Power material inspection, storage and distribution data collection and processing method and system based on artificial intelligence

Abstract

The present invention relates to the technical field of big data processing of electric power materials, and in particular to a method and system for collecting and processing data for inspection, storage and distribution of electric power materials based on artificial intelligence. It is applied to the detection, storage and distribution of electric power materials. A sample original database is established through sample pictures and sample videos, and the sample original database is stored in a local cache. The entry of sample original data is realized through the local cache, which is convenient for comparing samples and real objects in the future. The local cache retrieves the corresponding samples, and the photos of the real objects for on-site inspection are uploaded through the cloud server for comparison, and the human operation error is reduced through the comparison of artificial intelligence models. On-site inspectors can compare samples and real objects by taking pictures. The method of this application greatly reduces the requirements for shooting quality, and only requires orthographic projection shooting. There are no requirements for position, proportion, and centering. The aspect ratio is compared according to the principle of equal proportion, and the three-view comparison is realized by the left and right view groups, the pitch view group, and the front and back view groups. It takes into account both inspection efficiency and accuracy.

Description

Technical Field

[0001] The present invention relates to the field of big data processing technology for electric power materials, in particular to an artificial intelligence-based data collection and processing method and system for electric power material inspection, storage and distribution, which is applied to the inspection, storage and distribution of electric power materials. Background Art

[0002] Power material inspection, storage and distribution is a modern management system that combines inspection, storage and distribution functions, aiming to improve the targeting of grid material sampling, storage reliability and supply timeliness. Power material inspection, storage and distribution is a full-chain modern logistics system based on geographically close quality inspection centers, storage warehouses and distribution workshops, with material inspection, storage and distribution functions. The purpose of this system is to improve the targeting of grid material sampling, storage reliability and supply timeliness, thereby ensuring the stable operation and efficient service of the power system.

[0003] However, in actual applications, it is found that there are many types of power materials, with huge differences in material sizes and specifications. It is often necessary to manually check the power material input information through nameplates or bills. It is often easy to make mistakes, such as nameplate or bill batch errors, quantity or size mismatches, and it is often difficult for manual work to find the above problems, especially for some large equipment with strict size requirements. It is generally difficult for on-site staff to find size deviations. For example, the required size is 10 meters, but the nameplate or bill is marked as 8 meters. Due to the on-site observation angle and the deviation of the naked eye, it is often difficult to detect the size problem. In addition, if there is no handy tool, the size cannot be accurately measured. The above materials are stored and distributed according to the deviation size, and there is a risk of subsequent unusability, model replacement, and product return. The storage and transportation costs of large-scale power equipment are high, which affects the timeliness of supply and increases costs. At the same time, cables have a variety of different diameters, which further increases the difficulty of cable information entry and inspection. Power materials are not just cables. A large number of shells and components have various shapes, which further aggravates the difficulty of entry and inspection. At present, the method of increasing personnel and multi-post quality inspection is adopted to reduce error deviations, but the efficiency of entry and inspection is not high, and the reliability of entry and inspection cannot be effectively guaranteed. Summary of the invention

[0004] The present invention solves the deficiencies of the prior art and provides a convenient, efficient and highly operable artificial intelligence-based method and system for collecting and processing data on inspection, storage and distribution of electric power materials.

[0005] The present invention solves the technical problem by adopting the following technical solution: An AI-based data collection and processing system for power material inspection, storage and distribution, including raw data, inspection data, local cache, cloud server and camera equipment; The original data includes sample pictures and sample videos of power materials; The inspection data includes physical pictures and videos of power materials; Photographing equipment, which is used for photographing original data and inspection data; A local cache, which is used to receive the original data for local storage; The cloud server is used to query the original data in the local cache archive and upload the verification data for comparison with the original data; The original data is converted into sample pictures and sample videos through the shooting equipment and uploaded to the local cache to form the sample original database; the inspection data is converted into physical pictures and physical videos through the shooting equipment, which are compared with the sample original database through the cloud server, and the cloud server loads the target tracking algorithm model.

