Return sorting management system based on digital twinning

Abstract

The invention, which relates to the technical field of sorting management, discloses a digital-twinning-based returning sorting management system comprising a sorting detection unit, a feature acquisition unit, a twinning model library, a data analysis unit and a result display unit. Through comparison detection and classification analysis with a reference electric energy meter, various electric energy metering parameters are comprehensively considered; accurately determining whether the withdrawn electric energy meter has a fault or not; secondary classification analysis is carried out from the appearance and the weight by combining a twin model library, whether damage exists or not is judged, the fault judgment calculation step is rigorous, calculation of metering parameter errors, error analysis values and fault evaluation values is included, and comparison with preset threshold values is carried out; a series of processing such as a picture similarity algorithm is applied to appearance damage coefficient calculation, and weight damage coefficient calculation is simple and effective; the result display unit provides a clear decision basis for detection personnel, improves the sorting efficiency, saves manpower, time and cost, improves the resource utilization rate, and maintains and reuses the repairable electric energy meter.

Description

Technical Field

[0001] The present invention relates to the technical field of sorting management, and in particular to a return sorting management system based on digital twins. Background Art

[0002] With the continuous development of the power industry, the number of electric energy meters in use is increasing. During the life cycle of electric energy meters, electric energy meters may be returned due to various reasons. The returned electric energy meters need to be sorted and managed to determine which electric energy meters can be put into use again after repair and which electric energy meters need to be scrapped. Traditional return sorting management mainly relies on manual inspection and experience judgment, which has the problems of low efficiency and low accuracy.

[0003] At the same time, with the continuous development of digital twin technology, its application in various fields is becoming more and more extensive. Digital twin technology can achieve real-time monitoring, analysis and optimization of physical entities by establishing a corresponding relationship between virtual models and physical entities. Applying digital twin technology to the return sorting management system can improve the efficiency and accuracy of sorting management, reduce costs, and provide strong support for the sustainable development of the power industry.

[0004] Therefore, a return sorting management system based on digital twins came into being. The system realizes accurate classification and judgment of returned electricity meters through comparative detection and feature collection of returned electricity meters, combined with functions such as data analysis and result display, providing a scientific basis for the subsequent processing of returned electricity meters. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a return sorting management system based on digital twins.

[0006] The return sorting management system based on digital twin includes:

[0007] The sorting and testing unit is used to select an electric energy meter with good performance as a reference electric energy meter, then compare and test each returned electric energy meter with the reference electric energy meter, and collect electric energy metering parameters corresponding to each returned electric energy meter and the reference electric energy meter;

[0008] A feature collection unit, used to collect feature data of each returned electric energy meter;

[0009] Twin model library, used to store factory feature information of all types of electric energy meters;

[0010] A data analysis unit is used to classify and analyze the electric energy metering parameters obtained by the sorting and detection unit, and determine whether the returned electric energy meter is faulty based on the classification and analysis results, and is also used to combine the feature data of the returned electric energy meter with the twin model library for secondary classification and analysis, and determine whether the returned electric energy meter is damaged based on the secondary classification and analysis results;

[0011] The result display unit is used to display the determination result of the data analysis unit to relevant testing personnel.

[0012] As a further solution of the present invention: the comparative detection method is as follows:

[0013] Under the same load conditions, the reference electric energy meter and the returned electric energy meter are connected respectively, and the electric energy measurement parameters of the two electric energy meters within the specified period are recorded.

[0014] As a further solution of the present invention: the electric energy metering parameters include but are not limited to attribute parameters corresponding to electric quantity, power, voltage and current.

[0015] As a further solution of the present invention: the characteristic data includes the model, current appearance photo, and actual weight of the returned electric energy meter; the factory characteristic information includes factory photos of each surface of the electric energy meter and its factory weight.

[0016] As a further solution of the present invention: the classification analysis method is as follows:

[0017] Step 1.1, select a returned electric energy meter as an example;

[0018] Extract the electric energy measurement parameters corresponding to the returned electric energy meter and mark them as Ti; and extract the electric energy measurement parameters corresponding to the reference electric energy meter and mark them as Di;

[0019] Wherein, i=1, 2, ... n, n represents the number of attribute parameters in the electric energy metering parameters, and i represents the number of attribute parameters;

[0020] Step 1.2, let the value of i be 1, 2, ... n;

[0021] And through Ci=|Di-Ti|, the measurement parameter error Ci between the returned electric energy meter and the reference electric energy meter is calculated;

[0022] Step 1.3, calculate the absolute value of the difference between the measurement parameter error Ci between the returned electric energy meter and the reference electric energy meter and the preset allowable error threshold ui, and then record the result as the error analysis value Wi;

[0023] Wherein, ui represents the allowable error threshold preset according to the attribute parameter number;

[0024] Step 1.4, calculate the fault assessment value P of the returned electric energy meter through P=W1×α1+W2×α2+…Wn×αn;

[0025] Among them, αi is the corresponding preset weight value;

[0026] Step 1.5, calculate the fault assessment value of each returned electric energy meter according to the method from Step 1.1 to Step 1.4;

[0027] Then, the fault assessment value of each returned electric energy meter is compared with a preset fault assessment threshold, and the returned electric energy meter exceeding the fault assessment threshold is determined as a faulty electric energy meter; otherwise, it is determined as a normal electric energy meter.

