Energy management strategy for a solar race car including meteorologic and probabilistic variable

This thesis describes the energy management strategy for racing solar cars, the racing strategy is treated as an optimal control problem with random variables and uncertain predictions. A computational model is developed for estimating the vehicle performance under specific circumstances. Two evolutionary heuristic optimization methods are implemented and tested for this case, their effectiveness, convergence and efficiency is measured and compared to exhaustive search approaches. The dependency on solar radiation is validated using the computational model with different test cases. In order to reduce the uncertainties on the solar radiation estimation, satellite images are used as inputs to image processing and machine learning techniques, their efficacy is compared. Finally, a validation case is executed and different scenarios are evaluated with the inclusion of the proposed methods, the experimental performance of a vehicle obtained using the strategy in the World Solar Challenge 2015 is exposed and compared to the predicted results from the simulation.

• Autores: Betancur Valencia, Esteban
• Fecha de publicación: 2018
• Institución: Universidad EAFIT
• Repositorio: Repositorio EAFIT
• Idioma: spa
• OAI Identifier: oai:repository.eafit.edu.co:10784/13660
• Acceso en línea: http://repository.eafit.edu.co/handle/10784/13660
• Palabra clave: Carro solar; METEOROLOGÍA; ENERGÍA SOLAR; VARIABLES ALEATORIAS; PROGRAMACIÓN HEURÍSTICA; Solar Car; Race strategy; Energy management; Heuristic optimization; Solar resource
• License: Acceso abierto