Collaborative Optimization Algorithm of Deep Learning and Compliant Control in Online Trajectory Planning

With the rapid development of robotics and automation, online trajectory planning faces problems such as poor adaptability to dynamic environments and low computational efficiency. This study aims to propose a collaborative optimization algorithm that combines deep learning and compliant control to improve the efficiency and accuracy of online trajectory planning. First, CNN is used for feature extraction and a prediction model is established. Then a compliant control strategy is designed to deal with dynamic obstacles to ensure the flexibility and stability of the robot in trajectory planning. The trained CNN model is then used to generate real-time trajectory predictions. The robot motion is adjusted through a fuzzy logic controller to optimize the relative position of the trajectory and obstacles. The average obstacle avoidance success rate of the algorithm in dynamic environments reaches 97%. The collaborative optimization algorithm of deep learning and compliant control effectively improves the real-time and accuracy of online trajectory planning, providing a new solution for the application of intelligent robots in complex environments.