Autonomous Robotic Path Planning Based on the Gaussian Mixture Model in Complex Manufacturing Environment

Abstract

In the automotive manufacturing process, industrial robot path planning, which relies on offline programming by engineers, is usually time-consuming and difficult to migrate. To solve this problem, this paper proposes a path segment directed evolution algorithm (PSDEA) based on the Gaussian mixture model and a heuristic optimization algorithm. First, in the narrow and complex manufacturing environment, the Gaussian mixture model based on obstacles is used to calculate the collision probability of the robot arm in different poses. Secondly, the initial path is determined using the path generation configuration library, and a segmented path fitness function is introduced to evaluate the quality of the path quantitatively. Then, a directed evolution strategy is proposed to improve the genetic algorithm and achieve precise guidance of the path evolution direction; finally, the effectiveness of the proposed method is verified through simulation experiments and real scenarios. The results show that compared with the benchmark methods, the proposed method can improve the planning efficiency, reduce the path length, and be more robust. Therefore, the method proposed in this paper can quickly generate high-quality industrial robot terminal motion trajectories in complex manufacturing environments.

Keywords

• Manipulator
• path planning
• genetic algorithms
• Gaussian mixture model
• collision probability