Scalable Many-Objective Pathfinding Benchmark Suite

Scalable Many-Objective Pathfinding Benchmark Suite 是由奥托·冯·格里克马格德堡大学的研究人员开发的一个多目标路径规划基准数据集，包含491个实例。该数据集基于真实世界数据，旨在解决物流、移动机器人等领域的路径规划问题，特别是面对多个冲突目标时的优化挑战。数据集通过定义五个目标函数（距离、旅行时间、事故延迟、路线特定特征如曲率和海拔）来模拟真实环境中的复杂性。创建过程中，研究者分析了不同实例的难度，并应用了三种知名的进化多目标算法。该数据集不仅作为基准测试，还可直接应用于真实世界的路线规划问题，为研究人员和工程师提供了一个评估多目标方法的平台。

The Scalable Many-Objective Pathfinding Benchmark Suite is a multi-objective pathfinding benchmark dataset developed by researchers from Otto von Guericke University Magdeburg, comprising 491 instances. Rooted in real-world data, this dataset is designed to address pathfinding problems in domains such as logistics and mobile robotics, particularly the optimization challenges posed by multiple conflicting objectives. It simulates the complexity of real-world environments by defining five objective functions: distance, travel time, accident delay, and route-specific characteristics including curvature and elevation. During its development, the researchers analyzed the difficulty levels of different instances and applied three well-known evolutionary multi-objective algorithms. This dataset not only serves as a benchmark for testing, but can also be directly applied to real-world route planning scenarios, providing a platform for researchers and engineers to evaluate multi-objective approaches.