Enabling High-fidelity Modeling of Digital Twin for Hydraulic Systems: KP-PSO Based Parameter Identification

• Published in: 2023 International Conference on Advanced Robotics and Mechatronics (ICARM) pp. 663 - 668
• Main Authors: Cui, Zhexin, Yue, Jiguang, Tao, Wei, Xia, Qian, Chen, Jichang, Wu, Chenhao
• Format: Conference Proceeding
• Language: English
• Published: IEEE 08-07-2023
• Subjects: Computational modeling; Digital twins; Hydraulic systems; Mathematical models; Parameter estimation; Prototypes; Uncertainty
• DOI: 10.1109/ICARM58088.2023.10218888

Hydraulic systems have been widely applied in complex equipment due to their high power density, high reliability, and high servo precision. Digital twin (DT) is a key technology to support its monitoring, control, and predictive maintenance. However, enormous challenges are involved in the high-fidelity DT modeling of hydraulic systems, within the uncertainty and invisibility of time-varying parameters, leakage, and friction. This paper proposes a knowledge perturbation particle swarm optimization (KP-PSO) for parameter identification to bridge the discrepancy between the DT model and the physical equipment. A DT modeling framework for hydraulic systems is established, where the mathematical DT prototype is derived to simulate the physical entity. Since only the output of the physical entity is required, the framework accommodates information-poor and non-intrusive conditions. KP-PSO is presented and utilized to accurately identify the recessive parameters of DT, and its superiority is pre-verified on recognized test functions. Finally, the physical experiment demonstrates the effectiveness of the proposed method.