Digital Twin-Driven Controller Tuning Method for Dynamics

Abstract

The control performance of the control system directly affects the running performance of the product. In order to solve the problem that the dynamics characteristics of mechanical systems are affected by the performance degradation of the controller, a digital twin-driven PID controller tuning method for dynamics is proposed. In this paper, firstly, the structure and operation mechanism of digital twin model for PID controller tuning are described. By using the advantages of virtual real mapping and data fusion of digital twin model, combined with the online identification of controlled object model, the problems of real-time feedback of actual control effect of the controller and unreal virtual model of control system caused by time-varying working conditions are effectively solved, and the closed-loop self-tuning of PID controller is realized. At the same time, intelligent optimization algorithm is integrated to improve the efficiency and accuracy of PID controller parameter tuning. Secondly, the modeling method of digital twin model is described from three aspects of physical prototyping, twin service system, and virtual prototyping. Finally, the controller tuning for gear transmission stability is taken as an example to verify the practicability of the proposed method.

Future of precision manufacturing: Integrating advanced metrology and intelligent monitoring for process optimization

“…Furthermore, the digital twin approach enables predictive maintenance and virtual prototyping, allowing for the monitoring, simulation, control, optimization, and identification of defects and trends within ongoing processes (Warke et al, 2021). Digital twin technology also facilitates the synchronization between physical prototyping and virtual prototyping, providing strong support for efficient closed-loop self-tuning of controllers (Ikwue et al, 2023;He et al, 2021).…”

A Survey on AI-Driven Digital Twins in Industry 4.0: Smart Manufacturing and Advanced Robotics

“…In this respect, Kabaldin et al [ 154 ] selected RNN to estimate the statistical model of the dynamic state in cutting. ML models, such as ANN [ 155 ], and probabilistic modeling methods, such as the Gaussian process [ 156 , 157 , 159 ], could likewise be adopted to develop a surrogate model and implemented in a control context [ 157 , 158 , 160 ]. Alternatively, model order reduction techniques can transfer highly detailed and complex simulation models to other domain and life cycle phase, e.g., building efficient finite element model for dynamic structural analysis through reducing the degree of freedom, while maintaining required accuracies and predictability [ 161 , 162 , 163 ].…”