Hydraulic power plant security situation awareness system and method based on multi-modal fusion and reinforcement learning

Abstract

The invention relates to a hydraulic power plant security situation awareness system and method based on multi-modal fusion and reinforcement learning, and the system comprises a multi-modal data collection module which collects multi-modal data; the multi-modal feature fusion module is used for carrying out feature extraction and feature fusion; the security situation evaluation module is used for establishing an evaluation model and carrying out security early warning and model updating; the reinforcement learning module is used for performing reinforcement learning on the evaluation model, establishing a dynamic environment factor model and dynamically adjusting the safety early warning and model updating process; and the linkage early warning module is used for carrying out multi-dimensional early warning and carrying out model optimization on the dynamic environment factor model. A self-adaptive early warning mechanism based on reinforcement learning driving is established, dual adjustment based on environment perception and experience driving is realized, false alarm is effectively controlled according to a secondary verification mechanism, and end-to-end intelligent decision and automatic linkage are realized.

Images