Online Parameter Identification of Linear Induction Motors Based on Improved Interconnected Full-Order Observer

Due to the special structure and dynamic end effect of linear induction motors, the change mechanism and law of their excitation inductance and secondary loss resistance are complicated. In order to improve the identification accuracy and performance of the observer for excitation inductance and secondary loss resistance, an online dual-parameter identification method of linear induction motors based on an improved interconnected full-order observer was proposed. Firstly, based on the T-type equivalent circuit of the linear induction motor considering dynamic end effects, the state space equations with dual-parameter changes were established, and the influence of parameter changes and coupling characteristics on motor poles was analyzed. Secondly, to reduce the impact of parameter coupling on identification accuracy, a low-coupling identification model with dual-parameter interconnection was established, and an interconnected full-order adaptive observer was designed. The adaptive laws for online identification of excitation inductance and secondary resistance were derived using Popov hyperstability theory, realizing online dual-parameter identification. Then, to improve the stability and convergence speed of the observer, a feedback gain matrix was derived and designed by using a novel pole configuration method. Finally, a simulation model and hardware-in-the-loop identification model were built for experimental verification. The results show that the new full-order adaptive observer achieves excitation inductance and loss resistance identification errors of around 0.01% during the startup acceleration phase and around 0.03% during dynamic loading. © 2024 Science Press. All rights reserved.