研究目的
To obtain an optimal velocity profile for the AntuNekul II solar car to minimize the time spent in the 2018 Atacama Solar Race, subject to physical and regulatory constraints.
研究成果
The three metaheuristic algorithms (ILS, SA, GA) effectively optimize the velocity profile for the solar car, with SA providing the best solutions in terms of minimized race time, ILS having the shortest execution times, and GA performing worst. The methods are validated against real-world data, showing reliability for solar racing applications, and can be adapted to other instances with similar constraints.
研究不足
The study assumes similar solar irradiance between Chile and Australia for validation, which may not account for all geographical variations. The algorithms' performance depends on parameter calibration, and the feasibility of solutions is checked post-generation, which could be optimized. The model relies on specific parameters for the vehicle and environment that may not generalize to other conditions.
1：Experimental Design and Method Selection:
The study uses three metaheuristic algorithms (Simulated Annealing, Iterated Local Search, Genetic Algorithm) programmed in C++11 to optimize the velocity profile. The optimization aims to minimize the total race time based on a dynamic model of the hybrid electric vehicle and solar irradiance estimation.
2：Sample Selection and Data Sources:
Three test instances are used: two from the World Solar Challenge (26 and 52 segments) for calibration and validation, and one for the Atacama Solar Race 2018 (46 segments) created using Google Maps and Google Earth.
3：List of Experimental Equipment and Materials:
The solar car AntuNekul II with components including MPPT, photovoltaic cells, electric motor, controller, and battery. Software tools include Code::Blocks IDE for programming.
4：Experimental Procedures and Operational Workflow:
Algorithms generate initial feasible velocity profiles, iteratively improve them using methods like 2-opt for local search, and evaluate based on constraints such as battery state of charge and regulatory rules. Five replicas are run for each algorithm.
5：Data Analysis Methods:
Results are analyzed by comparing total race time and execution time of algorithms, with validation against actual race times from the World Solar Challenge.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
