Update Method of Digital Twin Model of Industrial Robot Based on Deep Reinforcement Learning
The digital twin model of industrial robot can simulate the behavior and performance of industrial robot in the real world,but its simulation accuracy will be reduced due to the influence of service conditions such as scene update and equipment wear.In this paper,an update method of digital twin model of industrial robots based on deep reinforcement learning is proposed.In this method,the simulation tool Coppeliasim is used to establish the digital twin model of industrial robots.At the same time,the key parameters of the digital twin model such as PID parameters and joint damping are optimized based on the depth deterministic strategy gradient(DDPG)algorithm,so as to realize the parameter update of the model and improve the model accuracy.Finally,the simulation experiment of ABB-IRB2400 industrial robot is carried out to verify the effectiveness of the proposed method.
Deep reinforcement learningIndustrial robotDigital twinDepth deterministic strategy gradient(DDPG)Model updating
万方数据
维普
