Optimal trajectory and genetic algorithm based energy minimization for robot manipulators

Abstract

Robotic Manipulators are playing the great role in any successful industry nowadays. Several categories exist to classify these manipulators, Rigid or Flexible manipulators is one classification. This work investigates both types, rigid and flexible, in terms of Mathematical Modelling, Trajectory Selection, and Optimal Trajectory Planning. Optimization of trajectory can be achieved using several criterions, such as: minimum time trajectory, minimum velocity and acceleration, minimum residual vibration, or minimum energy consumed. GA will search for the trajectory that will best satisfy the objective function set by the requirement of the application. First, a three D-O-F planar redundant manipulator that is working in the joint space is proposed, and the torques of the three joints have been simulated using Soft motion and Linear Segment with Parabolic Blends (LSPB) trajectories. Then, the proposed planar manipulator has been investigated in terms of optimal trajectory planning in the task space. The manipulator's end-effector was required to track two specified paths, a straight line and a circle. The angle that the last link makes with the horizontal line was assumed to endure a Cubic Spline trajectory while tracking both paths. GA was implemented to obtain the optimal trajectory by selecting the proper initial and final configurations of that angle. Finally, a Two-Links Flexible Manipulator is modelled and analyzed. It is required to optimize the actuators' energy consumption when the manipulator performs a certain movement in the joint space. Soft motion trajectory, which is a fully identified trajectory, is employed here also to benchmark any results resulting from a less identified trajectory as the fourth order polynomial trajectory. The fourth order trajectory's parameters are obtained using GA such that the resulting trajectory is the optimal one. The simulation results verified the effectiveness of GA, since the optimized fourth order polynomial trajectory needed less energy than the soft motion trajectory when the same movement is executed.