[论文解读] Vehicle Powertrain Connected Route Optimization for Conventional, Hybrid and Plug-in Electric Vehicles

Zhiqian Qiao, Orkun Karabaşoğlu|arXiv (Cornell University)|Dec 5, 2016

Electric Vehicles and Infrastructure参考文献 25被引用 16

一句话总结

本文提出了一种新型路径优化策略——车辆动力系统互联路径优化（VPCRO），将车辆动力系统特性（如电池荷电状态、部件效率及车辆类型）整合至路径规划中，以最小化行驶成本。通过考虑实时交通与动力系统特异性能耗，VPCRO相较于传统最短距离路径规划，使传统车辆的平均行驶成本降低高达60%，电动车降低30%。

ABSTRACT

Most navigation systems use data from satellites to provide drivers with the shortest-distance, shortest-time or highway-preferred paths. However, when the routing decisions are made for advanced vehicles, there are other factors affecting cost, such as vehicle powertrain type, battery state of charge (SOC) and the change of component efficiencies under traffic conditions, which are not considered by traditional routing systems. The impact of the trade-off between distance and traffic on the cost of the trip might change with the type of vehicle technology and component dynamics. As a result, the least-cost paths might be different from the shortest-distance or shortest-time paths. In this work, a novel routing strategy has been proposed where the decision-making process benefits from the aforementioned information to result in a least-cost path for drivers. We integrate vehicle powertrain dynamics into route optimization and call this strategy as Vehicle Powertrain Connected Route Optimization (VPCRO). We find that the optimal paths might change significantly for all types of vehicle powertrains when VPCRO is used instead of shortest-distance strategy. About 81% and 58% of trips were replaced by different optimal paths with VPCRO when the vehicle type was Conventional Vehicle (CV) and Electrified Vehicle (EV), respectively. Changed routes had reduced travel costs on an average of 15% up to a maximum of 60% for CVs and on an average of 6% up to a maximum of 30% for EVs. Moreover, it was observed that 3% and 10% of trips had different optimal paths for a plug-in hybrid electric vehicle, when initial battery SOC changed from 90% to 60% and 40%, respectively. Paper shows that using sensory information from vehicle powertrain for route optimization plays an important role to minimize travel costs.

研究动机与目标

• 解决传统导航系统仅优先考虑最短距离或最短时间，而忽略车辆特异性能耗动态的问题。
• 开发一种路径规划策略，通过整合电池荷电状态（SOC）和交通条件下部件效率等动力系统特异性因素，最小化总行程成本。
• 评估基于动力系统特性的路径优化在传统车辆、混合动力车和插电式电动车中的差异。
• 量化VPCRO相较于传统最短距离路径规划所实现的成本节约与路径变化。
• 分析电动车辆初始电池荷电状态（SOC）对最优路径选择的敏感性。

提出的方法

• VPCRO框架将车辆动力系统模型与交通及道路网络数据相结合，计算每个路径段的能量消耗。
• 采用动态规划或最优控制技术，结合实时交通与车辆特异性效率特性，计算最低能耗路径。
• 该方法考虑了内燃机、电动机和电池系统在不同驾驶条件下部件效率的差异。
• 将能耗建模为速度、加速度、道路坡度和车辆荷电状态（SOC）的函数，实现对每条路径的精确成本预测。
• 系统评估多个候选路径，并选择总能耗成本最低的路径，而非仅基于距离或时间。
• 通过将初始电池SOC从90%至40%进行变化，执行敏感性分析，以评估插电式混合动力电动车的路径变化。

实验结果

研究问题

• RQ1 将车辆动力系统特性整合至路径规划中，如何影响传统车辆、混合动力车和插电式电动车的最优路径选择？
• RQ2 与最短距离路径规划相比，基于VPCRO的路径在多大程度上降低了行驶成本？
• RQ3 在插电式混合动力电动车中，初始电池荷电状态（SOC）如何影响最优路径选择？
• RQ4 在不同车辆类型中，传统最短距离路径与VPCRO优化路径之间的路径偏差程度如何？
• RQ5 在交通条件下，部件效率的差异如何影响所选路径的成本最优性？

主要发现

• 对于传统车辆，81%的行程在使用VPCRO而非最短距离路径规划时选择了不同的最优路径，平均成本降低15%，最高达60%。
• 对于电动化车辆（EVs），58%的行程在VPCRO下选择了不同的最优路径，平均成本节省6%，最高达30%。
• 当插电式混合动力电动车的初始电池荷电状态（SOC）从90%降至60%时，3%的行程选择了不同的最优路径。
• 进一步将初始SOC降至40%后，路径选择发生变化的行程比例上升至10%。
• 研究证实，车辆动力系统提供的传感数据显著提升了路径优化的准确性与成本效率。
• 结果表明，最低成本路径与最短距离或最短时间路径存在显著差异，尤其在交通状况和动力系统条件变化时更为明显。