This paper proposes a data-based web platform to maximize the profit of a player participating in the Italian spot electricity markets. To ensure the scalability and portability of the platform, the micro-services approach are chosen with four stand-alone services communicating among one another. In the platform, a web crawler module retrieves data from the official and open sources, and a machine learning-based module is created to forecast the bidding strategies of all the market players for the next day across different spot markets. Based on the forecasted strategies, simplified market-clearing mechanisms are adopted to obtain the initial market information. Then by focusing on a single player, his/her strategy is optimized to maximize the daily profit over multiple spot markets with different payment schemes. The client can access the system through a simple, user-friendly web application. The simulation of the platform shows that around 92% of the market players can increase their profits. An example of a randomly selected player shows that the platform suggests a strategy leading to a 127.6% increase in its profit compared with its real bids in the markets.

A Data-based Platform for Supporting Profit-driven Strategy in the Italian Spot Electricity Markets / Huang, T.; Gioacchini, L.; Guaiana, F.; Valente, B.; Pio Domiziani, G.. - (2020), pp. 1-6. (Intervento presentato al convegno 55th International Universities Power Engineering Conference, UPEC 2020 tenutosi a ita nel 2020) [10.1109/UPEC49904.2020.9209787].

A Data-based Platform for Supporting Profit-driven Strategy in the Italian Spot Electricity Markets

Huang T.;Gioacchini L.;Guaiana F.;Valente B.;
2020

Abstract

This paper proposes a data-based web platform to maximize the profit of a player participating in the Italian spot electricity markets. To ensure the scalability and portability of the platform, the micro-services approach are chosen with four stand-alone services communicating among one another. In the platform, a web crawler module retrieves data from the official and open sources, and a machine learning-based module is created to forecast the bidding strategies of all the market players for the next day across different spot markets. Based on the forecasted strategies, simplified market-clearing mechanisms are adopted to obtain the initial market information. Then by focusing on a single player, his/her strategy is optimized to maximize the daily profit over multiple spot markets with different payment schemes. The client can access the system through a simple, user-friendly web application. The simulation of the platform shows that around 92% of the market players can increase their profits. An example of a randomly selected player shows that the platform suggests a strategy leading to a 127.6% increase in its profit compared with its real bids in the markets.
2020
978-1-7281-1078-3
File in questo prodotto:
File Dimensione Formato  
09209787.pdf

non disponibili

Tipologia: 2a Post-print versione editoriale / Version of Record
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 481.45 kB
Formato Adobe PDF
481.45 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2859299