An Introduction to MultiAgent Systems.

The increasing interest in integrating intermittent renewable energy sources into microgrids presents major challenges from the viewpoints of reliable operation and control. In this paper, the major issues and challenges in microgrid control are discussed, and a review of state-of-the-art control strategies and trends is presented; a general overview of the main control principles (e.g., droop control, model predictive control, multi-agent systems) is also included. The paper classifies microgrid control strategies into three levels: primary, secondary, and tertiary, where primary and secondary levels are associated with the operation of the microgrid itself, and tertiary level pertains to the coordinated operation of the microgrid and the host grid. Each control level is discussed in detail in view of the relevant existing technical literature.

Trends in Microgrid Control

Demand side management (DSM) is one of the important functions in a smart grid that allows customers to make informed decisions regarding their energy consumption, and helps the energy providers reduce the peak load demand and reshape the load profile. This results in increased sustainability of the smart grid, as well as reduced overall operational cost and carbon emission levels. Most of the existing demand side management strategies used in traditional energy management systems employ system specific techniques and algorithms. In addition, the existing strategies handle only a limited number of controllable loads of limited types. This paper presents a demand side management strategy based on load shifting technique for demand side management of future smart grids with a large number of devices of several types. The day-ahead load shifting technique proposed in this paper is mathematically formulated as a minimization problem. A heuristic-based Evolutionary Algorithm (EA) that easily adapts heuristics in the problem was developed for solving this minimization problem. Simulations were carried out on a smart grid which contains a variety of loads in three service areas, one with residential customers, another with commercial customers, and the third one with industrial customers. The simulation results show that the proposed demand side management strategy achieves substantial savings, while reducing the peak load demand of the smart grid.

Demand Side Management in Smart Grid Using Heuristic Optimization

This paper proposes a hybrid ac/dc micro grid to reduce the processes of multiple dc-ac-dc or ac-dc-ac conversions in an individual ac or dc grid. The hybrid grid consists of both ac and dc networks connected together by multi-bidirectional converters. AC sources and loads are connected to the ac network whereas dc sources and loads are tied to the dc network. Energy storage systems can be connected to dc or ac links. The proposed hybrid grid can operate in a grid-tied or autonomous mode. The coordination control algorithms are proposed for smooth power transfer between ac and dc links and for stable system operation under various generation and load conditions. Uncertainty and intermittent characteristics of wind speed, solar irradiation level, ambient temperature, and load are also considered in system control and operation. A small hybrid grid has been modeled and simulated using the Simulink in the MATLAB. The simulation results show that the system can maintain stable operation under the proposed coordination control schemes when the grid is switched from one operating condition to another.

A Hybrid AC/DC Microgrid and Its Coordination Control

In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications, e.g., for medical purposes and in vehicular networks. Traditional cloud-based Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL.

/pdf/federated-learning-in-mobile-edge-networks-a-comprehensive-5dgelc5m6h.pdf

Federated Learning in Mobile Edge Networks: A Comprehensive Survey

Multi-agent system for energy resource scheduling of integrated microgrids in a distributed system

This paper presents a multiagent system (MAS) for real-time operation of a microgrid. The proposed operational strategy is mainly focused on generation scheduling and demand side management. In generation scheduling, schedule coordinator agent executes a two-stage scheduling: day-ahead and real-time scheduling. The day-ahead scheduling finds out hourly power settings of distributed energy resources (DERs) from a day-ahead energy market. The real-time scheduling updates the power settings of the distributed energy resources by considering the results of the day-ahead scheduling and feedback from real-time operation of the microgrid in real-time digital simulator (RTDS). A demand side management agent performs load shifting before the day-ahead scheduling, and does load curtailing in real-time whenever it is necessary and possible. The distributed multiagent model proposed in this paper provides a common communication interface for all components of the microgrid to interact with one another for autonomous intelligent control actions. Furthermore, the multiagent system maximizes the power production of local distributed generators, minimizes the operational cost of the microgrid, and optimizes the power exchange between the main power grid and the microgrid subject to system constraints and constraints of distributed energy resources. Outcome of simulation studies demonstrates the effectiveness of the proposed multiagent approach for real-time operation of a microgrid.

Multiagent System for Real-Time Operation of a Microgrid in Real-Time Digital Simulator

Multi-agent system (MAS) is one of the most exciting and the fastest growing domains in agent oriented technology which deals with modeling of autonomous decision making entities. This paper presents an application of MAS for distributed energy resource (DER) management in a MicroGrid. MicroGrid can be defined as low voltage distributed power networks comprising various distributed generators (DG), storage and controllable loads, which can be operated as interconnected or as islands from the main power grid. By representing each element in MicroGrid as an autonomous intelligent agent, multi agent modeling of a MicroGrid is designed and implemented. JADE framework is proposed for the modeling and reliability of the MicroGrid is confirmed with PowerWorld Simulator. Further, the FIPA contract net coordination between the agents is demonstrated through software simulation. As a result, this paper provides a MicroGrid modeling which has the necessary communication and coordination structure to create a scalable system. The optimized MicroGrid management and operations can be developed on it in future.

Multi-agent coordination for DER in MicroGrid

Microgrids are low voltage intelligent distribution networks comprising various distributed generators, storage devices and controllable loads which can be operated as interconnected or as islanded system. The optimal generation scheduling is one of the important functions for the Microgrid operation. This paper describes a three-step efficient method for the optimal generation scheduling of a Microgrid in island operation. The first step of the method is to set up an initial feasible solution for thermal unit commitment and the next step is to solve the thermal unit commitment problem. The final step is to optimize the renewable-thermal dispatch based on thermal unit commitment results. Solving the thermal unit commitment problem has more opportunity to minimize the operating cost. Therefore, few algorithms such as Lagrangian relaxation, genetic algorithm and a hybrid algorithm of Lagrangian relaxation and genetic algorithm have been used to find the least operating cost. Microgrid which is considered in the case study, consists of a PV system, a wind plant, 10 thermal units and a battery bank.

Thillainathan Logenthiran

Papers

Demand Side Management in Smart Grid Using Heuristic Optimization

Multi-agent system for energy resource scheduling of integrated microgrids in a distributed system

Multiagent System for Real-Time Operation of a Microgrid in Real-Time Digital Simulator

Multi-agent coordination for DER in MicroGrid

Short term generation scheduling of a Microgrid