Substation auxiliary equipment monitoring system

Abstract

The invention discloses a transformer substation auxiliary equipment monitoring system, particularly relates to the field of transformer substation supervision, and is used for solving the problems of temperature and humidity sampling drift, regulation and control misalignment and maintenance lag caused by local blockage of a traditional monitoring system. A pressure difference-fractal coupling feature vector and a wind speed-entropy change feature vector are extracted by adopting a layered trend-removed residual image convolutional network and wavelet multi-scale decomposition, a blocking interaction spectrum value is generated by combining a polar coordinate fusion model, a fingerprint vector is constructed, and a dynamic cluster center trajectory is generated by utilizing dimension reduction mapping of a self-organizing mapping network to quantify the blocking depth; an equivalent environment field is reconstructed based on a heat and humidity coherence model to compensate sampling deviation, a fuzzy control strategy is driven to generate a segmented smooth load instruction to achieve air conditioner-dehumidification collaborative optimization, accurate maintenance is triggered through maintenance cache increment comparison and linear regression man-hour prediction, and full-process closed-loop intelligent monitoring of the transformer substation auxiliary equipment room is achieved.

Images