Considering the increasing demand for transport systems that can efficiently utilize energy, minimize travel time and enhance the driving experience, this thesis focuses on the development of energy-efficient speed planning for connected electric vehicles (EVs) using Model Predictive Control (MPC) based optimization and Vehicle-to-Infrastructure (V2I) communication. By integrating real time traffic light data, V2I systems enable EVs to predict when the traffic light will turn red so as to slow down and reduce stops and energy consumption. The methodological approach involves a simulation framework that combines traffic light configurations for urban and suburban road scenarios with a Signal Phase and Timing (SPaT) model. This model facilitates the evaluation of three Tesla Model Y configurations (RWD, Long Range, and Performance) under non-connected EV and connected EV with rule-based or MPC-based speed planning strategies, with prediction horizons set at varying levels (N=5, N=10, N=15, and N=20). The rule-based strategy serves as a baseline, while the MPC-based approach optimizes speed, State of Charge (SOC), and passenger comfort by balancing efficiency and computational cost. Results indicate that the MPC-based connected EV approach substantially reduces energy consumption up to 24.4% in Urban Road and 24.32% in Suburban Road compared to non-connected EVs. Higher prediction horizons further enhance energy savings, though they increase computational demand, presenting a trade-off that must be considered for real-time applications. The analysis of urban and suburban conditions demonstrates that the higher frequency of stops is beneficial for energy conservation in urban environments. In all the tests, the connected EVs with MPC optimized exhibit smoother driving trajectories, which is a clear indication of the benefits of V2I systems in enhancing the performance and energy efficiency of EVs. All simulations have been carried out via MATLAB.
Considerando la crescente domanda di sistemi di trasporto che possano utilizzare l’energia in modo efficiente, ridurre i tempi di viaggio e migliorare l’esperienza di guida, questa tesi si concentra sullo sviluppo di una pianificazione della velocità a basso consumo energetico per veicoli elettrici (EV) connessi, utilizzando un’ottimizzazione basata sul Controllo Predittivo Modello (MPC) e la comunicazione Veicolo-Infrastruttura (V2I). Integrando i dati in tempo reale dei semafori, i sistemi V2I consentono ai veicoli elettrici di prevedere quando il semaforo diventerà rosso, in modo da rallentare e ridurre le fermate e il consumo di energia. L’approccio metodologico prevede un quadro di simulazione che combina configurazioni semaforiche per scenari stradali urbani e suburbani con un modello di Fase e Temporizzazione del Segnale (SPaT). Questo modello facilita la valutazione di tre configurazioni di Tesla Model Y (RWD, Long Range e Performance) in base a strategie di pianificazione della velocità basate su regole e su MPC, con orizzonti di previsione impostati a diversi livelli (N=5, N=10, N=15 e N=20). La strategia basata su regole funge da riferimento, mentre l’approccio basato su MPC ottimizza la velocità, il consumo dello Stato di Carica (SOC) e il comfort dei passeggeri, bilanciando efficienza e costo computazionale. I risultati indicano che l’approccio basato su MPC riduce sostanzialmente il consumo di energia fino al 24,4% nelle strade urbane e al 24,32% nelle strade suburbane rispetto ai veicoli elettrici non connessi. Orizzonti di previsione più ampi migliorano ulteriormente il risparmio energetico, anche se aumentano la domanda computazionale, presentando un compromesso da considerare per applicazioni in tempo reale. L’analisi delle condizioni urbane e suburbane dimostra che la maggiore frequenza delle fermate è vantaggiosa per la conservazione dell’energia negli ambienti urbani. In tutti i test, i veicoli elettrici connessi ottimizzati con MPC mostrano traiettorie di guida più fluide, il che è un chiaro segno dei benefici dei sistemi V2I nel migliorare le prestazioni e l’efficienza energetica dei veicoli elettrici. Tutte le simulazioni sono state eseguite tramite MATLAB.
Energy-efficient based model prediction for speed planning of connected electric vehicles
Patlar, Alper
2023/2024
Abstract
Considering the increasing demand for transport systems that can efficiently utilize energy, minimize travel time and enhance the driving experience, this thesis focuses on the development of energy-efficient speed planning for connected electric vehicles (EVs) using Model Predictive Control (MPC) based optimization and Vehicle-to-Infrastructure (V2I) communication. By integrating real time traffic light data, V2I systems enable EVs to predict when the traffic light will turn red so as to slow down and reduce stops and energy consumption. The methodological approach involves a simulation framework that combines traffic light configurations for urban and suburban road scenarios with a Signal Phase and Timing (SPaT) model. This model facilitates the evaluation of three Tesla Model Y configurations (RWD, Long Range, and Performance) under non-connected EV and connected EV with rule-based or MPC-based speed planning strategies, with prediction horizons set at varying levels (N=5, N=10, N=15, and N=20). The rule-based strategy serves as a baseline, while the MPC-based approach optimizes speed, State of Charge (SOC), and passenger comfort by balancing efficiency and computational cost. Results indicate that the MPC-based connected EV approach substantially reduces energy consumption up to 24.4% in Urban Road and 24.32% in Suburban Road compared to non-connected EVs. Higher prediction horizons further enhance energy savings, though they increase computational demand, presenting a trade-off that must be considered for real-time applications. The analysis of urban and suburban conditions demonstrates that the higher frequency of stops is beneficial for energy conservation in urban environments. In all the tests, the connected EVs with MPC optimized exhibit smoother driving trajectories, which is a clear indication of the benefits of V2I systems in enhancing the performance and energy efficiency of EVs. All simulations have been carried out via MATLAB.File | Dimensione | Formato | |
---|---|---|---|
2024_12_Patlar_Thesis_01.pdf
solo utenti autorizzati a partire dal 11/11/2027
Descrizione: Thesis File
Dimensione
50.05 MB
Formato
Adobe PDF
|
50.05 MB | Adobe PDF | Visualizza/Apri |
2024_12_Patlar__Executive Summary_02.pdf
solo utenti autorizzati a partire dal 11/11/2027
Descrizione: Executive Summary File
Dimensione
10.06 MB
Formato
Adobe PDF
|
10.06 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/229907