A Genetic Algorithm Based, Non-linear Bi-level Optimization Path Planning Research For A Robotic Spray System

Robotic trajectory planning plays a very important role in robotic control domain. This thesis is presented the milestone and necessarily of robotic trajectory planning firstly. Then analyses several different types of robotic trajectory planning algorithm, the strong points of these algorithms are reviewed, the weak points are highlighted as well.This thesis is devoted to the study of generating a trajectory under the painting coverage area and the constraints. By the researching processes of concrete analysis of trajectory planning on robotic manipulator arm, imitation of trajectory based on kinematics and optimization of trajectory in the articulation space, this paper formulizes general regularity of trajectory planning.A novel approach is employed to generate this trajectory based on the requirements and constraints of spray robot-----Bi-level Trajectory Optimal Planning Problem(BLTOPP). The method researched resolves contradiction between real-time and accuracy in the operation space, and problem of minimal optimization about time.In order to increase the flexibility of robot arm to improved the productivity of paint spray. Thus, the changeable stream length method is employed in this thesis. The coverage area is changed with the stream length. Due to the influence of robot pose selection, coverage quality (minimum overlap area and no bank area) and coupled parameters, this optimal problem is a bi-level problem. In process of the trajectory planning depending on robotic inverse and forward kinematics, MTTP is adopted to optimal the middle nodal point. Due to the constraint of robotic dynamic is enhanced, the results is more accuracy. Genetic algorithm is employed to optimal the painting quality. In the end of this thesis, the simulation is conducted by MATLAB, and experiment is conducted on the real-world robot. The method is provided reasonable by results.