شناسایی پارامترهای بارهای الکتریکی با استفاده از ساختار چند متغیره مبتنی بر یادگیری عمیق

محورهای موضوعی : انرژی های تجدیدپذیر

امید ایزدی قهفرخی ¹ , مزدا معطری ² , احمد فروزان تبار ³

1 - دانشکده مهندسی برق- واحد مرودشت، دانشگاه آزاد اسلامی، مرودشت، ایران
2 - مرکز تحقیقات مکاترونیک و هوش مصنوعی- واحد مرودشت، دانشگاه آزاد اسلامی، مرودشت، ایران
3 - دانشکده مهندسی برق- واحد مرودشت، دانشگاه آزاد اسلامی، مرودشت، ایران

تاریخ دریافت : 1400/05/02 تاریخ پذیرش : 1401/03/24 تاریخ انتشار : 1402/12/01

کلید واژه: مدل‌سازی بار, سیستم اندازه‌گیری گسترده, ساختار یادگیری عمیق چند متغیره, تابع تلفات, شبکه بازگشتی حافظه‌دار,

چکیده مقاله :

مدل‌سازی بار یکی از وظایف ضروری در مطالعات سیستم‌های قدرت محسوب می شوند. با توسعه سیستم‌های قدرت این مسئله بیش از پیش پیچیده تر شده ‌است. روش‌های پیشین مدل‌سازی بار دارای عیوب اساسی مانند الف) حساسیت بالا به نویز، ب) عدم لحاظ همگرایی بارهای الکتریکی در یک شبکه، ج) وابستگی به مدل ریاضی، د) بار محاسباتی بالا و ه) وابستگی به اندازه‌گیری محلی هستند. برای رفع این مشکلات، در این مقاله یک ساختار مبتنی بر یادگیری عمیق توسعه داده شده است که قادر به شناسایی تعداد زیادی از پارامترهای بار به صورت همزمان با سرعت و دقت مطلوب است. ساختار طراحی شده قادر به درک کامل ویژگی‌های زمانی بر مبنای یک ساختار حافظه‌دار بازگشتی است. همچنین، برای تخمین تعداد متغیرهای زیاد یک روش اختصاص‌دهی وزن برای این مدل توسعه داده شده ‌است. نهایتأ، یک تابع تلفات فرمول‌بندی شده ‌است تا مقاوم ‌بودن ساختار در برابر با نویز را افزایش دهد. مطالعات عددی بر روی شبکه 68-شینه IEEE موثر بودن و برتری روش پیشنهادی را در مقایسه با تعدادی از روش‌های کم‌-عمق و عمیق را نشان می دهد.

چکیده انگلیسی:

Electrical load modeling has been considered an essential task in power system studies. With the recent development of power systems, load modeling is becoming more and more challenging. The previous methods on load modeling are suffered from: i) high sensitivity to noise; ii) neglecting the load correlation in a power system, iii) high computational burden, and iv) dependency on the local measurement devices. To address these problems, this paper develops a deep neural network-based structure that can identify a large number of parameters simultaneously with fast performance as well as high accuracy. The designed network can fully understand the temporal features using a gated recurrent neural network-based structure. Furthermore, to provide the ability to estimate a large number of load parameters, a technique to assign the learning weight has been developed. Consequ­ently, to enhance the robustness of the designed network considering noisy conditions, a loss function has been developed in this paper. The numerical results on the IEEE 68-bus system demonstrate the effectiveness and superiority of the proposed network in comparison with several shallow-based and deep-based structures. 

منابع و مأخذ:

[1] O.I. Ghafarokhi, M. Moattari, A. Forouzantabar, "Composite load modeling by spatial-temporal deep attention network based on wide-area monitoring systems", Journal of Intelligent and Fuzzy Systems, vol. 40, no. 6, pp. 12215-12216, June 2021 (doi: 10.3233/JIFS-210296) .

[2] A. Arif, Z. Wang, J. Wang, B. Mather, H. Bashualdo, D. Zhao, "Load modeling- A review", IEEE Trans. on Smart Grid, vol. 9, no. 6, pp. 5986-5999, Nov. 2018 (doi: 10.1109/TSG.2017.2700436).

[3] X. Zhang, D.J. Hill, C. Lu, "Identification of composite demand side model with distributed photovoltaic generation and energy storage", IEEE Trans. on Sustainable Energy, vol. 11, no. 1, pp. 326-336, Jan. 2020 (doi: 10.1109/TSTE.2019.2890868).

