Charging guidance method and system based on gated graph convolution and madmpg
By constructing a traffic-electricity coupled network based on gated graph convolution and MADDPG, a multi-agent deep deterministic policy gradient algorithm is constructed by extracting spatiotemporal features and combining them with a priority experience replay mechanism. This solves the collaborative optimization problem of multi-vehicle charging guidance, realizes efficient and stable decision-making for charging station selection and route planning, and improves charging efficiency and economy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- STATE GRID JIANGSU ELECTRIC POWER CO LTD RESEARCH INSTITUTE
- Filing Date
- 2026-06-12
- Publication Date
- 2026-07-17
AI Technical Summary
Existing charging guidance technologies struggle to achieve collaborative optimization of multi-vehicle charging guidance in scenarios where transportation networks and charging networks are coupled, and suffer from insufficient collaborative decision-making capabilities, low training efficiency, and unstable optimization results.
A charging guidance method based on gated graph convolution and MADDPG is adopted. By constructing a traffic-electricity coupled network, the spatiotemporal features are extracted by gated temporal trend attention graph convolution and gated temporal convolution. Combined with a multi-agent deep deterministic policy gradient algorithm with a priority experience replay mechanism, policy learning is performed, and charging guidance results that take into account both time and economic costs are output.
It effectively solves the collaborative optimization problem of multi-vehicle charging guidance, improves training efficiency and strategy convergence, outputs charging guidance results that take into account both time and economic costs, and enhances the operational efficiency and economic benefits of charging stations.
Smart Images
Figure CN122402283A_ABST