Browsing by Author "Andrade, Rui Filipe Azevedo"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- A Multi-Agent System Simulation Model for Trusted Local Energy MarketsPublication . Andrade, Rui Filipe Azevedo; Pereira, Isabel Cecília Correia da Silva Praça GomesThe energy market and electric grid play a major role in everyday life. Most areas in modern society, such as: communication, health, transportation, the financial system and many others; require electrical energy to operate properly. Traditionally energy grids operate in a centralized manner. Consumers are connected to centralized utilities in the grid and energy flows from producers to Consumers. However, the rising in popularity in Renewable Energy Sources (RES) such as photovoltaic panels installed in households, small commerce and small industry wide spread the use of distributed energy generation, which the main energy grid was not designed to support. One of the possible solutions for this problem is the creation of a Local Energy Market (LeM). A LeM is a market that operates in a small physical area such as a neighborhood. Traditional consumers can become active market participants under a LeM. That is possible because the LeM is structured in such a way as to enable small-scale negotiations and energy exchanges between participants, who traditionally would only be final consumers. The LeM is capable of dealing with distributed energy generation from RES because negotiations and distribution happen at a local level, thus reducing problems with the main grid. Furthermore, the participation in the local market can reduce energy costs or even create profits for consumers, while contributing to easy the management of the grid and associated technical losses. This work explores the concept of LeM and is focused on two main objectives: designing and developing a system that allows the simulation of LeM, and designing and developing a mechanism that allows trusted negotiations in this market. To accomplish these objectives a Multi-Agent System (MAS) architecture is proposed to model and allow the simulation of LeM. Furthermore to support the market it is also proposed a trust model used to evaluate the behavior of participants and detecting faulty or malicious activities. The developed MAS models a LeM based on a Smart Grid, that is an energy grid with a cyber-physical system with smart meters and communications mechanisms. The MAS was developed with agents to model sensors, market participants and a Market Interaction Manager (MIM) agent that is responsible for managing the negotiations and for applying trust mechanisms. The trust mechanism was designed to attribute a dynamic trust value to each participant, which is reviewed during the all negotiation period. This evaluation of the participant’s trust is based on the analysis of historical data, contextual data, such as weather conditions, and by using forecasting methods to predict the participant expected behavior, allowing to penalize the ones that are exhibiting a questionable behavior in the market. A case study simulation was made with the objective of understanding how the proposed trust mechanism performed, and how the use of different forecasting methods can interfere with it. The results obtained allowed us to conclude that the trust methodology is able to update the trust of each participant, during the negotiation period, and when paired with a well performing forecasting mechanism it is able to achieve a trusted evaluation of the participants behavior. Taking into consideration these results we believe that the proposed trust methodology is capable of providing a valuable trust assessment when used by the MIM agent. This Master Thesis is developed within the scope of a project called Secure interactions and trusted Participation in local Electricity Trading (SPET), a FCT-SAICT2017 funded Research & Development project. SPET project envisions the development of a MAS that is designed to model and simulate the operations of a LeM, taking a focus on security and market trust necessary in this negotiation environment.