Development of a methodology for vehicle systems using artificial neural networks based on experimental tests

• Grant number: 24/08117-9
• Support Opportunities: Scholarships abroad - Research Internship - Doctorate
• Start date: November 04, 2024
• End date: May 18, 2025
• Field of knowledge: Engineering - Mechanical Engineering
• Principal Investigator: Ludmila Corrêa de Alkmin e Silva
• Grantee: Matheus Henrique Rodrigues Miranda
• Supervisor: Tiago Alexandre Abranches Teixeira Lopes Farias
• Host Institution: Faculdade de Engenharia Mecânica (FEM). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Institution abroad: Universidade de Lisboa, Portugal
• Associated to the scholarship: 21/14026-8 - Optimization of energy storage and the power management using artificial neural networks, BP.DR

Abstract

In order to boost the advancement of mobility technologies, it is essential to understand the dynamic behavior of vehicles, as well as their systems and components. Therefore, the main objective of the internship abroad is to carry out experimental tests to collect data from vehicles in real operating conditions, to compare and if possible validate the battery model already developed, and to apply machine learning techniques, specifically the artificial neural network method, to develop a representative model of pollutant emissions. This proposal directly complements ongoing research in the country, as it will allow us to use and validate the battery model developed, and improve the modeling of the internal combustion engine, which is fundamental to the development of the plug-in hybrid electric vehicle model. To achieve this goal, it is necessary to establish a sequence of experimental tests and analyze the vehicle's behavior over specific driving cycles. This approach will enable the creation of a robust and representative database, which is essential for training the artificial neural networks. As a result, power management systems will have more accurate and robust responses, based on realistic models. This work aims to contribute to the development of advanced power management strategies for vehicles by proposing the adoption of a model capable of realistically representing the emissions profile, taking into account variables such as vehicle operation, climatic conditions, and the type of fuel used.