[0006] It also includes standard reference data, which includes a comparison picture of the actual power material and the standard reference object, and the actual power material and the standard reference object are located on the same picture. The standard reference data is formed by a shooting device to form a comparison picture, and the standard reference object sets a standard size, which is 1m, 1cm or 1mm, and the corresponding size of the actual object in the comparison picture is obtained by calculation through a cloud server.

[0007] The method for using the power material inspection, storage and distribution data acquisition and processing system based on artificial intelligence includes the following steps: S1. Raw data extraction Select a single-frame screenshot of a video of a power material sample or a sample picture; calculate the aspect ratio value A of the sample edge through the target tracking algorithm model; S2. Inspection data extraction Use a single-frame video screenshot or sample picture of the actual power material; calculate the aspect ratio value B of the edge of the inspection object through the target tracking algorithm model; S3. Comparison of the length-to-width ratio between the sample and the actual object Calculate the ratio = A / B; if the ratio result is within the range of 95%~105%, the test is passed; otherwise, if the ratio result is not within the range of 95%~105%, the test is failed.

[0008] The following steps are also included: S4. View direction replacement check The single-frame screenshot of the video or the sample image is taken using orthogonal projection; it is divided into left and right view group, pitch view group, front and back view group according to the orthogonal projection direction; a single-frame screenshot of the video or a sample image is selected for each view group; after replacing the view, the aspect ratio value A′ of the sample edge is calculated by the target tracking algorithm model, and the aspect ratio value B′ of the edge of the physical object to be inspected is calculated by the target tracking algorithm model; S3 is used to calculate the ratio = A′ / B′; the three groups of ratio results are all within the range of 95%~105% and pass the inspection; otherwise, any group of the three groups of ratio results is not within the range of 95%~105% and fails the inspection.

[0009] S5. Physical size inspection Select the physical object that has passed the inspection in step S4, place a standard reference object next to the physical object and take a comparison picture by orthogonal projection, set the standard size of the standard reference object, the standard size is 1m, 1cm or 1mm, and obtain the corresponding size of the physical object in the comparison picture through calculation on the cloud server.

[0010] The beneficial effects of the present invention are as follows: The sample original database is established through sample pictures and sample videos. The sample original database is stored in the local cache. The sample original data is entered through the local cache, which is convenient for comparing samples and real objects in the future. The local cache retrieves the corresponding samples, and uploads the photos of the real objects for on-site inspection through the cloud server for comparison. The artificial intelligence model comparison reduces human operation errors. On-site inspectors can compare samples and real objects only by taking photos. The application method greatly reduces the requirements for shooting quality. It only requires orthographic projection shooting. There are no requirements for position, proportion, and centering. The aspect ratio is compared according to the principle of equal proportion, and the three-view comparison is achieved through the left and right view groups, the pitch view group, and the front and back view groups. It takes into account both inspection efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a schematic diagram of the data collection and processing system for inspection, storage and distribution of electric power materials of the present invention. DETAILED DESCRIPTION

[0012] An AI-based data collection and processing system for power material inspection, storage and distribution, including raw data, inspection data, local cache, cloud server and camera equipment; The original data includes sample pictures and sample videos of power materials; The inspection data includes physical pictures and videos of power materials; Photographing equipment, which is used for photographing original data and inspection data; A local cache, which is used to receive the original data for local storage; The cloud server is used to query the original data in the local cache archive and upload the verification data for comparison with the original data; The original data is converted into sample pictures and sample videos through the shooting equipment and uploaded to the local cache to form the sample original database; the inspection data is converted into physical pictures and physical videos through the shooting equipment, which are compared with the sample original database through the cloud server, and the cloud server loads the target tracking algorithm model.

[0013] It also includes standard reference data, which includes a comparison picture of the actual power material and the standard reference object, and the actual power material and the standard reference object are located on the same picture. The standard reference data is formed by a shooting device to form a comparison picture, and the standard reference object sets a standard size, which is 1m, 1cm or 1mm, and the corresponding size of the actual object in the comparison picture is obtained by calculation through a cloud server.

