Multiagent-Based Route Guidance for Increasing the Chance of Arrival on Time
Authors: Zhiguang Cao, Hongliang Guo, Jie Zhang, Ulrich Fastenrath
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on real road networks justify its ability to increase the chance of arrival on time. We conduct experiments in various settings to extensively compare our route guidance approach with existing methods, showing its advantages of increasing the chances of reaching destination before deadline for all vehicles. All experiments are conducted on SUMO (Behrisch et al. 2011). |
| Researcher Affiliation | Collaboration | Zhiguang Cao1,2, Hongliang Guo1,2, Jie Zhang1, and Ulrich Fastenrath3 1School of Computer Engineering, Nanyang Technological University, Singapore 2Energy Research Institute @NTU, Nanyang Technological University, Singapore 3Department of Traffic Information Management and Routing Optimization, BMW Group, Germany |
| Pseudocode | Yes | Algorithm 1: Multiagent-based Route Guidance |
| Open Source Code | No | The paper cites tools like SUMO and LIBSVM that were used, but it does not provide any link or explicit statement about releasing the source code for the methodology or approach described in this paper. |
| Open Datasets | No | The paper mentions using 'real road networks' (Singapore and New York) and running simulations to get 'historical traffic data', but it does not provide concrete access information (e.g., links, DOIs, specific repository names, or citations to data sources) for these road networks or the generated historical data. |
| Dataset Splits | No | The paper describes running simulations 'for 250 times to get an expected travel time' and 'for 500 times under each setting' and generating random origins and destinations, but it does not specify any training, validation, or test dataset splits in terms of percentages, sample counts, or predefined partition strategies. |
| Hardware Specification | Yes | Particularly, all experiments are conducted on an ordinary PC with Intel Core i7-3540M processor and 8.00 GB RAM. |
| Software Dependencies | No | The paper mentions using 'SUMO (Behrisch et al. 2011)', 'Pyomo (www.pyomo.org)', and 'SVR (Chang and Lin 2011)', but it does not provide specific version numbers for these software components. |
| Experiment Setup | Yes | The configurations of vehicles are as follows: length is 5m; minimal gap is 2.5m; car following model is Krauss (Behrisch et al. 2011); origins and destinations are randomly generated; traffic light duration: Tg = Tr = 20s; vehicles will not occupy road resources when reaching destinations. In addition, we introduce a positive parameter α to denote different levels of deadlines. ... We run the simulation for 500 times under each setting. ... adopt λ=0.8 and 0.4. ... α is set as 0.8 for tight deadline, and 1.2 for loose deadline. |