2021/01/13 by Balakumar Balasingam, Balasingam, Balakumar, Krishna R. Pattipati +1
Computer Science · Engineering · #Advanced Battery Technologies Research #Blind Source Separation Techniques #Control Systems and Identification #FOS: Electrical engineering #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2101.05349
openalex publication_date 2021/01/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Real-time identification of electrical equivalent circuit models is a\ncritical requirement in many practical systems, such as batteries and electric\nmotors. Significant work has been done in the past developing different types\nof algorithms for system identification using reduced equivalent circuit\nmodels. However, little work was done in analyzing the theoretical performance\nbounds of these approaches. Proper understanding of theoretical bounds will\nhelp in designing a system that is economical in cost and robust in\nperformance. In this paper, we analyze the performance of a linear recursive\nleast squares approach to equivalent circuit model identification and show that\nthe least squares approach is both unbiased and efficient when the\nsignal-to-noise ratio is high enough. However, we show that, when the\nsignal-to-noise ratio is low - resembling the case in many practical\napplications - the least squares estimator becomes significantly biased.\nConsequently, we develop a parameter estimation approach based on total least\nsquares method and show it to be asymptotically unbiased and efficient at\npractically low signal-to-noise ratio regions. Further, we develop a recursive\nimplementation of the total least square algorithm and find it to be slow to\nconverge; for this, we employ a Kalman filter to improve the convergence speed\nof the total least squares method. The resulting total Kalman filter is shown\nto be both unbiased and efficient in equivalent circuit model parameter\nidentification. The performance of this filter is analyzed using real-world\ncurrent profile under fluctuating signal-to-noise ratios. Finally, the\napplicability of the algorithms and analysis in this paper in identifying\nhigher order electrical equivalent circuit models is explained.\n