Induction Motors Parameters Identification by Starting Process Using Quantum Particle Swarm Optimization-Trust Region Algorithm (QPSO-TRA)

A thorough understanding of the induction machine parameters is necessary for practically all applications’ control techniques in order to maintain high-quality dynamic and steady-state drive performance. However, as it directly affects the effectiveness of the control systems, the higher accuracy of IM parameters is crucial for all industrial processes. Out of the five electrical parameters ($R_{\mathrm{s}},\ R_{\mathrm{r}},\ X_{\mathrm{s}},\ X_{\mathrm{r}}$, and $X_{\mathrm{m}}$), only four parameters, $R_{\mathrm{s}},\ R_{\mathrm{r}},\ X_{\mathrm{s}}$, and $X_{\mathrm{m}}$, can be independently estimated by the parameter estimation algorithms disclosed in the current literature, while the core loss resistance $R_{\mathrm{f}\mathrm{e}}$ is ignored. The method for predicting induction machine parameters during startup is presented in this study without the use of any assumptions. Two instantaneous voltage and current waveforms recorded during a direct startup are used to estimate the parameters. The six electrical parameters of the IM (Rs, Rr, Rfe, Xs, Xr, and Xm) can all be independently determined using the suggested parameter estimating approach utilized in this study. The weight of computing is also a significant issue for real-time data processing. This work offers a less time-consuming approach based on QPSO-TRA approach to address these issues. The suggested approach involves transforming the drawbacks of earlier algorithms to create a trustworthy hybrid algorithm to discover all IM parameters devoid of any assumptions. Comparing this hybrid method to the traditional genetic algorithm (GA), the computation load can be cut in 1/3. A 5. 5kW induction motor was used to assess the effectiveness of the suggested strategy using the Matlab software program.