System identification by using RGA with a reduced-order robust observer for an induction motor. (October 2018)

• Record Type: Journal Article
• Authors: Chen, Kun-Yung; Yang, Wen-Hao; Fung, Rong-Fong
• Is Part Of: Mechatronics. Volume 54(2018)
• Page Start: 1
• Page End: 15
• Publication Date: 2018-10
• Subjects: Induction motor (IM) -- Real-coded genetic algorithm (RGA) -- Reduced-order robust observer (RRO) -- System identification; Computer integrated manufacturing systems -- Periodicals; Flexible manufacturing systems -- Periodicals; Mechatronics -- Periodicals; Productique -- Périodiques; Fabrication, Systèmes flexibles de -- Périodiques; Mécatronique -- Périodiques; Computer integrated manufacturing systems; Flexible manufacturing systems; Mechatronics; Periodicals; 629.89
• DOI: 10.1016/j.mechatronics.2018.07.001
• Languages: English
• ISSNs: 0957-4158

Abstract

Abstract: In this paper, system identification by using real-coded genetic algorithm (RGA) with a reduced-order robust observer (RRO) for an induction motor (IM) is successfully proposed and realized. Because the rotor flux linkages of an IM are not easily measurable, the RRO is proposed to observe them. Different state errors including the measurable and observed states are used in the fitness functions, which are optimized by the RGA to find the unknown system's parameters. From numerical simulations and experimental results, unknown parameters of an IM are successfully identified by the RGA with state-error fitness functions. The contributions of this paper are: (i) the RRO is combined in the mathematical model of the IM to observe the un-measureable rotor flux linkages; (ii) The observer gain of the reduced-order observer is given as the unknown parameters, which is selected by the RGA approach to satisfy the linear matrix inequality (LMI) conditions; (iii) Three fitness functions are successfully proposed, compared and optimized by using the RGA. It is found that the one with more real state errors has the best identify performance than the other two; (iv) From numerical and experimental results, it can be concluded that the more system's states are measurable and used in the fitness function, the more system's parameters are accurately identified.