Transformer area phase sequence identification method and device based on multi-layer stacked neural network

Abstract

The invention relates to a transformer area phase sequence identification method and device based on a multilayer stacked neural network, and the method comprises the steps: obtaining time sequence voltage sample data between each phase low-voltage outgoing line of a distribution transformer and each user ammeter in a certain time period of a target transformer area, and carrying out the preprocessing of the time sequence voltage sample data; generating a training set and a test set for the sample set obtained after processing, training the training set by using a CNN network to obtain time sequence features, and training the time sequence features and the training set by using an LSTM network to establish a phase sequence prediction model; and predicting the phase sequence of each phase of the user ammeter and the distribution transformer in the test set by using the phase sequence prediction model. According to the transformer area phase sequence identification method, on the premise that other terminal equipment does not need to be externally hung in the target transformer area, the phase sequence affiliation relation of the user electricity meters can be accurately sorted, the cost is low, the engineering application value is high, and the problem that terminal equipment is additionally added in existing transformer-line-phase-household physical topology identification of the transformer area is solved.

Description

technical field [0001] The invention relates to the technical field of low-voltage power distribution network, in particular to a method and device for identifying phase sequence of a station area based on a multi-layer stacked neural network. Background technique [0002] In the low-voltage distribution network, the traditional low-voltage operation and maintenance management mode is used for management. The traditional low-voltage operation and maintenance management mode lacks the support of the topological relationship of the station area, which easily leads to untimely power outage notification, untimely rush to repair power, and low-voltage solution time. Long-term or unresolved problems, frequent changes in the station area, abnormal line loss in the station area, etc., which in turn lead to dissatisfaction with electricity users. For this reason, it is particularly important to study the physical topology recognition technology of "transformation-line-phase-household...

AI Technical Summary

Problems solved by technology

[0003] The embodiment of the present invention provides a multi-layer stacked neural network-based phase sequence recognition method and device for the station area, which is used to solve the problem of using the injection signal method, The identification of the "change-line-phase-household" physical topology of the station area by the data label method or data analysis method has technical problems such as additional terminal equipment, large investment, and heavy operation and maintenance pressure

Images