Distribution Optimization for Connected Autonomous Vehicles (CAV) Considering Fuel Consumption Optimization

With the maturity of autonomous driving and Internet of Vehicles, as well as the increasingly serious climate change problem, the research and application of smart logistics and green logistics have attracted much attention. With urban distribution as the research object, a distribution optimization model considering connected autonomous vehicles (CAV) for fuel consumption optimization at intersections is constructed, and an enhanced differential evolutionary algorithm (EDEA) is designed to find the optimal solution. The model considers intersection fuel consumption, time window constraint and penalty cost, and aims at minimizing the sum of distribution cost, carbon tax cost and penalty cost, focusing on optimizing the speed strategy of CAV at intersections to reduce fuel cost and carbon emission. EDEA generates the initial population by backward learning strategy and random strategy, which effectively improves the quality of the initial population. The intersection operator and variation operator in the differential evolution algorithm are tested, and the operator with the best adaptation degree is selected from them. The distribution optimization model and EDEA are applied to verify the distribution problem of autonomous vehicles in a courier distribution center in Xi'an. The results show that the constructed model considers fuel consumption at intersections and can decrease the carbon tax cost, distribution cost and penalty cost arising from the speed change of autonomous vehicles at the intersection. The EDEA proposed in this paper has a better performance in finding the optimal value, and the optimization target value is 18.3% lower than the Differential Evolutionary Algorithm (DEA) and 35.9% lower than the Standard Genetic Algorithm (SGA).