轨迹。为了提高关节空间中轨迹的平滑度和效率，在传统粒子群优化（PSO）算法的基础上引入了动态非线性学习因子。模拟了受GKS启发的四种不同生物行为。在时间最优的前提下，设置目标函数以在可行空间内进行有效优化。在确定铸件分拣机器人手臂的工作任务后进行了仿真和验证。结果表明，优化后的机器人手臂实现了平滑连续的轨迹速度，同时在给定约束内优化了整体运行时间。对传统PSO算法和改进的PSO算法进行了比较，结果表明改进后的算法具有更好的收敛性。此外，基于GKS行为的规划方法陷入局部最优的可能性降低，从而证实了所提算法的有效性。
