POTENTIAL OF REINFORCEMENT LEARNING IN ENERGY MANAGEMENT STRATEGY OF HYBRID ELECTRIC VEHICLE

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21220%2F23%3A00368219" target="_blank" >RIV/68407700:21220/23:00368219 - isvavai.cz</a>
Result on the web
<a href="https://sites.google.com/vutbr.cz/koka2023" target="_blank" >https://sites.google.com/vutbr.cz/koka2023</a>
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
POTENTIAL OF REINFORCEMENT LEARNING IN ENERGY MANAGEMENT STRATEGY OF HYBRID ELECTRIC VEHICLE
Original language description
Reinforcement learning is gaining popularity in automotive industry as a form of control due to relative simplicity of its development, theoretical ability to handle complex issues and capability to abstract patterns from stochastic phenomena by letting the agent interact with the environment. This paper describes implementation of reinforcement learning algorithm called Deep Deterministic Policy Gradient to infer energy management strategy for Hybridised delivery vehicle (7.5 t). Both the algorithm and the training environment are created in Python 3.9. programming language. Results of the reinforcement learning are tested on multiple drive cycles and compared to benchmarks that are set by an optimal control algorithm based on Pontryagin’s Minimum Principle.
Czech name
—
Czech description
—

Project
<a href="/en/project/TN02000054" target="_blank" >TN02000054: Božek Vehicle Engineering National Center of Competence</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year
2023
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Similar results(10)