System security management method
By combining reinforcement learning, optimal control, and game theory models, the system dynamically adjusts access control, solving the problems of lagging risk assessment and imbalance between security and resource utilization in existing technologies. This enables the system to respond quickly and optimize resources when faced with dynamic user behavior.
Patent Information
- Application Number
- CN202510275523.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-10
- Publication Date
- 2025-11-28
- Estimated Expiration
- 2045-03-10
AI Technical Summary
Existing access control methods lack dynamic adjustment capabilities, have lagging risk assessments, and cannot balance security and resource utilization. This results in the system being unable to respond in a timely manner when faced with dynamic and complex user behaviors, increasing security risks and potentially leading to resource waste.
By combining reinforcement learning and optimal control theory, user behavior is monitored in real time. The interaction between the system and the user is optimized through a game theory model, and permissions are dynamically adjusted by an adaptive feedback mechanism to achieve a balance between security and resource utilization.
It enables rapid response to abnormal behavior, improves system flexibility and response speed, ensures that the system provides high security while making reasonable use of resources and avoiding excessive restriction of permissions.
Smart Images

Figure CN120105460B_ABST
Abstract
Citation Information
Patent Citations
Network security evaluation system and method based on dynamic attack and defense game model
CN119544307A