Parameter identification of induction motors using dynamic encoding algorithm for searches (DEAS)

Abstract

In this paper, a newly developed optimization algorithm, called the dynamic encoding algorithm for searches (DEAS), is introduced and applied to the parameter identification of an induction motor for vector control and fault detection. Digital simulations are conducted on startup with no load and normal operation with load perturbations. DEAS is compared with the continuous-time prediction error method and the genetic algorithm via identification performance using the startup signals. The capability of onload identification using the proposed technique is also verified with transient signals. Consequently, DEAS is shown to locate more precise parameter values than both the compared methods especially with much faster execution time than the genetical algorithm.