D2Park: Diversified Demand-aware On-street Parking Guidance
summary
D2Park formulates parking as a Heterogeneous-Agent Dynamic Resource Allocation problem and builds a working system that combines GCN and LSTM sequence prediction with a model predictive control guided allocator. It implements separate mechanisms — DG-MRC cooperation for reserved users and DG-MNC competition for normal users — plus a three-step joint guidance algorithm. Evaluation uses four months and ~62 million on-street records from Shenzhen and reports at least 33% reduced average parking delay time.
AI-written summary · about
chapters (3)
related talks
18mRecognition and Localization of Parking Signs using Deep Learning
06mRISC: Resource-Constrained Urban Sensing Task Scheduling Based on Commercial Fleets
27mSolving Madrid's parking problem with Airflow and Streamlit
06mFairCharge: A Data-Driven Fairness-Aware Charging Recommendation System for Large-Scale Electric ...
06mData-Driven Battery-Lifetime-Aware Scheduling for Electric Bus Fleets
21mData Space to boost the European data economy
57mSmarter Tools for (Citi)Bike-Sharing
06mReal-time Travel Time Estimation with Sparse Reliable Surveillance Information