Fractional-Order Modeling and Identification for Dual-Inertia Servo Inverter Systems with Lightweight Flexible Shaft or Coupling

To effectively mitigate resonance in dual-inertia servo inverter systems with a lightweight flexible shaft or coupling, the precise modeling of the dual-mass mechanism is essential. This paper proposes a fractional-order modeling and identification methodology tailored for a dual-mass loading permanent magnet synchronous motor (PMSM) servo inverter system. By extending the traditional integer-order model to a more precise fractional-order one, the accuracy of resonance capture can be enhanced within the dual-inertia mechanism. Model parameters are identified using an output error approach combined with the Levenberg–Marquardt (LM) algorithm for fractional-order identification. To validate the effectiveness of this proposed methodology, a PMSM servo inverter experimental platform was developed, and identification experiments were conducted on this platform. The experimental results demonstrate that the proposed fractional-order modeling and parameter identification method significantly improves the system characterization accuracy of the dual-inertia servo inverter system.