Method and system for predicting gas content in transformer oil based on joint model

Abstract

The invention discloses a method and a system for predicting gas content in transformer oil based on a joint model, and belongs to the field of transformer fault prediction, and the method comprises the steps: determining the type of a to-be-predicted gas related to a fault and a time sequence, and employing empirical mode decomposition and local mean decomposition to process an original sequencefor the non-stationary features of a concentration sequence of a dissolved gas in the oil; normalizing each subsequence component, and dividing a training sample and a test sample; respectively constructing a DBN prediction model for each subsequence component for training, building a DBN model through superposition and reconstruction, carrying out the feature extraction and classification of multi-dimensional data of the fault, and evaluating the prediction performance of the prediction model through the calculation of an error index. According to the method, the time sequence of the concentration of the dissolved gas in the oil is decomposed into components with higher stability and regularity, and the advantages of the deep belief network for multi-layer network feature extraction are utilized to achieve the purpose of predicting the state of the transformer so as to identify the fault in time.

Description

technical field [0001] The invention belongs to the field of transformer fault prediction, and more specifically relates to a method and system for predicting gas content in transformer oil based on a joint model. Background technique [0002] The safe operation of power equipment is the basis for the safe and stable operation of the power grid, especially as the key hub equipment of the power system, the safe and stable operation of power transformers is related to the power supply for the normal operation of society. At present, the fault diagnosis of power transformers is mainly divided into offline monitoring and online monitoring. The analysis of dissolved gas in oil is an important online measurement method for transformers. The prediction of the concentration of dissolved gas in oil can provide an important prediction basis for power transformer fault diagnosis and early warning. It is still recognized as an effective method to discover defects and latent faults of po...

AI Technical Summary

Problems solved by technology

[0005] In view of the above defects or improvement needs of the prior art, the present invention proposes a method and system for predicting gas content in transformer oil based on a joint model, which decomposes the time series of dissolved gas concentration in power transformer oil into more stable and regular Components, and then use the advantages of deep belief network to extract multi-layer network features to achieve the purpose of predicting the state of transformers to facilitate timely identification of faults, thereby solving the limitations of existing single diagnosis methods, and the combined model relies on expert experience. Determining the weight is too subjective, or there is a technical problem that the weight is negative

Images