Evaluation of Particle Swarm Optimization and Adaptive Genetic Algorithm for Motion Planning in Minimally Invasive Surgery

ABSTRACT: This paper evaluates Adaptive Genetic Algorithm (AGA) and Particle Swarm Optimization (PSO) to find a time optimal quadratic polynomial trajectory of an anthropomorphic manipulator. This robot that is used in minimally invasive surgery (MIS) must achieve motions under the constraints of displacement, velocity, acceleration and jerk of each joint. The modeling and resolution of the constraints are presented. PSO and different selection methods of the genetic algorithm are evaluated and compared in order to define the best one according to convergence speed and optimal time. These methods can solve the premature convergence and slow convergence problems in MIS. Simulation and experimental results for the grasper of a compact laparoscopic surgical robot prototype system validate the algorithms.