Parameter Identification of Motors by Cuckoo Search Using Steady-State Relations
| Autor: | Carlos Fuentes-Silva, Omar Rodríguez-Abreo, Jose Miguel Hernandez-Paredes, Alejandro Flores Rangel, Francisco Velasquez |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
0209 industrial biotechnology
Steady state (electronics) General Computer Science Mean squared error Computer science 020208 electrical & electronic engineering General Engineering Cuckoo search metaheuristic 02 engineering and technology Function (mathematics) DC motor Steiglitz-McBride algorithm TK1-9971 020901 industrial engineering & automation 0202 electrical engineering electronic engineering information engineering General Materials Science Electrical engineering. Electronics. Nuclear engineering parameter estimation Metaheuristic Algorithm Induction motor Voltage |
| Zdroj: | IEEE Access, Vol 9, Pp 72017-72024 (2021) |
| ISSN: | 2169-3536 |
| Popis: | The direct current (DC) motors are widely used; therefore, they are subject to multiple studies, different control techniques or analyses require a dynamic DC motor model. The parameters are needed to complete the model, which can be challenging to obtain. Therefore, multiple parametric estimation techniques have been developed. This paper presents a metaheuristic cuckoo search algorithm modified for motors as a parametric estimation tool. A cost function is based on the current and velocity error obtained when an input voltage step is applied to the motor. The main difference with similar works is that we used the steady-state equations to determine the parameters. The algorithm proposed is compared with the Steiglitz–McBride and the original cuckoo search algorithms to evaluate its performance objectively. Simulated and experimental results show that the algorithm proposed can calculate the parameters with better accuracy than the original cuckoo search and Steiglitz–McBride. The modifications made to the original algorithm of the cuckoo search allowed finding the values of the parameters motor with a root mean square error of less than 0.1% for signals obtained with simulation and less than 1% for real signals sampled at 0.001 s. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|