A Method for Predicting the Error of Electronic Transformer

Abstract

The invention discloses a method for predicting errors of electronic transformers, comprising the following steps: collecting error data and environmental parameter data of electronic transformers to generate sample data, and eliminating abnormal data among them; Carry out standardization processing; perform clustering processing on the historical data of environmental parameters, and establish an error prediction model of electronic transformers through training and learning; predict the ratio difference and angle difference of electronic transformers based on the prediction model according to the environmental parameter values. Advantages: the present invention does not need to establish any physical model, based on the multi-dimensional data-driven method, according to the error data of the electronic transformer and the environmental parameter data, it can realize the online estimation of the error of the electronic transformer, and solve the problem of the error of the electronic transformer and There is no problem of determining the functional relationship of the environmental parameters, which is conducive to improving the accuracy of the error prediction of the electronic transformer.

Description

technical field [0001] The invention belongs to the field of status evaluation and fault diagnosis of power transmission and distribution equipment, and more specifically relates to a method for predicting errors of electronic transformers based on a clustering neural network. Background technique [0002] The electronic transformer is one of the key equipment for the digitization, automation, informatization and interaction of smart substations. After the electronic transformer is adopted, the optical fiber of signal transmission fundamentally solves the problem of additional error in the transmission of the electromagnetic transformer to the secondary equipment, and greatly improves the accuracy of the measurement and metering system. However, from the point of view of field operation problems, the accuracy of electronic transformers still occupies a large proportion. Although all the electronic transformers in operation have passed the type test and factory test, their e...

AI Technical Summary

Problems solved by technology

However, this method does not consider the influence of outliers in the sample data on the structure and parameters of the model, and the model also has shortcomings such as easy to fall into local minimum and slow convergence.

[0006] To sum up, the existing technology cannot evaluate the long-term operation error of the electronic transformer, and the on-site implementation is difficult, requiring heavy labor operations, greatly affected by the external environment, and the on-site prediction results may have large errors and are unreliable, etc.

Images