#	Pagina
attuale pagina	/open-h2020/projects/202193/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - ADAPT (Adaptive Decision support for Agents negotiation in electricity market and smart grid Power Transactions)

Teaser

Summary

ADAPT explores and proposes new solutions for the decision support of playersâ€™ negotiations in the complementary environments of electricity markets and smart grids. The EU priority of increasing the penetration of renewable energy sources in the power system has led to the wide acceptance of the smart grid paradigm as the solution for several emerging challenges. Another EU priority is the creation of a Pan-European Energy Market. A relevant step was taken in 2015 with the coupling of several regional European markets. Despite the significant investments in the fields of smart grids and electricity markets, the two fronts are evolving quite independently from each other, evidencing the sector difficulties in profiting from the strong relationships among them. A relevant drawback is the lack of decision support solutions to enable involved players improving their outcomes from energy transactions. The management and planning tools of players in the power and energy sector are directed to the traditional market models; while the emerging small players, such as renewable energy generators and small-sized consumers, were not able to participate directly in wholesale electricity markets, without the support of intermediaries such as retailers and aggregators. Hence, such players do not have the know-how or experience in market participation. ADAPT has developed solutions to support these players actions in a market context that includes the participation in future wholesale electricity markets and in local transactions at a smart grid level. The solutions developed by the project include a planning component, which helps market players deciding in which market opportunities to participate; and a component to assist players during the actual negotiation process.

Work performed

ADAPT has proposed a multi-agent decision support system to assist the negotiation process of smart grid and electricity market players. This decision support system integrates two main components, which refer to the planning phase (helping players deciding in which market opportunities to participate and the volumes of energy to negotiate) and to the actual negotiation phase (supporting playersâ€™ negotiation process itself, when participating in the market). The planning component consists in a portfolio optimization model that identifies the most gainful negotiation opportunities, considering the available markets, including negotiations at a smart grid level. The decision support for actual negotiations applies novel machine learning approaches to choose the most suitable negotiation strategies for each context. The contexts refers to the factors that influence the negotiation conditions, such as the expected volume of renewable generation at each time, the expected market prices, the players that are expected to be involved in the particular market negotiations, etc. The opponent players negotiation profiles are also considered as means for the system to adapt its suggested actions according to the expected behaviour of the opponents.
The ADAPT system will be used by players involved in electricity transactions, including consumers, generators and aggregators. These players are able to improve their electricity market participation by being endowed with solutions that help them in the planning and negotiation processes. The ADAPT system can be used by both the traditional players usually involved on wholesale electricity market negotiations, and also by the small consumers and generators that are just starting or about to start a more active participation in electricity markets, especially through local trading.
The project results are have been widely disseminated through the traditional channels, such as the project website and social networks (Facebook, LinkedIn, Twitter and YouTube), but also through the participation in multiple international events. It is worth highlighting the organization of 13 workshops and seminars in international venues; the delivery of 8 invited talks on the project topics; the edition of 2 special issues in scientific journals and one book; and the publication of 12 SCI indexed journal papers and 20 publications in international conferences and congresses in the fields of artificial intelligence and power and energy systems.

Final results

The scientific results of the project open new paths for research in both the fields of artificial intelligence and power and energy systems. The new machine learning, optimization and negotiation approaches proposed by ADAPT create a solid basis for further exploration of the theoretical artificial intelligence models and their application to the evolving electricity market sector, e.g. the ongoing transitions towards local energy markets, including peer-to-peer transactions.
The findings achieved during the project complemented by defining as output a potentially commercial product (ADAPT system), by including a solid exposure to the industry (through ENGIE) and by exploiting the project results appropriately, ADAPT has a direct impact on the EU power sector, the industry and ultimately the society as a whole, by contributing towards:
(i) the increase of the power sector efficiency by proposing models that optimize the involved entitiesâ€™ operation in an integrated electricity market and smart grid environment;
(ii) the EUâ€™s return of investment in smart grids by proposing innovative solutions for smart grids effective integration in the electricity market;
(iii) the sustainability of investing in renewable energy sources, by proposing solutions that enable owners to maximize their gains from the sale of power;
(iv) the decrease of electricity prices by potentiating the increase of competitiveness in the market through the introduction of a decision support system whose ultimate goal is to support players decisions in electricity market negotiations.
Additionally, the project findings have been enabled by the inter-sectoral and multidisciplinary nature of the involved partners. The industry driven expertise brought by world leaders in the energy field - ENGIE, complemented by the unique know-how of Professor Michael. Wooldridge (University of Oxford) in multi-agent systems and automated negotiation, bring together an unparalleled set of competences that has resulted in significant contributions in the fields of artificial intelligence and power systems. The partners background and expertise has also ensured an effective cross-sectoral transfer of knowledge, from academic to industrial institutions and vice-versa.