应自动化学院邀请，俄罗斯科学院Denis Sidorov教授、伊尔库茨克国立理工大学Aliona Dreglea教授一行3人受邀参加自动化学院智慧能源先进控制研究所召开的科技部“一带一路”创新人才交流项目“Research on prediction and control technology of energy-storage power”系列研讨会，并作报告。
报告安排如下：
报告时间：2023年10月29号9:30-12:00
报告地点：中南大学民主楼210
1.俄罗斯科学院Denis Sidorov教授， Volterra integral equations with jump kernels: theory and numerical methods with application for storage control
2.伊尔库茨克国立理工大学Aliona Dreglea教授, Towards robust control of melt spinning process: theory and numerical methods
报告时间：2023年10月30号14:00-16:00
报告地点：中南大学科教北楼101
1.俄罗斯科学院Denis Sidorov教授，Volterra Black-box models identification methods:direct collocation
报告时间：2023年10月31号14:00-16: 00
报告地点：中南大学新校区D203
1.俄罗斯科学院Denis Sidorov教授，Design and Optimization of Net-zero Communities
报告时间：2023年11月1号9:00-12: 00
报告地点：中南大学信息楼316
1.俄罗斯科学院Denis Sidorov教授，Volterra Black-box models identification methods: least squares
2.伊尔库茨克国立理工大学Aliona Dreglea教授，Existence and numerical solutions of nonlinear BVP in the theory of boundary
3.伊尔库茨克国立理工大学Song Liu：Neural network fusion optimization for photoviltaic power forecasting
报告时间：2023年11月2号上午9:00-12: 00
报告地点：中南大学信息楼316（腾讯会议845438628）
1.俄罗斯科学院Denis Sidorov教授，Toward General Theory of Differential-Operator and Kinetic Models
2.伊尔库茨克国立理工大学Aliona Dreglea教授，Robust approach to detection of bubbles based on image analysis
3.伊尔库茨克国立理工大学Samad Noeiaghdam副教授：A novel numerical optimality technique to find the optimal results of Volterra integral equation of the second kind with discontinuous kernel（online）
报告时间：2023年11月3号上午9:00-12: 00
报告地点：中南大学信息楼316（腾讯会议753575327）
1.俄罗斯科学院Denis Sidorov教授，A Dynamic Analysis of Energy Storage With Renewable and Diesel Generation Using Volterra Equations
2.伊尔库茨克国立理工大学Aliona Dreglea教授，Risk stratification of intestinal anastomosis using machine learning methods
3.伊尔库茨克国立理工大学Samad Noeiaghdam副教授：The Best Approximation of Generalized Fuzzy Numbers Based on Scaled Metric（online）
报告摘要：
1.The Volterra integral-functional series is the classic approach for nonlinear black box dynamical systems modeling. It is widely employed in many domains including radiophysics, aerodynamics, electronic and electrical engineering and many other. Identifying the time-varying functional parameters, also known as Volterra kernels, poses a difficulty due to the curse of dimensionality. This refers to the exponential growth in the number of model parameters as the complexity of the input-output response increases. The least squares method (LSM) is widely acknowledged as the standard approach for tackling the issue of identifying parameters. Unfortunately, the LSM suffers with many drawbacks such as the sensitivity to outliers causing biased estimation, multicollinearity, overfitting and inefficiency with large datasets.This paper presents alternative approach based on direct estimation of the Volterra kernels using the collocation method. Two model examples are studied.It is found that the collocation method presents a promising alternative for optimization, surpassing the traditional least squares method when it comes to the Volterra kernels identification including the case when input and output signals suffer from considerable measurement errors.
2.A key role in the theory of fiber spinning is played by hydrodynamic models described by nonlinear boundary layer problems. The report examines such boundary value problems using modern methods of functional analysis, hydrodynamics and computational mathematics. An analytical method for constructing solutions in one-dimensional stationary models with boundary conditions has been developed. A numerical method is proposed whose relative errors in determining the characteristics of the heat flow for a flat plate do not depend on the Reynolds number. Theorems on the existence of solutions to boundary value problems arising in physical and technical models of polymers are proven using the Cauchy-Kovalevskaya theorem and the Schauder fixed point principle. Using general existence theorems for solutions of nonlinear equations with a vector parameter, the existence of solutions to a boundary-value problem from boundary layer theory on an unbounded half-interval is proven. The report is intended for specialists in the field of industrial mathematics, its applications in chemical technology, in the theory of materials, as well as scientists and engineers interested in nonlinear boundary value problems.
