Implement of Edge Computing for Cyber-Physical System in the Smart Grids Quasi-Real Time Control Context

An increase in the number of power consumers leads to scale up a power supply grids, and the introduction of Smart Grid (SG) for connecting components and subsystems of distributed generation, increases the complexity of the interaction configuration and, accordingly, grids control. At the same time, with peak loads cases, when there is a lack of power, the probability of big-scale fluctuations increases, which turns the energy system from deterministic to stochastic. Unpredictable manifestations of big-scale fluctuations spread in the form of a cascade of failures through the grid, which lead to the loss of reliability of the entire power grid. To simplify the controllability of complex big-scale power grids and prevent cascading failures in the supply of electric power, they are divided into a big number of microgrids. To increase observability and controllability, big-scale SGs are implementing innovative intelligent products and cyber-physical systems (CPS) that connect the virtual world with the world of things. CPSs allow to carry out instant monitoring and display of the states of power system components in large geographical areas, however, due to the concentration of control in the center, it does not ensure control efficiency in critical peak load times. Thus, it is important to increase the energy efficiency of smart grids, in particular by reducing power dispersion and increasing the flexibility, safety and reliability of energy supply. Such requirements lead to rethinking the structure of both the physical and cybernetic subsystems of the CPS, with which users and suppliers of energy resources interact. To meet these requirements, it is proposed to develop the structure of the CPS and improve computation based on the processing of “instant data” with the use of edge computations in quasi-real time, in order to ensure an increase in the efficiency of a big-scale complex power networks control.

I. Introduction

The electric power production and supply is of significant importance for ensuring the wellbeing and safety of any country, and its value is crucial for the development of human activities and life quality. The development of social and economic activities of mankind requires the constant increase in the consumption of electric power. As the result of this development, the scale (increase in size) of electric power grids are growing. When there is a lack of power at times of peak loads in a big-scale grid, the probability of big-scale fluctuation increases, which turns the system from deterministic to stochastic. Unpredictable manifestations of big-scale fluctuation spread in the form of a cascade of failures through the grid, which lead to the loss of reliability of the operation of the power grid, and the occurrence of global interruptions in the power supply of big cities up to the disconnection of a big number of consumers (big-scale power outages in the USA in 2003). This state of affairs is connected with the fact that the existing system of centralized (dispatching) management of the supply of electric energy and the existing information support for decision-making cannot promptly detect trends of approaching supply problems and, accordingly, respond to them.