Modeling and Control Methods of Cyber-Physical Systems in Multi-Energy Microgrids

International Project of ESI SB RAS (Russia, PI), CSU (China) and IIT Roorkee (India).

The concept of intelligent multi-energy microgrids is based on the development and the introduction of cyberphysical systems based on effective mathematical models, telecommunication and information management technologies that allow efficiently generate, store, distribute and deliver energy to the consumer from sources of both conventional and renewable generation. This project is aimed at systematising and obtaining new scientific knowledge in the field of computational modeling and control of cyber-physical systems in multi-energy microgrids, including small unit power generation units based on thermo chemical conversion of low-grade fuels and renewable energy sources based on wind generators and solar modules. Integral dynamic models and efficient predictive models and methods based on deep learning recursive networks will make it possible to implement the concept of a digital twin for controlling a multi-element hybrid energy storage system based on energy accumulators and super capacitors. Multi-energy microgrid plays an important role in improving the comprehensive utilisation rate of multi-energy on the user side. With the wide interconnection of source-storage-load equipment at the multi-energy microgrid level through wired/wireless communication networks, the multi-energy microgrid has gradually evolved into a highly coupled cyber-physical system, and the traditional operation and control methods are difficult to apply. Russian Team (project coordinator) will focus on theoretical studies for novel mathematical models development. India team will contribute with its vast knowledge in converters modeling and development and China team will contribute on automation including situation awareness. Based on the research foundation of China, Russia and India in the field of multi-energy microgrid, this project is carried out from four aspects: coupling interaction mechanism of information flow/energy flow, mathematical modeling, situation perception and cooperative control in multi-energy microgrid. It will overcome three scientific difficulties: dynamic modeling of cyber-physical system of multi-energy microgrid, situation perception of time series data driven multi-energy microgrid and resource-storage equipment, and distributed cooperative control of flexible controllable resources of multi-energy microgrid. And the joint simulation model of cyber-physical system in multi-energy microgrid and the typical demonstration platform of optical-storage-charge multi-energy microgrid are established. Through the research of this project, the original innovations such as the coupling interaction mechanism and the fusion interaction architecture of multiple information flow/energy flow, situation perception method of the source-storage stability, the distributed energy storage cooperative control, and the frequency-voltage flexible control will be achieved. These will enhance the situation perception ability, multi-energy collaborative level, and dynamic stability of the multi-energy microgrid, realize self-balance and safe operation of the multi-energy microgrid, and improve the economic benefits of the cyber-physical system in multi-energy microgrids and multi-energy consumption efficiency.

Results for 2022 – the first stage of the project

During 2022, the first phase of work on “Modeling and methods of control of cyber-physical systems in multi-energy microgrids (MEMG)” was performed. The following results were obtained:

1. The analysis of scientific and technical literature, normative and technical documentation was carried out, which showed that the general processes of decentralization, decarbonization and digitalization, as well as the formation of new energy practices, the introduction of new technologies, the penetration of renewable energy sources (RES) and power electronics stimulate the increasing development of microgrids and their transformation into MEMG as autonomous and independent energy structures. At the same time, the stochastic nature of RES generation, which is actively used in MEMG, adversely affects their stable operation and complicates the prediction and optimization of their modes.
2. The patent research on the project topic was carried out. Checking of patents showed that there are no patents with broad legal protection in the area of interest, there are no patents preventing free development in the considered area.
3. Justification of the chosen research direction is given. It is shown that MEMG control is a complex and urgent task, when solving which it is necessary to take into account not only energy but also information resources. The conducted analysis of the research results presented in scientific publications has shown that a ready-made solution that fully satisfies the solution of this problem has not been developed to date. This problem can be formulated as an optimal control problem. As initial data for optimization, the forecast of changes in the mode parameters for a given time is used. Distributed optimization allows effectively solving the problem of optimal control in the case when the creation of a single control center, centralized data collection and formation of a single MEMG model is not rational. At the same time, the existing methods of distributed optimal control require development. A promising approach to MEMG control is the multi-agent approach. Multi-agent technologies appear to be an effective approach not only for distributed control of MEMG modes, but also for building its reliable information network at the level of medium and low voltage systems.
4. modeling of connections between energy flow and information flow of multi-energy network is carried out. The description of the proposed solution of multi-agent control of MEMS is given. The developed MEMS control model realizes a dynamic multi-agent control system based on universal node controllers. Each of the node controllers implements a universal adaptive algorithm, the joint work of which realizes the optimal control of the whole MEMS. A dynamic communication and energy structure of the MEMS is assumed, which takes into account information flows during multi-agent control.
5. Methods and models of multi-criteria optimization of the structure of multi-energy microgrids according to the criteria of economic efficiency and minimization of environmental impact are described and developed. Multicriteria target functions formed in accordance with the multicriteria value theory are used for two-level optimization. Their use allows mathematically justified formation and use of the decision maker (DM) preference model for two-level optimization. The developed two-level hierarchy of objectives and criteria allows optimization within a single multi-criteria model. In this case, the minimization of information requests from the DM is provided.
6. An experimental sample of a hybrid power plant (as part of MEMG) was designed and installed. Its test tests were carried out. The test tests showed the performance of the hybrid power plant in both normal and emergency operating modes.

