Application of Pareto-based Multiobjectives Genetic Algorithm in Minimum Time Motion Planning


This paper looks into the application of Pareto-based Genetic Algorithm (GA) in obtaining the minimum time motion planning for an industrial robot. A common practice in multiobjective optimisation GAs for minimum time motion planning is to apply classical aggregation approach to the objective formulation while in this study, Pareto-ranking method is used. A suitable objective vector is organised to include the total travel time and the two joint constraints: velocity and acceleration limits. The objective is to obtain a optimal motion with minimum travel time and within the kinematics limitations. The optimisation process involves first producing a fixed number of joint displacements using the genetic operators, and then scaling the travel time such that it is not violating the kinematics constraints. The feasibility of this method is shown by simulation results with an RTX SCARA robot. Cubic spline functions are used in the construction of the joint trajectory.