The Vehicle Routing Problem (VRP) has been widely studied throughout its history as a way of optimizing routes by minimizing distances, and the issue of risk in VRP has been received less attention, which is essential to increase transport safety, to reduce accident costs and to improve delivery reliability. In this way, this paper aims to support decision makers to plan routes for a road freight company considering both, i.e. logistics cost and safety. An analytical approach based on statistics was developed in which official government data of accidents occurrences and data from cargo insurance companies were used to estimate the risk cost of routes using the Monte Carlo simulation. The Capacitated Vehicle Routing Problem (CVRP) was employed and logistics and risk costs were minimized by varying a specific safety level coefficient and the model generated solutions with safer routes, reducing risk cost by up to approximately 18%. The accidents probabilities and risk costs of each route represented values as expected and the main contributions of this paper is the applicability of the approach to support decision markers to choose routes considering safety and logistics costs, and to be a simple and adaptable methodology for any VRP model. In addition Knime Analytics Platform was also used to estimate the accidents probabilities and to simplify data exploration, analysis, visualization and interpretation.
翻译:历来对车辆出行问题进行了广泛研究,作为通过尽可能缩短距离优化路线的一种方式,历来对车辆出行问题进行了广泛研究,对车辆出行风险问题重视度较低,这对于提高运输安全性、降低事故成本和提高交货可靠性至关重要,因此,本文件旨在支持决策者规划公路货运公司路线,既考虑物流成本和安全,又考虑物流成本和安全;根据统计数据制定了分析方法,利用政府官方事故数据以及货物保险公司的数据来估算使用蒙特卡洛模拟的路线风险成本;采用了机动车辆出行问题(CVRP),采用不同的具体安全等级系数将物流和风险成本降低到最低程度,采用更安全路线的模型产生解决方案,将风险成本降低到大约18%;每种路线的意外概率和风险成本按预期值计算,以及本文件的主要贡献是采用这一方法支持决策标志选择安全和后勤成本,并成为任何VRP模型的简单和调整方法;此外,除了Knime Analy分析外,还使用模型和视觉分析模型来估算风险。