Parameter Identification of Flexible Drive Systems using Particle Swarm Optimization

A Master of Science thesis in Mechatronics Engineering by Ishaq Hafez entitled, “Parameter Identification of Flexible Drive Systems using Particle Swarm Optimization”, submitted in April 2023. Thesis advisor is Dr. Rached Dhaouadi. Soft copy is available (Thesis, Completion Certificate, Approval Signatures, and AUS Archives Consent Form).

Abstract

This thesis proposes a hybrid optimization method called Hybrid Particle Swarm Optimization with Quasi-Newton (HPSO-QN) for accurately identifying the mechanical parameters of a two-mass model (2MM) system commonly used in high-performance electric drive systems with elastic joints. Accurate identification of mechanical parameters, such as motor and load inertias, shaft stiffness, and friction disturbance coefficients, is essential for achieving optimal control performance. The HPSO-QN method integrates the Quasi-Newton’s (QN) local exploitation capabilities with Particle Swarm Optimization’s (PSO) global exploration capabilities to achieve accurate parameter identification. The frequency response of the system was obtained by employing several excitation signals and analyzed using the Frequency Response Function (FRF) analysis method to estimate the resonant frequencies. The FRF and time-response analysis were then utilized to extract the benchmark parameters that served as a reference for the system parameters identification using PSO methods. The experimental results from a 2MM system validated the performance of the proposed optimization method, which demonstrated superior accuracy and efficiency compared to standard PSO algorithms. The optimization method effectively identified the mechanical parameters of the 2MM system, with potential implications for improving the modeling of the 2MM, leading to better performance and stability. The proposed method builds on previous work in the field, including the use of stochastic algorithms, FRF, and time-response analysis methods for parameter identification. The HPSO-QN method demonstrates improved accuracy and performance and can be extended to other systems with flexible shafts and couplings. This thesis contributes to the development of more accurate and effective parameter identification methods for complex systems, highlighting the importance of accurate parameter estimation for optimal control performance and stability.