[4] N.K. Neto, A.R. Abaide, V. Miranda, P.V. Gomes, L. Carvalho, J. Sumaili, D.P. Bernardon, "Load model­ing of active low-voltage consumers and comparative analysis of their impact on distribution system expans­ion planning", International Trans. on Electrical Energy Systems, vol. 29, no. 8, e12038, Aug. 2019 (doi: 10.1002/2050-7038.12038).

[5] H. Renmu, M. Jin, D.J. Hill, "Composite load modeling via measurement approach", IEEE Trans. on Power Systems, vol. 21, no. 2, pp. 663-672, May 2006 (doi: 10.1109/TPWRS.2006.873130).

[6] P. Regulski, D.S. Vilchis-Rodriguez, S. Djurović, V. Terzija, "Estimation of composite load model par­am­eters using an improved particle swarm optimization method", IEEE Trans. on Power Delivery, vol. 30, no. 2, pp. 553-560, Feb. 2015 (doi: 10.1109/TPWRD.2014.2301219).

[7] V. Knyazkin, C.A. Canizares, L.H. Soder, "On the parameter estimation and modeling of aggregate power system loads", IEEE Trans. on Power Systems, vol. 19, no. 2, pp. 1023-1031, May 2004 (doi: 10.5370/JE­ET.2­0­12.7.3.304).

[8] P. Jazayeri, W. Rosehart, D.T. Westwick, "A multistage algorithm for identification of nonlinear aggregate power system loads", IEEE Trans. on Power Systems, vol. 22, no. 3, pp. 1072-1079, July 2007 (doi: 10.1­00­7/s­40565-018-0469-2).

[9] Q. Liu, Y. Chen, D. Duan, "The load modeling and parameters identification for voltage stability analysis", Int. Conf. on Power System Technology, vol. 4, pp. 2030-2033, Oct. 2002 (doi: :10.1109/PSCE.200­4.13­97­702).

[10] S.M. Rizvi, K.S. Sajan, A.K. Srivastava, "Real-time parameter tracking of power-electronics interfaced composite ZIP load model", IEEE Trans. on Smart Grid, Oct. 2021 (Early Access) (doi: 10.1109/TSG.20­21.3119507).

[11] A. M. Najafabadi A. T. Alouani, "Real time estimation of sensitive parameters of composite power system load model", Proceeding of the IEEE/TDC, pp. 1-8, May, Orlando, FL, USA, 2012 (doi: 10.1109/TDC.201­2.62­81427).

[12] A. Rouhani, A. Abur, "Real-time dynamic parameter estimation for an exponential dynamic load model", IEEE Trans. on Smart Grid, vol. 7, no. 3, pp. 1530-1536, May 2016 (doi: 10.1109/TSG.2015.2449904).

[13] S. Afrasiabi, M. Afrasiabi, B. Parang, M. Mohammadi, S. Kahourzade, A. Mahmoudi, "Two-stage deep learning-based wind turbine condition monitoring using SCADA data", Proceeding of the IEEE/PEDES,  pp. 1-6, Jaipur, India, Dec. 2020 (doi: 10.1109/PEDES49360.2020.9379393).

[14] A. Qian, G.B. Shrestha, "An ANN-based load model for fast transient stability calculations", Electric Power Systems Research, vol. 76, no. 4, pp. 217-227, Jan. 2006 (doi: 10.1016/j.epsr.2005.06.001).

[15] M. Ganjouri, M. Moattari, A. forouzantabar, M. Azadi, "Short-term load forecasting using a graph-based deep learning structure", Journal of Novel Researches on Electrical Power, vol. 9, no. 4, pp. 37-46, Feb. 2021.

[16] J. Ma, Z.Y. Dong, P. Zhang, "Using a support vector machine (SVM) to improve generalization ability of load model parameters", Proceeding of the IEEE/PES, pp. 1-8, Seattle, WA, USA, Mar. 2009 (doi: 10.110­9/PS­CE.2009.4839969)

[17] L. Chávarro-Barrera, S. Pérez-Londoño, J. Mora-Flórez, "An adaptive approach for dynamic load modeling in microgrids", IEEE Trans. on Smart Grid, vol. 12, no. 4, pp. 2834-2843, Mar. 2021 (doi: 10.1109­/TSG.2­021.3­064046).

