This paper presents the development of a Hardware-in-the-loop Simulation (HILS) environment for a single-link robot system and experimentally evaluates its ability to effectively represent the dynamic characteristics of the actual system. Simulation-b...
This paper presents the development of a Hardware-in-the-loop Simulation (HILS) environment for a single-link robot system and experimentally evaluates its ability to effectively represent the dynamic characteristics of the actual system. Simulation-based design has become an essential tool in controller development; however, model simplifications and the exclusion of real hardware effects often lead to discrepancies between simulation results and real-world system behavior. Although HILS techniques have been introduced to address these limitations, existing studies have primarily focused on large-scale and high-cost systems, while applications and experimental validations for relatively simple robotic systems remain limited.
To address this gap, a single-link robot incorporating both electrical and mechanical dynamics is selected as the control target, and a HILS environment employing a real microcontroller unit (MCU) as the controller is constructed. The controlled plant is implemented on a real-time simulator using a dynamic-equation-based model. To ensure consistency with the actual system, key parameters of the motor, gear, load link, and angular sensor are identified through a combination of experimental results, datasheet information, and CATIA-based modeling. In particular, for a DC motor with limited specification data, a rated-input-based parameter estimation procedure is applied, enabling the construction of a model that simultaneously satisfies experimental responses and manufacturer specifications.
The validity of the proposed HILS environment is verified by comparing simulation results, HILS experiments, and actual hardware experiments conducted under identical controller structures and control parameter settings. The comparison results show that the transient response characteristics are largely consistent across the three environments, while some discrepancies near steady-state conditions are observed due to friction and other nonlinear effects. Furthermore, response comparisons under varying controller gains confirm that the proposed HILS environment consistently reflects the behavior of the actual system without being restricted to specific operating conditions.
The results demonstrate that even for a simple robotic system, a HILS environment can serve as an effective intermediate validation stage in controller design and verification processes. This study extends the applicability of conventional HILS approaches beyond complex large-scale systems and provides a safe and efficient validation methodology for early-stage robot controller development.