Li Tianyang, He Zhangyi, Wu Yanling
Computer Science, Northeast Electric Power University, Jilin, Jilin, China.
Jiangxi New Energy Technology Institute, Xinyu, Jiangxi, China.
PeerJ Comput Sci. 2022 Dec 1;8:e1170. doi: 10.7717/peerj-cs.1170. eCollection 2022.
With the rapid growth of express delivery in urban areas, the use of driverless vehicles as an alternative to traditional human delivery can reduce costs and improve efficiency. The route planning of driverless vehicles is crucial in realizing autonomous navigation, which improves the working level and ensures improvements in efficiency. However, it is difficult to reasonably organize the real-time delivery, taking into account several factors that influence the planning of routes, such as load capabilities, power limits and traffic conditions. To deal with this concern, we propose an integrated approach including a multistage model and improved genetic algorithm to obtain the optimal delivery plan for driverless vehicles. The experimental results in an urban scenario with a realistic delivery service show the superiority of our proposition in the delivery efficiency.
随着城市地区快递业务的快速增长，使用无人驾驶车辆替代传统人工配送能够降低成本并提高效率。无人驾驶车辆的路径规划对于实现自主导航至关重要，这能提升工作水平并确保效率的提高。然而，考虑到影响路径规划的多个因素，如载重能力、功率限制和交通状况，合理组织实时配送具有一定难度。为解决这一问题，我们提出一种综合方法，包括多阶段模型和改进的遗传算法，以获得无人驾驶车辆的最优配送方案。在具有实际配送服务的城市场景中的实验结果表明了我们的方案在配送效率方面的优越性。