[18] S.M.H. Rizvi, K. S. Sajan, A. K. Srivastava, "Synchrophasor based ZIP parameters tracking using ML with adaptive window and data anomalies", IEEE Trans. on Power Systems, vol. 37, no. 1, pp. 3-13, June 2022 (doi: 10.1109/TPWRS.2021.3088903).

[19] M. Cui, M. Khodayar, C. Chen, X. Wang, Y. Zhang, M.E. Khodayar, "Deep learning-based time-varying parameter identification for system-wide load modeling", IEEE Trans. on Smart Grid, vol. 10, no. 6, pp. 6102-6114, Jan. 2019 (doi: 10.1109/TSG.2019.2896493).

[20] C. Wang, Z. Wang, J. Wang, D. Zhao, "SVM-Based parameter identification for composite ZIP and electronic load modeling", IEEE Trans. on Power Systems, vol. 34, no. 1, pp. 182-193, Jan. 2019 (doi: 10.1109/TPWRS.2018.2865966).

[21] M. Afrasiabi, M. Mohammadi, M. Rastegar, L. Stankovic, S. Afrasiabi, M. Khazaei, "Deep-based conditi­onal probability density function forecasting of residential loads", IEEE Trans. on Smart Grid, vol. 11, no. 4, pp. 3646-3657, July 2020 (doi: 10.1109/TSG.2020.2972513).

[22] M. Afrasiabi, M. Mohammadi, M. Rastegar, A. Kargarian, "Probabilistic deep neural network price forec­ast­ing based on residential load and wind speed predictions", IET Renewable Power Generation, vol. 13, no. 11, pp. 1840-1848, Aug. 2019 (doi: 10.1049/iet-rpg.2018.6257).

[23] M. Afrasiabi, M. Mohammadi, M. Rastegar, S. Afrasiabi, "Advanced deep learning approach for probabilistic wind speed forecasting", IEEE Trans. on Industrial Informatics, vol. 17, no. 1, pp. 720-727, Jan. 2021 (doi: 10.1109/TII.2020.3004436).

[24] S. Afrasiabi, M. Afrasiabi, M. Mohammadi, B. Parang, "Fault localisation and diagnosis in transmission networks based on robust deep Gabor convolutional neural network and PMU measurements", IET Generation, Transmission and Distribution, vol. 14, no. 26, pp. 6484-6492, Dec. 2020 (doi: 10.1049/iet-gtd.2020.0856).

[25] S. Afrasiabi, M. Mohammadi, M. Afrasiabi, B. Parang, "Modulated gabor filter based deep convolutional network for electrical motor bearing fault classification and diagnosis," , IET Science, Measurement and Technology, vol. 15, no. 2, pp. 154-162, Mar. 2021 (doi: 10.1049/smt2.12017).

[26] S. Afrasiabi, M. Afrasiabi, B. Parang, M. Mohammadi, M.M. Arefi, M. Rastegar, "Wind turbine fault diagnosis with generative-temporal convolutional neural network", Proceeding of the IEEE/EEEIC, pp. 1-5, Genova, Italy, June 2019 (doi: 10.1109/EEEIC.2019.8783233).

[27] H. Samet, S. Ketabipoor, M. Afrasiabi, S. Afrasiabi,M. Mohammadi, "Deep learning forecaster based controller for SVC: Wind farm flicker mitigation", IEEE Trans. on Industrial Informatics, Sept. 2020 (Early Access) (doi: 10.1049/iet-gtd.2019.128).

[28] D. Bahdanau, K. Cho, Y. Bengio, "Neural machine translation by jointly learning to align and translate", arXiv Preprint, Article Number: 1409.0473, Sept. 2014 (doi: 10.48550/arXiv.1409.0473).

[29] D.P. Kingma, J. Ba, "Adam: A method for stochastic optimization", arXiv Preprint, Article Number: 1412.6980, Dec. 2014 (doi: 10.48550/arXiv.1412.6980).

[30] S. Afrasiabi, M. Afrasiabi, B. Parang, M. Mohammadi, "Designing a composite deep learning based differential protection scheme of power transformers", Applied Soft Computing, vol. 87, Article Number: 105975, Feb. 2020 (doi: 10.1016/j.asoc.2019.105975).

[31] A.K. Singh, B.C. Pal, "IEEE PES task force on benchmark systems for stability controls report on the 68-bus 16-machine 5-area system", Technical Report, IEEE Power and Energy Society, Dec. 2013 (https://eprints.lincoln.ac.uk/id/eprint/28771).