Reinforcement learning based adaptive trajectory planning and control system for industrial robots

Abstract

The present application relates to the technical field of intelligent control of industrial robots, in particular to an adaptive trajectory planning and control system for industrial robots based on reinforcement learning, comprising a data acquisition module, a state reward construction module, a decision model training module, an online decision control module and an instruction execution module. The data acquisition module collects multi-modal sensing data of the industrial robot, the rotary positioner and the sensing network of the processing area, the state reward construction module constructs model input states and generates reward signals according to the requirements of the composite manufacturing process, the decision model training module sets continuous actions containing joint micro position increments, laser power adjustments and feed speeds, and completes model training in a digital twin environment. The online decision control module outputs action sequences according to real-time states, and the instruction execution module converts them into corresponding driving instructions. The system realizes multi-parameter collaborative adaptive control, improves the synchronicity and real-time adaptation capability of trajectory planning and process adjustment.

Images