Loading...
Publication
Algoritmo de heurística para o carregamento de veículos elétricos em ambiente residencial
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 13.57 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Nesta dissertação é proposto e validado um algoritmo heurístico para o carregamento
de veículos elétricos (VE) em ambiente residencial, integrando produção fotovoltaica
(PV) e sistema de armazenamento (ESS). A motivação surge da necessidade de explorar
metodologias de maximização do autoconsumo de energia solar para o carregamento
de VE, onde se procura minimizar a dependência da rede pública, e em
simultâneo, assegurar o cumprimento dos requisitos de autonomia definidos pelos
utilizadores. O algoritmo de controlo foi desenvolvido e validado experimentalmente
no laboratório x-Energy do INESC TEC. O algoritmo de controlo, implementado
em Python sobre uma Raspberry Pi, recolhe ciclicamente medições de potência do
PV, estado de carga do ESS e parâmetros do VE (SOC inicial e hora prevista de
saída), definidos via dashboard em Node-RED. Com base na energia necessária e
na janela temporal disponível, o algoritmo aplica uma hierarquia de decisão que
prioriza o uso de energia solar, seguindo-se a utilização do ESS quando necessário
e, em último recurso, recorrer à rede pública. As ordens de potência são transmitidas
ao carregador de VE através de uma API. A monitorização em tempo real é
assegurada por dashboards desenvolvidos em Node-RED. A solução foi avaliada em
múltiplos cenários de operação, desde carregamento convencional até condições de
baixa produção no sistema de PV.
In this dissertation, a heuristic algorithm for charging electric vehicles (EVs) in a residential environment is proposed and validated, integrating photovoltaic (PV) production and storage system (ESS). The motivation arises from the need to explore methodologies for maximizing solar energy self-consumption for EV charging, seeking to minimize dependence on the public grid, and at the same time, ensure compliance with the autonomy requirements defined by the users. The control algorithm was developed and experimentally validated at the x-Energy laboratory of INESC TEC. The control algorithm, implemented in Python on a Raspberry Pi, cyclically collects measurements of PV power, ESS state of charge and EV parameters (initial SOC and estimated departure time), defined as dashboard in Node-RED. Based on the required energy and the available time window, the algorithm applies a decision hierarchy that prioritizes the use of solar energy, followed by the use of the ESS when necessary and, as a last resort, resorting to the public grid. Power orders are transmitted to the EV charger via an API. Real-time monitoring is ensured by dashboards developed in Node-RED. The solution has been evaluated in multiple operating scenarios, from conventional charging to low-output conditions in the PV system.conditions in the PV system.
In this dissertation, a heuristic algorithm for charging electric vehicles (EVs) in a residential environment is proposed and validated, integrating photovoltaic (PV) production and storage system (ESS). The motivation arises from the need to explore methodologies for maximizing solar energy self-consumption for EV charging, seeking to minimize dependence on the public grid, and at the same time, ensure compliance with the autonomy requirements defined by the users. The control algorithm was developed and experimentally validated at the x-Energy laboratory of INESC TEC. The control algorithm, implemented in Python on a Raspberry Pi, cyclically collects measurements of PV power, ESS state of charge and EV parameters (initial SOC and estimated departure time), defined as dashboard in Node-RED. Based on the required energy and the available time window, the algorithm applies a decision hierarchy that prioritizes the use of solar energy, followed by the use of the ESS when necessary and, as a last resort, resorting to the public grid. Power orders are transmitted to the EV charger via an API. Real-time monitoring is ensured by dashboards developed in Node-RED. The solution has been evaluated in multiple operating scenarios, from conventional charging to low-output conditions in the PV system.conditions in the PV system.
Description
Keywords
Veículos Elétricos Autoconsumo PV Algoritmo heurístico Armazenamento de energia Raspberry Pi Node-RED Veículos elétricos Autoconsumo PV ALgoritmo heurístico Armazenamento de energia
Pedagogical Context
Citation
Publisher
CC License
Without CC licence