TLab | Seminar 预告 交通系统建模优化与超大规模并行计算

作者：发布时间：2020-12-22浏览次数：85




seminar预告




      本周seminar我们有幸邀请到了天津大学朱宁副教授和东南大学刘志远教授，为大家带来“Team Orienteering with Time-Varying Profit”学术报告和“Distributed Computing Approaches for Large-scale Traffic Assignment Problem”学术报告，欢迎大家参会交流。


主讲人介绍


图片

朱宁，天津大学管理与经济学部信息管理与管理科学系副教授，硕士生导师。主要研究方向是交通与物流系统的建模与优化。关注于交通与物流系统中的选址问题、线路优化问题及资源分配问题的应用。主持国家自然科学基金项目3项，教育部项目2项。发表高水平论文20余篇，包括INFORMS Journal on Computing Transportation Science, Transportation Research Part B,C, European Journal of Operational Research, Journal of Scheduling等期刊。



     刘志远，东南大学交通学院教授、博导、副院长，复杂交通网络研究中心主任，东南大学网络空间安全学院博导。主要研究领域包括交通网络规划与管理、交通大数据分析与建模、公共交通、多模式物流网络等。迄今为止在这些领域中发表学术论文百余篇，其中被SCI/SSCI期刊检索70余篇，论文被引用2000余次。担任交通研究领域知名SCI期刊ASCE Journal of Transportation Engineering以及IET Intelligent Transport Systems副主编，担任国际期刊Transportation Research Part E（SCI/SSCI）、Transportation Research Record（SCI）、Journal of Transport and Land Use（SSCI）编委。指导学生获得多项国内外大数据算法比赛奖项（皆为前三名），包括被誉为“大数据比赛世界杯”的KDD CUP冠军（2020年）、与第二名（2019年），及其他同为人工智能三大国际顶级赛事的IJCAI（冠军，2019年）等。

  


讲座预览


第一场: Team Orienteering with Time-Varying Profit



主讲人：朱宁
Abstract:This paper studies the team orienteering problem, where the arrival time and service time affect the collection of profits. Such interactions result in a non-concave profit function. This problem integrates the aspect of time scheduling into the routing decision, which can be applied in humanitarian search and rescue operations where the survival rate declines rapidly. Rescue teams are needed to help trapped people in multiple affected sites whereas the number of people who could be saved depends as well onhow long a rescue team spends at each site. Efficient allocation and scheduling of rescue teams is critical to ensure a high survival rate. To solve the problem, we formulate a mixed integer non-concave programming model and propose a Benders branch-and-cut algorithm, along with valid inequalities for tightening the upper bound. To solve it more effectively, we introduce a hybrid heuristic that integrates a modified coordinate search (MCS) into an iterated local search. Computational results show that valid inequalities significantly reduce the optimality gap, and the proposed exact method is capable of solving instances where the mixed integer nonlinear programming solver SCIP fails in finding an optimal solution. In addition, the proposed MCS algorithm is highly efficient compared to other benchmark approaches whereas the hybrid heuristic is proven to be effective in finding high-quality solutions within short computing times. We also demonstrate the performance of the heuristic with the MCS using instances with up to 100 customers.

第二场: Distributed Computing Approaches for Large-scale Traffic Assignment Problem



主讲人：刘志远
Abstract: Traffic assignment is a fundamental tool to evaluate the flowdistribution pattern in a transport network. As one of the most recognized theory for traffic assignment, user equilibrium is widely investigated and implemented. Most of the existing algorithms for the user equilibrium-based traffic assignment problem are developed and implemented sequentially. This study aims to study and investigate the parallel computing approach to utilize the widely available parallel computing resources. The Parallel Block-Coordinate Descent (PBCD) algorithm is developed based on the path-based algorithm, i.e, the improved path-based gradient projection algorithm (iGP). A parallel block-coordinate method is proposed to replace the widely used Gauss-Seidel method for the procedure of path flow adjustment. To further improve the robustness/performance of the algorithm, the iPBCD algorithm is developed based on the state-of-the-art PBCD algorithm. A different type of flow update policy is studied and investigated intensively. The block size is determined using a sensitive test, and five indices-grouping rules are compared. Besides, a greedy update order of block indexes is introduced to compare with the cyclic scheme. Moreover, a new algorithm is developed based on the alternating direction method of multipliers (ADMM). In order to take use of the ADMM, the network links should be grouped into several blocks, where the linksin the same block are disconnected. This link grouping problem falls into the category of edge-coloring problem in graph theory, and it follows the Vizing theorem. Numerical examples show that the proposed algorithms perform well in convergence and efficiency and can significantly reduce the computing time.


参会方式

本次讲座为线上线下同步进行。

时间：2020/12/25 （周五） 19:00

线下会议地点：交通学院三楼322会议室

线上参会方式：腾讯会议 ID: 791 7485 8563


图片

欢迎交通学院所有师生

前来交流与探讨！