面向变电站智能安监的行为识别与时空特征决策方法

打开文本图片集

中图分类号：TP277 文献标志码：A 文章编号：1008-0562（2026）02-0242-07

Abstract：To address the isues of insufficient granularity in feature extractionand limited adaptability to complex video surveillnce scenarios in substations inherent in traditional power personnel behavior recognition methods，this study investigates behavior recognition technology tailored to the needs of power operation and maintenance.Anend-to-end video behavior recognition framework is adopted to directlymodelraw surveillance videos，and a key frame extraction method based on spatiotemporal features is designed to improve inference efficiency.Abehaviorclassification decoder is constructedto enhance the discriminative ability for multiple types of operational actions.The experimental results on the real substation operation video dataset show that the proposed method achieves an overall recognition rate of 93.7% ，significantly outperforming traditional image recognition methods such assupport vector machine （SVM） and multi-layer perceptron （MLP） in both recognition accuracy and processng speed. The research conclusion provides a technical reference for improving theintelligent monitoring ability of power field.

Keywords：video surveilance;behaviorrecognition;spatiotemporal features;keyframe extraction;end-to-end; power operation and maintenance

0 引言

变电站作为电力系统的核心组成部分，承担着电压变换、电流分配等重要功能。（剩余11700字）