3.The energy communities based on the integration of microgrids make it possible to gain economic, environmental, technical, and social benefits. The paper aims to propose a unified multi-criteria approach covering both the planning stage and the stage of managing the energy community, in the context of various interests of its participants. Planning stage should take into account the long-term goals of the community and possible changes in external conditions. Therefore, we suggest an approach relying on the multi-attribute value theory considering the uncertainty of decision makers' preferences. Interval estimators used to express preferences enable a choice of community configuration with robust performance under changing conditions within some limits. In the operation stage, the new multi-criteria model of an intelligent "energy community operator" is proposed. It is based on bi-level programming and reinforcement learning, implementing the structure of a fair local market for sustainable development of the community. To optimize the operation of individual microgrids within the community, the multi-objective Monte-Carlo Tree Search (MCTS) algorithm is used, which helps to improve the convergence in the Stackelberg game. The multi-criteria version of the MCTS algorithm allows implementing an adaptive local automation model to solve a multi-objective lower-level problem: minimize operating costs, risk of power shortage, and CO2 emissions; smooth load peaks, and optimize power exchange between microgrids. At the top level, a management strategy that will be beneficial to all members of the community is chosen to guarantee their long-term aggregation. The effectiveness of the proposed approach is demonstrated by the example of an energy community created for three remote villages located on the coast of the Baikal lake. The natural and climatic conditions of the area allow the efficient use of wind, solar, and biomass resources. Building the community involves the consideration of three scenarios, in which priority is given to economic efficiency, environmental efficiency, or balanced development。
4.This lecture provides soft introduction to the conteporary theory of linear and nonlinear Volterra (evolutionary) integral equations of the first kind: essential tool for hereditary dynamic systems modeling. In our approach we focus on the case of piecewise continuous (jump) Volterra kernels. The existence and uniquness theory of the local global solutions will be presented. The robust numerical methods will discussed for both cases: integer and fractional order. The application of the costructed theory to energy balance and strorage control will be discussed.
5.The datasets representing the different seasons including the photovoltaic data and meteorological data were employed. The neural networks fusion method is proposed using LSTM, RNN and Dense archetectures. Three kinds of neural networks to train and predict each season , the final prediction model is formed by weighting are combined , and the weights are optimized using the Nelder-Mead method. The model is tested by randomly selecting data from other years and adding artificial disturbances. The test results prove that the weighted model has strong stability and generalization ability, and the prediction error is smaller than that of the single model.
报告人简介：
1.Denis Sidorov is an Leading Researcher of Operations Research Laboratory of Energy Systems Institute of Russian Academy of Sciences. He has completed his PhD in 1999 and Habilitation (DSc) in 2014. He is a Distinguished Guest Professor of Hunan University, Changsha, China and Visiting Research Professor of QUB, UK. He has held Vision Engineer Lead position at ASTI Holdings Pte Ltd, Singapore (2005-2008), Research Fellow ay Trinity College Dublin, Ireland (2001-2002) and CNRS, Compiegne, France (2003-2004). He is IEEE Senior Research Fellow, Expert of the Russian Science Foundation and the Russian Foundation for Basic Research. Reviewer of Mathematical Reviews and Zentralblatt fur Mathematik. Dr Sidorov is IEEE Chapter Chair of IEEE PES Russia. He serves as Member of the Editorial Boards of the journals: Renewable and Sustainable Energy Reviews (Elsevier, Q1); International Journal of Artificial Intelligence (Scopus Q2). He gained Publons Award as one of the top 1 per cent of peer reviewers in Engineering in 2018 and Publons Award "Sentinels of Science" in Mathematics in 2016. His main research interests include machine learning, power engineering, image processing, numerical methods and integral equations. He is Professor of Russian Academy of Sciences.
2. Aliona Dreglea was born in Cantemir, Moldova, in 1977. She received the M.Sc. degree in mathematics from the University of Oradea, Oradea, Romania, in 2000, and the Ph.D. degree in fluid mechanics from Irkutsk State University, Irkutsk, Russia, in 2013.,She gained her postgraduate (M.Phil.) research experience in fluid dynamics in Technological University Dublin, Dublin, Ireland, supported by Klüber Librication in 2002–2006. She is currently a Professor with the Baikal School of BRICS, Irkutsk National Research Technical University, Irkutsk, and a Research Fellow with Melentiev Energy Systems Institute, Siberian Branch of the Russian Academy of Sciences, Irkutsk. Her research interests include partial differential equations, machine learning, computational fluid dynamics, wind energy, and ml applications in medical studies.
3. Liu Song received a bachelor's degree in engineering from Harbin University of Technology in 2020. He is studying for a master's degree at Irkutsk State Polytechnic University and as a researcher in Computer Vision Laboratory at Irkutsk State Polytechnic University. His main research directions include: the application of machine learning algorithms in the fields of power systems and time series prediction, as well as the application of machine learning in image segmentation, recognition and other fields. Participated in the smart microgrid research project supported by the Russian Science Foundation.
4. Dr Samad Noeiaghdam was born in 1983 in Ardabil, Iran. He got his PhD in 2018 from the Islamic Azad University. Since 2019 he serves as Associate professor of Irkutsk National Research Technical University, Irkutsk, Russian Federation. His research interests include Applied Mathematics with application in energy, Numerical Analysis, Integral Equations, DAE, ODEs and PDEs and various Ill-posed Problems in engineering. He serves as Guest Editor of Symmetry (Q1 MDPI), Editorial board member of the Mathematical Modelling of Engineering Problems Journal.