China team member postgrad Mr Peng Yuheng’s talk on 27th of May 2023

Results for 2023 – second stage of the project

During 2023, the second stage of work on the topic “Modeling and methods of control of cyber-physical systems in MEMG” was performed. The following results were obtained in the reporting period:

1. An effective method of modeling the cyber-physical system of a multi-energy network has been developed, which takes into account active elements in MEMG involved in the formation of a solution with regard to their own goals in the process of organizing an efficient power supply. The method allows to decompose a general complex control problem into several simpler ones with the organization of information communication between the elements;
2. The theory of mathematical modeling of connections between energy flow and information flow of a multi-energy microgrid is developed, which is reduced to the theory of optimal control, since reliability and stability constraints are taken into account as constraints on the parameters of stationary modes during optimization;
3. A mathematical model of distributed energy generator is developed;
4. A method for estimating the state of charge of energy accumulators in a multi-energy microgrid with time series control of a multi-energy microgrid is developed.
5. New forecasting methods based on the use of ensemble method and deep machine learning methods based on recurrent neural network algorithms have been developed. The method were validated on the real dataset from the wind farms.
6. Methods and models of multi-criteria optimization of the structure of multi-energy microgrids by the criteria of reliability and quality of generated electricity have been developed. A modified MAVT method is developed, which allows ranking the variants of the structure of multi-energy microgrids under uncertainty using quantitative indicators. The quality of the obtained solutions using the proposed approach is higher in comparison with the traditional single-criteria approach, since the single-criteria approach ensures that only one criterion is taken into account, and the estimates for the other criteria may be unsatisfactory. The comparative analysis showed that changing the scaling coefficients of the criteria leads to a corresponding change in the resulting estimates for the criteria. This ensures high quality of the decisions made, as the necessary criteria are taken into account in proportion to their importance for the decision maker in a particular situation.
7. Experimental studies were conducted on the developed hybrid power plant, which showed that in this configuration, autonomous operation was maintained and satisfied the demand of consumers. Due to photovoltaic energy it was possible to maintain autonomous operation during the daytime, and the battery pack accumulated excess electricity for the subsequent operation of the microgrid at night. If necessary, the microgrid was automatically connected to an external network (generator). Based on the obtained data, it can be concluded that the developed hybrid microgrid functions and allows to conduct research in the field of improving the reliability and efficiency of hybrid energy systems using renewable energy sources and can serve as a basis for further research in the field of hybrid energy systems. Thus, the data on the operation of storage units were used to develop a method for assessing the state of charge of energy accumulators in a multi-energy microgrid with time series control.

Summary of the results for 2024 – third stage of the project

Models and methods for controlling multi-energy microgrids with biofuel gas generators are developed. A virtual model of inertia control considering the limitation on the storage charge level is proposed. A method for estimating the charge level of energy storage devices based on Kolmogorov-Arnold neural networks is designed. A multilevel strategy for controlling the response of a multi-energy microgrid based on network data and a distributed control strategy for energy storage devices are proposed. Methods and models for multi-criteria selection of the association structure of multi-energy microgrids are proposed. Experimental studies have been carried out and the optimal modes of operation of the hybrid power plant have been revealed. The technical and economic evaluation of the research results has been carried out. The research showed that intelligent management of microgrids with small-scale renewable generation can reduce the carbon footprint by 40% and cut costs for consumers by 30-50%, contributing to improved energy efficiency and sustainable development of regions in Russia and BRICS countries with high renewable energy potential.

Multi-energy microgrid monitoring system dataset: https://isem.irk.ru/microgrid/

Selected publications for 2024

1. V. Badenko, A. Kozlov. Primary analysis of the operation of a laboratory-scale hybrid microgrid using a monitoring system. Information and mathematical technologies in science and management. 2024, Vol. (35), p.73-81 (in Russian).
2. V. Shakirov, I. Popov. Multi-criteria design of multi-energy system for remote area using NSGA-III and fuzzy TOPSIS. Journal of Renewable and Sustainable Energy, 2024, 16.6.
3. M. Dao, F. Liu, D. Sidorov. Kolmogorov-Arnold neural networks technique for the state of charge estimation for Li-ion batteries. Bulletin of the South Ural State University, Series: Mathematical Modelling, Programming and Computer Software, 2024, 17(4), pp. 22–31.
4. N. Tomin and L. Gurina. Optimal Management of a Multi-Energy Microgrid Community Based on Building-to-Grid Technology, 2024 International Ural Conference on Electrical Power Engineering (UralCon), Magnitogorsk, Russian Federation, 2024, pp. 220-224.
5. W. Guo, F. Liu, Y. Wang, D. Sidorov, J. Wu, Adaptive Event-Triggered Sliding Mode Load Frequency Control for Cyber-Physical Power Systems Under False Data Injection Attacks, IEEE Transactions on Industrial Informatics, 2024, 1–10.
6. A. Domyshev, D. Sidorov, F. Liu, Microgrids multi-agent control based on power distribution and communication constraints, Proc. of Chinese Control Conference (43rd IEEE Chinese Control Conference, 2024), IEEE, 2024, 5812–5816.