A Two-Stage Mono- and Multi-Objective Method for the Optimization of General UPS Parallel Manipulators

This paper introduces a two-stage method based on bio-inspired algorithms for the design optimization of class general Stewart platforms. The first stage performs mono-objective in order to reach, with sufficient dexterity, regular target workspace while minimizing elements’ lengths. For this problem, we compare three algorithms: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Boltzman Univariate Marginal Distribution (BUMDA). second looks most suitable gains Proportional Integral Derivative (PID) control via minimization two conflicting objectives: one energy consumption tracking error trajectory. To effect, multi-objective Multiobjective Evolutionary Decomposition (MOEA/D) Non-dominated Sorting Algorithm-III (NSGA-III). main contributions lie model, proposal method, findings performance different each stage. Furthermore, show optimized designs delivered by proposed provide directions best-performing through metrics statistical hypothesis tests.