Method and device for fault monitoring of transformer respirator

Abstract

The embodiment of the invention provides a method and a device for monitoring the failure of a transformer respirator, and relates to the technical field of image processing. The method comprises the following steps: firstly, obtaining an orthographic projection image of a color card and an image of a to-be-detected feature body; then, parsing the color degree of the image of the to-be-detected feature body; finally, judging whether the parsed color degree of the image of the to-be-detected feature body is consistent with the prestored standard color degree or not, and generating failure prompt information if the color degree of the image of the to-be-detected feature body is not consistent with the prestored standard color degree. The method and the device for monitoring the failure of the transformer respirator, provided by the invention, have the characteristics of more instantaneity and accuracy, and certain manpower and material resources can be reduced.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a transformer respirator fault monitoring method and device. Background technique [0002] Excess air inside the transformer is exhaled when the transformer oil temperature rises, and outside air is sucked in when the transformer oil temperature cools. Since the external air contains water, which will pollute the oil after entering the transformer, a transformer respirator is installed. Transformer respirator is also called transformer hygroscopic device, which can absorb moisture and sundries in the air, so that the transformer oil is not intruded by moisture in the external air, and ensures the insulation strength of the transformer oil. [0003] Whether the transformer respirator is in normal operation can be judged by the change of the color of the feature body to be detected in the transformer. At present, operation and maintenance personnel generally inspect the ...

AI Technical Summary

Problems solved by technology

At present, the operation and maintenance personnel generally inspect the operation status of the transformer respirator on a regular basis. This traditional detection method not only wastes a lot of manpower and material resources, but also cannot quickly and accurately realize 24-hour remote monitoring.

Images