Selective Publications代表著作 Grants/Contracts科研项目 Awards科技奖励 Patents/Software Copyright专利/软件著作权 Selective Presentations国际报告 Models & Algorithms模型与算法

Models & Algorithms

    模型与算法

Our group develops reusable epidemic models and reinforcement learning algorithms for spatiotemporal epidemic simulation, risk assessment, intervention evaluation, and intelligent prevention and control. The model resources emphasize transparent simulation components that can be reused for scenario analysis, while the algorithm resources focus on optimization methods for adaptive and cost-aware epidemic control. Together, these resources support reproducible research and public health decision analysis.

本课题组围绕传染病时空建模、风险评估、干预评估与智能防控，持续整理和维护可复用的模型与算法资源。模型资源强调可复用、可解释的仿真组件，算法资源侧重面向多区域、多个体和不完全观测条件的防控策略优化。这些资源可服务于情景推演、干预评估、方法复现与公共卫生决策分析。

Epidemic Models / 传染病模型

Metapopulation ODE Model / 元群体 ODE 模型

The metapopulation ODE model captures epidemic transmission across connected urban regions by coupling compartmental disease dynamics with inter-regional mobility. It can be used to evaluate region-level intervention policies under different transmission scenarios.

该模型通过耦合分区疾病传播动力学与区域间人口流动，刻画城市内部多个区域之间的传播过程，可用于评估不同传播情景下的区域级防控策略。

Metapopulation ODE epidemic trajectory
Example trajectory generated from the packaged metapopulation ODE model with Shenzhen data. / 基于深圳示例数据运行元群体 ODE 模型得到的传播曲线。

GitHub / Download: metapop_ode

Agent-based Model / 个体级 ABM 模型

The agent-based model simulates epidemic spread at the individual level and supports population attributes, spatial structure, disease progression, interventions, and mobility-related responses.

该模型从个体尺度模拟疫情传播过程，支持人口属性、空间结构、疾病状态演化、干预措施以及个体流动与响应行为等要素的表达。

Agent-based epidemic model overview
Model component overview generated from the ABM module design. / 根据 ABM 模块设计生成的模型组成示意图。

GitHub / Download: abm

Reinforcement Learning Algorithms / 强化学习算法

DRL_EPC_STO

This algorithm optimizes epidemic control strategies with a focus on coordinated interventions across spatial and temporal dimensions.

该算法面向传染病防控策略优化，强调干预措施在时间与空间维度上的协调性与有序性。

DRL_EPC_STO framework
Framework of DRL_EPC_STO. / DRL_EPC_STO 算法框架。

GitHub: https://github.com/lexie1216/DRL_EPC_STO

Paper: https://www.tandfonline.com/doi/full/10.1080/13658816.2024.2431904

Li, X., Yin, L., Liu, K., Zhu, K., & Cui, Y. (2024). Deep-reinforcement-learning-based optimization for intra-urban epidemic control considering spatiotemporal orderliness. International Journal of Geographical Information Science, 1-26.

Hi_RICE_EPC

This framework optimizes large-scale individual mobility interventions through a hierarchical region-individual control structure, balancing epidemic mitigation and intervention costs.

该框架通过区域-个体分层协同控制结构优化大规模个体流动干预策略，在疫情抑制效果与干预成本之间实现平衡。

Hi_RICE_EPC framework
Framework of Hi_RICE_EPC. / Hi_RICE_EPC 算法框架。

GitHub: https://github.com/YuxiaoLuo0013/Hi_RICE_EPC

Paper: https://www.sciencedirect.com/science/article/abs/pii/S0198971525000651

Luo, Y., Yin, L., Zhu, K., & Liu, K. (2025). Architecting urban epidemic defense: A hierarchical region-individual control framework for optimizing large-scale individual mobility interventions. Computers, Environment and Urban Systems, 121, 102312.

EpidemicControl-IncompleteObs

This method studies epidemic control under incomplete observation and focuses on robust intervention decisions when the true epidemic state cannot be fully observed.

该方法关注不完全观测条件下的疫情防控问题，研究在真实疫情状态无法被完全观测时如何进行稳健的干预决策。

EpidemicControl-IncompleteObs framework
Framework of EpidemicControl-IncompleteObs. / 不完全观测条件下疫情防控算法框架。

GitHub: https://github.com/xiangsiersheng-web/EpidemicControl-IncompleteObs