Optimal trajectory planning of industrial robot with evolutionary algorithm

This paper proposes new method based on evolutionary algorithm for optimal trajectory planning of robot manipulator (PUMA 560). The intelligent method is elitist non dominated sorting based genetic algorithm (NSGA-II). This method is superior as it considers multiple criterion to be optimised simultaneously. The multicriterion cost function has to minimize with defined constraints. The problem has 5 objective functions, 32 constraints and 252 variables. Pareto optimal is set of compromise solutions instead of single optimal solution. Pareto optimal fronts are evaluated by the solution spread measure and ratio of non dominated individuals. The B-spline function is used to define robot trajectories.