In the contemporary energy landscape, characterized by a global commitment to sustainability, the effective management and forecasting of energy consumption play pivotal roles in achieving environmental and economic goals. As nations strive to meet sustainable development targets, optimizing energy use becomes imperative. This paper addresses these challenges by focusing on load forecasting and energy management within the context of a Savona campus microgrid. In this thesis, the NNR algorithm based load profile prediction model was proposed. The development process involved a detailed exploration of the correlation between weather information and electricity consumption. Furthermore, the outputs of the load forecasting model, namely the predicted load profiles, were subsequently utilized in the Energy Management System (EMS) to optimally manage power flows in the campus microgrid using an Internet of things as a service (IoTaaS) with day-ahead forecasting model. The overall results of the model evaluation across all periods reveal a Mean Absolute Error (MAE) of 9.63 kW, a Coefficient of Determination (R2) of 0.79, and a Mean Absolute Percentage Error (MAPE) of 9.02%. These metrics provide a comprehensive assessment of the model's performance across various temperature conditions. The proposed load profile forecasting model was integrated into the Energy Management System (EMS) developed for Savona campus microgrid in Italy. The findings provide a valuable framework for optimizing microgrid operations, aligning with global sustainability objectives.
Energy Management System for the Campus Microgrid Using an Internet of Things as a Service (IoTaaS) with Day-ahead Forecasting
Bracco S.
2024-01-01
Abstract
In the contemporary energy landscape, characterized by a global commitment to sustainability, the effective management and forecasting of energy consumption play pivotal roles in achieving environmental and economic goals. As nations strive to meet sustainable development targets, optimizing energy use becomes imperative. This paper addresses these challenges by focusing on load forecasting and energy management within the context of a Savona campus microgrid. In this thesis, the NNR algorithm based load profile prediction model was proposed. The development process involved a detailed exploration of the correlation between weather information and electricity consumption. Furthermore, the outputs of the load forecasting model, namely the predicted load profiles, were subsequently utilized in the Energy Management System (EMS) to optimally manage power flows in the campus microgrid using an Internet of things as a service (IoTaaS) with day-ahead forecasting model. The overall results of the model evaluation across all periods reveal a Mean Absolute Error (MAE) of 9.63 kW, a Coefficient of Determination (R2) of 0.79, and a Mean Absolute Percentage Error (MAPE) of 9.02%. These metrics provide a comprehensive assessment of the model's performance across various temperature conditions. The proposed load profile forecasting model was integrated into the Energy Management System (EMS) developed for Savona campus microgrid in Italy. The findings provide a valuable framework for optimizing microgrid operations, aligning with global sustainability objectives.File | Dimensione | Formato | |
---|---|---|---|
Kaz1.pdf
accesso chiuso
Tipologia:
Documento in versione editoriale
Dimensione
807.87 kB
Formato
Adobe PDF
|
807.87 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.