CN-115035431-B - Unmanned aerial vehicle-based power supply site inspection method and system

CN115035431BCN 115035431 BCN115035431 BCN 115035431BCN-115035431-B

Abstract

The invention provides a power supply site inspection method and system based on an unmanned aerial vehicle, wherein the method comprises the steps of receiving first image data of a power site shot by the unmanned aerial vehicle, carrying out first feature extraction on the first image data, inputting the first feature into a first abnormality judgment model, judging whether an output result of the first abnormality judgment model is abnormal, if so, determining and acquiring second image data according to the first image data, carrying out second feature extraction on the second image data, inputting the second feature into a second abnormality judgment model, and determining whether the power site is abnormal according to an output result of the second abnormality judgment model. According to the invention, the first characteristic of the power equipment and the second characteristic which possibly causes false abnormality are subjected to abnormality judgment at the same time, so that the judgment accuracy of the abnormality of the power equipment in the unmanned aerial vehicle inspection process can be remarkably improved, and the false alarm probability is reduced.

Inventors

LIU BIN
CHEN YAN
LIU XIAOXUAN
LUO JIANQIN
YAN JING
WANG WEI
ZHANG CHENG
JIANG LE
WANG HUA
YANG TAO
LI DUANHUAN

Assignees

国网湖北省电力有限公司黄石供电公司
上海舒盈科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20220713

Claims (7)

1. A power supply field inspection method based on an unmanned aerial vehicle is characterized by comprising the following steps of receiving first image data of a power field shot by the unmanned aerial vehicle, carrying out first feature extraction on the first image data, inputting the first feature into a first abnormality judgment model, judging whether an output result of the first abnormality judgment model is abnormal, if so, determining and acquiring second image data according to the first image data, carrying out second feature extraction on the second image data, inputting the second feature into a second abnormality judgment model, and determining whether the power field is abnormal according to an output result of the second abnormality judgment model; The method comprises the steps of identifying each object in first image data, extracting contour line characteristics of each object based on the identification processing, judging whether each object meets a second preset condition according to the contour line characteristics, taking the image data of the area where the object which does not meet the second preset condition is located as second image data, wherein the second image data corresponds to a characteristic area of potential worker maintenance power equipment; The first characteristic extraction of the first image data comprises the steps of carrying out electric equipment object identification processing on the first image data, carrying out line characteristic extraction on the electric equipment object based on the identification processing, determining the inspection difficulty of the electric equipment object according to the line characteristic extraction, determining the amplification factor of the first image data according to the inspection difficulty, and carrying out amplification processing on the first image data according to the amplification factor to obtain third image data; The method comprises the steps of inputting the first characteristic into a first abnormality judgment model, judging whether an output result of the first abnormality judgment model is abnormal, wherein the step of inputting the first characteristic into the first abnormality judgment model, the step of outputting an abnormality evaluation value by the first abnormality judgment model, the step of judging that the output result of the first abnormality judgment model is abnormal if the abnormality evaluation value is larger than or equal to a first threshold value, the step of inputting the first characteristic into a third abnormality judgment model if the abnormality evaluation value is smaller than the first threshold value and larger than or equal to a second threshold value, and the step of judging that the output result of the first abnormality judgment model is abnormal if a matching result output by the third abnormality judgment model is not null.
2. The method for inspecting a power supply site based on an unmanned aerial vehicle according to claim 1, wherein the method further comprises the steps of judging whether a manual control instruction is received, and/or judging whether the current date and/or time meets a first preset condition, and/or judging whether an alarm signal of a power alarm system is received, and if so, generating an inspection control instruction, wherein the inspection control instruction is used for triggering the unmanned aerial vehicle to inspect the power supply site.
3. The unmanned aerial vehicle-based power supply site inspection method of claim 1, wherein after determining whether the power site has an abnormality according to the output result of the second abnormality determination model, further comprises outputting an abnormality determination result to a worker for confirmation, and determining an output object of the abnormality determination result according to the confirmation result.
4. The unmanned aerial vehicle-based power supply field inspection method of claim 1, wherein the first anomaly determination model and the second anomaly determination model are both constructed through a deep learning algorithm.
5. An artificial intelligence-based inspection scheme making system comprises a receiving module, a processing module and a storage module, wherein the processing module is connected with the receiving module and the storage module, the storage module is used for storing executable computer program codes, the receiving module is used for receiving first image data of an electric power field shot by an unmanned aerial vehicle and transmitting the first image data to the processing module, and the processing module is characterized by executing the method according to any one of claims 1-4 by calling the executable computer program codes in the storage module.
6. An electronic device comprising a memory storing executable program code, a processor coupled to the memory, wherein the processor invokes the executable program code stored in the memory to perform the method of any of claims 1-4.
7. A computer storage medium having a computer program stored thereon, characterized in that the computer program when run by a processor performs the method according to any of claims 1-4.