[0014] The method for using the power material inspection, storage and distribution data acquisition and processing system based on artificial intelligence includes the following steps: S1. Raw data extraction Select a single-frame screenshot of a video of a power material sample or a sample picture; calculate the aspect ratio value A of the sample edge through the target tracking algorithm model; S2. Inspection data extraction Use a single-frame video screenshot or sample picture of the actual power material; calculate the aspect ratio value B of the edge of the inspection object through the target tracking algorithm model; S3. Comparison of the length-to-width ratio between the sample and the actual object Calculate the ratio = A / B; if the ratio result is within the range of 95%~105%, the test is passed; otherwise, if the ratio result is not within the range of 95%~105%, the test is failed.

[0015] The following steps are also included: S4. View direction replacement check The single-frame screenshot of the video or the sample image is taken using orthogonal projection; it is divided into left and right view group, pitch view group, front and back view group according to the orthogonal projection direction; a single-frame screenshot of the video or a sample image is selected for each view group; after replacing the view, the aspect ratio value A′ of the sample edge is calculated by the target tracking algorithm model, and the aspect ratio value B′ of the edge of the physical object to be inspected is calculated by the target tracking algorithm model; S3 is used to calculate the ratio = A′ / B′; the three groups of ratio results are all within the range of 95%~105% and pass the inspection; otherwise, any group of the three groups of ratio results is not within the range of 95%~105% and fails the inspection.

[0016] S5. Physical size inspection Select the physical object that has passed the inspection in step S4, place a standard reference object next to the physical object and take a comparison picture by orthogonal projection, set the standard size of the standard reference object, the standard size is 1m, 1cm or 1mm, and obtain the corresponding size of the physical object in the comparison picture through calculation on the cloud server. Example

[0017] The system and method of the present application are combined with artificial intelligence models for intelligent data collection of electric power material inspection, storage and distribution. The target tracking technology used on the monitoring equipment mainly captures the image of the target through the monitoring camera, and then uses computer vision technology to detect and analyze the image features and motion models between frames to obtain the tracking results. The system and method of the present application greatly facilitates the comparison and inspection of the size of the physical object and the sample, and reduces the errors caused by manual inspection observation angles and human eye deviations. Reduce the number of manual workers, improve the efficiency of data entry and inspection, and effectively ensure the reliability of data entry and inspection.

[0018] The innovative technical points of this application are as follows: 1. Establishing the original sample database - Establishing the original sample database through sample pictures and sample videos. The original sample database is stored in the local cache. The original sample data is input through the local cache, which is convenient for comparing samples and real objects in the future. The local cache retrieves the corresponding samples, and uploads the photos of the real objects for on-site inspection through the cloud server for comparison. The artificial intelligence model comparison reduces the human operation error. On-site inspectors can compare samples and real objects by taking photos.

[0019] 2. Comparison and inspection of the aspect ratio of multiple views of samples and real objects - There are many types of power materials, with huge differences in material size and specifications. The overall comparison of photos will cause errors in the processing model or increase the load. The overall comparison of photos requires many issues such as the shooting angle, shooting distance, the ratio of the photographed object to the picture frame, and whether the shooting is centered. Detailed shooting quality requirements and detailed technical standards are required for the photos. If there is a problem with the photo collection, it will directly lead to the inability to complete the subsequent inspection. This application method greatly reduces the shooting quality requirements. It only requires orthographic projection shooting. There are no requirements for position, proportion, and centering. The aspect ratio is compared based on the principle of equal proportion, and the three-view comparison is achieved through the left and right view groups, the pitch view group, and the front and back view groups. Take into account both inspection efficiency and accuracy.

