针对Stewart 平台传统结构优化设计中,存在的设计过程低效、优化方案不全面以及优化结果不直观等问题,在深入分析Stewart 平台运动特性的基础上,建立平台的运动学方程,并通过仿真技术求解出平台的可达工作空间以及灵活工作空间。文中将多目标蚁狮(MOALO)算法应用于Stewart 平台的结构优化设计,以雅可比矩阵条件数以及可用操作度为优化目标,通过仿真软件得到多组优化解,即帕雷托优化解集;以用作运动模拟器的Stewart 平台为例进行具体的优化设计分析,通过对灵活工作空间体积占比的求解,验证了该算法的有效性和可行性。在Stewart 平台的结构优化设计中,MOALO 算法相较进化遗传算法、多目标粒子群算法等,在多目标优化问题上具有更好的收敛性和覆盖性,更符合实际多目标优化工程设计。

In response to the problems of inefficient design process, incomplete optimization schemes, and non-intuitive optimization results in the traditional structural optimization design of the Stewart platform, based on deep analysis of the kinematic characteristics of the Stewart platform, the kinematic equations of the platform are established, and the reachable workspace and flexible workspace of the platform are obtained through simulation technology. The multi-objective ant lion (MOALO) algorithm is applied to the structural optimization design of the Stewart platform, and with the Jacobian matrix condition number and available operability as optimization objectives, multiple sets of optimization solutions, namely Pareto optimization solution sets, are obtained through simulation software. Taking the Stewart platform used as a motion simulator as an example for specific optimization design analysis, the effectiveness and feasibility of the algorithm are verified by solving the volume ratio of flexible workspace. In the structural optimization design of the Stewart platform, the MOALO algorithm has better convergence and coverage in multi-objective optimization problems compared to evolutionary genetic algorithm, multi-objective particle swarm optimization algorithm, etc. , and it is more in line with practical multi-objective optimization engineering design.

TH112

西安市科技计划资助项目(22GXFW0065)