Description

Unmanned aerial vehicle-based power supply site inspection method and system Technical Field The invention relates to the technical field of power inspection, in particular to a power supply field inspection method, a power supply field inspection system, electronic equipment and a computer storage medium based on an unmanned aerial vehicle. Background The intelligent inspection unmanned aerial vehicle is an important product of new century artificial intelligence, virtual reality technology and the like. The intelligent inspection unmanned aerial vehicle can timely and completely record various data in the electric power inspection practice, realize comprehensive monitoring, timely find hidden danger in production and ensure safe and reliable operation of an electric power system. In the prior art, unmanned aerial vehicle inspection generally utilizes detection means such as image recognition and thermal imaging to extract characteristics of power equipment, and judges whether the unmanned aerial vehicle inspection belongs to an abnormal state or not based on a preset judgment rule or depth recognition model. However, the power equipment is in an "abnormal" state and has a true or false score, for example, when a worker maintains the power equipment, the state of the power equipment is also "abnormal", and the prior art has little research on the situation, and cannot effectively solve the problem of misjudgment in the situation. Disclosure of Invention In order to at least solve the technical problems in the background art, the invention provides a power supply field inspection method, a system, electronic equipment and a computer storage medium based on an unmanned aerial vehicle. The first aspect of the invention provides a power supply field inspection method based on an unmanned aerial vehicle, which comprises the following steps: Receiving first image data of an electric power field shot by an unmanned aerial vehicle, and carrying out first feature extraction on the first image data; inputting the first characteristic into a first abnormality judgment model, and judging whether the output result of the first abnormality judgment model is abnormal or not; If yes, determining and acquiring second image data according to the first image data, extracting second features of the second image data, and inputting the second features into a second abnormality judgment model; And determining whether the power field is abnormal according to the output result of the second abnormality judgment model. Further, the method further comprises: judging whether a manual control instruction is received, and/or judging whether the current date and/or time meets a first preset condition, and/or judging whether an alarm signal of an electric power alarm system is received; if yes, generating a patrol control instruction, wherein the patrol control instruction is used for triggering the unmanned aerial vehicle to patrol the power supply site. Further, the performing a first feature extraction on the first image data includes: Performing power equipment object identification processing on the first image data, performing line feature extraction on the power equipment object based on the identification processing, and determining the inspection difficulty of the power equipment object according to the line feature extraction; determining the amplification factor of the first image data according to the inspection difficulty, and amplifying the first image data according to the amplification factor to obtain third image data; and carrying out first feature extraction on the third image data. Further, the determining and acquiring the second image data according to the first image data includes: performing recognition processing on each object in the first image data, and performing contour line feature extraction on each object based on the recognition processing; Judging whether each object meets a second preset condition according to the contour line characteristics; And taking the image data of the area where the object does not meet the second preset condition as the second image data. Further, the inputting the first feature into a first abnormality determination model, determining whether an output result of the first abnormality determination model is abnormal, includes: Inputting the first characteristic into a first abnormality judgment model, wherein the first abnormality judgment model outputs an abnormality evaluation value; If the abnormality evaluation value is greater than or equal to a first threshold value, judging that the output result of the first abnormality judgment model is abnormal; And if the abnormal evaluation value is smaller than a first threshold value and larger than or equal to a second threshold value, inputting the first characteristic into a third abnormal judgment model, and if the matching result output by the third abnormal judgment model is not null, judging that the output result of the