Transformer state knowledge acquisition method based on genetic algorithm and attribute support degree and equipment

Abstract

The invention discloses a transformer state knowledge acquisition method based on a genetic algorithm and an attribute support degree and equipment. According to the method, a decision-making table isestablished through transformer fault cases; the state knowledge under the specific state abnormal information of a transformer is solved by combining a genetic algorithm and an attribute support degree and the transformer state evolution rule is discovered, so that the calculation complexity is greatly reduced by being compared with the traditional method, and the condition attribute most related to the transformer fault type in the decision table can be efficiently extracted. The method can overcome the defects in the prior art, can be widely used for knowing the transformer state development rule, achieving transformer state evaluation, guiding transformer state monitoring and making differentiated operation and maintenance strategies; and the method is simple, efficient and beneficialto improving the operation safety of the power transformer.

Description

technical field [0001] The invention relates to fault diagnosis and state evaluation of power transformers, in particular to a transformer state knowledge acquisition method and equipment. Background technique [0002] In recent years, under the background of the energy Internet, with the large-scale development of the national smart grid, the application of various state sensing and detection technologies has become more mature, and intelligent detection and analysis technologies have begun to emerge in various fields of the power grid, such as smart patrolling. Inspection robots, unmanned inspection helicopters, and intelligent analysis and diagnosis systems, etc., my country's power grid has initially possessed the ability to grasp real-time data such as power equipment operation status, environmental changes, and operation inspection information through information technology. Diagnosis and risk prediction provide the basis. As one of the important equipment in the power ...

AI Technical Summary

Problems solved by technology

However, this kind of method relies heavily on the learning algorithm, and the algorithm input requires a lot of state information, and the relationship between each state information is unclear, and it is impossible to effectively understand the influence of each state information on the transformer fault; (3) Comprehensive evaluation: In order to Indicate the location and even the severity of the fault, and use as comprehensive information as possible to obtain a more refined transformer state conclusion by introducing fuzzy mathematics, rough set and other theories

However, at present, the above methods only focus on transformer fault diagnosis and fault cause analysis.

In addition, in the existing transformer state evaluation system, most of the research work only single-scores or simply weights the results represented by different state information, and seldom carries out the extraction and analysis of transformer state knowledge and the research on state development laws.

[0003] Looking at the existing transformer fault diagnosis and state evaluation methods, there are still the following problems: (1) The transformer fault diagnosis model relies on neural network algorithms, and the relationship between each state information is unclear, and it is impossible to effectively understand the impact of each state information on transformer faults It is not conducive to the extraction and analysis of the key state information of the transformer; (2) The existing information fusion method only calculates the weighted multi-dimensional state information of the transformer, and cannot find the difference between different abnormal state information, abnormal state information and different fault types. Progressive relationship, cannot efficiently form transformer state knowledge from fault cases

Images