Practical simulator generation using a simple feedforward element design method for precision motion systems
Year: 2025
Type: article
Abstract: The growing demand for automation and labor-saving solutions has intensified interest in digital technology applications that replicate real-world systems and utilize analytical results in physical settings. Fulfilling these needs requires the development of a high-precision simulator that can be automatically and effortlessly generated from data easily obtainable from a physical machine. This paper proposes an innovative method for generating a precision simulator for motion systems. The proposed method uses feedforward elements derived through our feedforward design methodology. In this approach, following a specified procedure, a learning controller distinguishes four operational characteristics based on specific types of motion profiles, and the derived elements are combined according to predetermined rules to produce a simulator. We applied this method to a ball-screw mechanism and assessed the performance of the generated simulator by comparing it with experimental data. In the experimental setup, we observed position-dependent submicrometer vibrations, which cannot be represented by conventional mechanical models. The simulator, constructed by combining FF elements from high-resolution measured or reconstructed data, replicated vibrations with similar frequencies. However, its ability to effectively mimic certain responses, including the observed submicrometer vibrations, was limited. Notably, the vibration amplitude varied with velocity. Nevertheless, the results demonstrated that straightforward refinement of the simulator generation process could produce a simulator with high simulation accuracy.
Source: Precision Engineering
Authors: Kaiji Sato, Mizuki Takeda
Institution: Toyohashi University of Technology
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,340
Funder: Tokyo Institute of Technology
OpenAlex ID: W4416543416