[0020] 3. Accurate identification of physical objects of the same type - many materials will be scaled proportionally according to the usage scenario, such as cables of the same cross-sectional diameter but different lengths or cables of the same length but different cross-sectional diameters. In the prior art, they can only be distinguished by personnel marking and measurement methods. First, the physical object is checked through steps S1 to S3 of the application method of this case, and multi-angle views are used to assist in confirmation and inspection through step S4. Step S5 provides standard reference objects to distinguish products of the same type but different sizes, and combines with cloud server calculations to obtain the corresponding sizes of the physical objects in the comparison pictures to achieve specific size measurement. No measuring tools are required on site. On-site personnel only need to select standard sizes according to the size of the physical object, and select one of 1m, 1cm or 1mm for size measurement. The 1m standard size is selected for physical size at the meter level, and the 1cm standard size is selected for physical size at the centimeter level to improve measurement accuracy.

Claims

1. The power material inspection, storage and distribution data acquisition and processing system based on artificial intelligence is characterized by Includes original data, verification data, local cache, cloud server and shooting equipment; The original data includes sample pictures and sample videos of power materials; The inspection data includes physical pictures and videos of power materials; Photographing equipment, which is used for photographing original data and inspection data; A local cache, which is used to receive the original data for local storage; The cloud server is used to query the original data in the local cache archive and upload the verification data for comparison with the original data; The original data is converted into sample pictures and sample videos through the shooting equipment and uploaded to the local cache to form the sample original database; the inspection data is converted into physical pictures and physical videos through the shooting equipment, which are compared with the sample original database through the cloud server, and the cloud server loads the target tracking algorithm model.

2. According to claim 1, the artificial intelligence-based power material inspection, storage and distribution data acquisition and processing system is characterized in that It also includes standard reference data, which includes a comparison picture of the actual power materials and the standard reference objects, and the actual power materials and the standard reference objects are located on the same picture.

3. According to claim 2, the artificial intelligence-based power material inspection, storage and distribution data acquisition and processing system is characterized in that The standard reference data is used to form a comparison picture through a shooting device. The standard reference object sets a standard size, which is 1m, 1cm or 1mm. The corresponding size of the actual object in the comparison picture is obtained through calculation by a cloud server.

4. The method for using the artificial intelligence-based power material inspection, storage and distribution data acquisition and processing system according to any one of claims 1 to 3 is characterized in that The steps include: S1. Raw data extraction Select a single-frame screenshot of a video of a power material sample or a sample picture; calculate the aspect ratio value A of the sample edge through the target tracking algorithm model; S2. Inspection data extraction Use a single-frame video screenshot or sample picture of the actual power material; calculate the aspect ratio value B of the edge of the inspection object through the target tracking algorithm model; S3. Comparison of the length-to-width ratio between the sample and the actual object Calculate the ratio = A / B; if the ratio result is within the range of 95%~105%, the test is passed; otherwise, if the ratio result is not within the range of 95%~105%, the test is failed.

5. The method for using the artificial intelligence-based power material inspection, storage and distribution data acquisition and processing system according to claim 4 is characterized in that The following steps are also included: S4. View direction replacement check The single-frame screenshots or sample images of the video are taken using orthogonal projection; they are divided into left and right view groups, pitch view groups, and front and back view groups according to the orthogonal projection direction; each view group uses a single-frame screenshot or sample image of the video; After replacing the view, the aspect ratio A′ of the sample edge is calculated through the target tracking algorithm model, and the aspect ratio B′ of the edge of the actual object is calculated through the target tracking algorithm model; S3 is used to calculate the ratio = A′ / B′; the three groups of ratio results are all within the range of 95%~105% and pass the inspection; otherwise, any of the three groups of ratio results is not within the range of 95%~105% and fails the inspection.

6. The method for using the artificial intelligence-based power material inspection, storage and distribution data acquisition and processing system according to claim 5 is characterized in that The following steps are also included: S5. Physical size inspection Select the physical object that has passed the inspection in step S4, place a standard reference object next to the physical object and take a comparison picture by orthogonal projection, set the standard size of the standard reference object, the standard size is 1m, 1cm or 1mm, and obtain the corresponding size of the physical object in the comparison picture through calculation on the cloud server.

Images