[0028] As a further solution of the present invention: the secondary classification analysis method is as follows:

[0029] Step V1. Select a returned electric energy meter, and obtain the factory photos and factory weight of the electric energy meter of the same model from the twin model library according to the model of the returned electric energy meter;

[0030] Step V2, then compare the current appearance photo and actual weight of the returned electric energy meter with the factory photo and factory weight of the same model electric energy meter, and obtain the appearance damage coefficient and weight damage coefficient of the returned electric energy meter;

[0031] Step V3, multiply the weight damage coefficient and the appearance damage coefficient by the corresponding preset weight coefficient, then add the results to obtain the damage assessment coefficient;

[0032] Step V4, compare the damage assessment coefficient with the preset damage assessment threshold. When the damage assessment coefficient exceeds the damage assessment threshold, the returned electricity meter is determined to be a damaged electricity meter. Otherwise, the returned electricity meter is determined to be a normal electricity meter.

[0033] As a further solution of the present invention: the comparison processing method in Step V2 is as follows:

[0034] Step 0.1, a current appearance photo, combined with the image similarity algorithm, and then extracting a factory photo with the highest similarity to the current appearance photo from all factory photos;

[0035] Step 0.2, adjust the main body size and resolution of the extracted factory photo and the current appearance photo to be consistent, and the total number of pixels of the main body of the electric energy meter in the factory photo and the current appearance photo after adjustment is consistent;

[0036] Step 0.3, convert the factory photos and current appearance photos corresponding to the electric energy meter into grayscale images;

[0037] Step 0.4, select the pixel values ​​of all pixel coordinate points from the two grayscale images, then select the pixel value of the same pixel coordinate point from the two grayscale images, and calculate the pixel difference of the same pixel coordinate point;

[0038] Step 0.5, obtain the pixel difference of all the same pixel coordinate points in the two grayscale images, and mark it as Xj, j = 1, 2, ... m, m represents the total number of pixels of the subject in the photo;

[0039] Step 0.6, calculate the average value of the pixel differences corresponding to all the same pixel coordinate points;

[0040] Then, the standard deviation of the pixel differences corresponding to all the same pixel coordinate points is calculated based on the average value;

[0041] Step 0.7, calculate the sum of the mean value and the standard deviation, then compare the pixel difference values ​​of each pixel coordinate point with the sum of the mean value and the standard deviation one by one, and obtain the number of pixel differences of the same pixel coordinate point that exceed the sum of the mean value and the standard deviation, and then mark it as the abnormal pixel quantity v;

[0042] Step 0.8, calculate the proportion of v in m and record it as the appearance damage coefficient;

[0043] Step 0.9. Extract the factory weight and actual weight of the same model of electricity meter, calculate the absolute value of the difference between them, and then divide the result by the factory weight, and then mark the result as the weight loss coefficient.

[0044] As a further solution of the present invention: in Step 0.7, each pixel point is arranged in a 9-square grid;

[0045] When the pixel difference of the same pixel coordinate point exceeds the sum of the mean and the standard deviation, and the pixel difference of its adjacent 8 pixel coordinates does not exceed the sum of the mean and the standard deviation, the v value is reduced by one; otherwise, the v value remains unchanged.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] The present invention can accurately determine whether the returned energy meter is faulty by comparing the returned energy meter with the reference energy meter and performing classification analysis. A comprehensive evaluation is performed using multiple energy metering parameters, taking into account multiple attribute parameters such as electricity, power, voltage, and current, thereby improving the accuracy of fault determination.

[0048] The present invention combines the twin model library to conduct secondary classification analysis on the returned energy meter, and can determine whether the returned energy meter is damaged. The evaluation is conducted from two aspects: appearance and weight. By comparing with the factory photos and weight, the physical state of the energy meter is comprehensively considered.

[0049] The present invention adopts detailed calculation steps for electric energy meter fault determination. By calculating the metering parameter error, error analysis value and fault assessment value and comparing them with the preset threshold value, the determination process is scientific and objective. At the same time, the preset weight value can be adjusted according to the actual situation, which improves the flexibility of the method.

[0050] The present invention uses a series of scientific processing methods such as image similarity algorithm, image adjustment, grayscale conversion, pixel difference calculation, etc. to calculate the appearance damage coefficient, which can accurately reflect the degree of change in the appearance of the electric energy meter. The calculation of the weight damage coefficient is simple and direct, and by comparing it with the factory weight, it effectively measures the weight change of the electric energy meter.

[0051] In the present invention, the result display unit displays the determination result of the data analysis unit to relevant inspection personnel, providing them with a clear basis for decision-making, and facilitating rapid sorting and processing of returned electric energy meters.

[0052] The present invention has a high degree of automation of the entire system, which can greatly improve the sorting efficiency of returned electric energy meters and save manpower and time costs.

[0053] The present invention can repair and reuse the repairable electric energy meters and reasonably dispose of the electric energy meters that cannot be repaired by accurately determining the fault and damage status of the returned electric energy meters, thereby improving the utilization rate of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a system block diagram of the present invention;

[0055] Figure 2 The process diagram of the data analysis unit of the present invention is shown in FIG. Figure 1 ;

[0056] Figure 3 The process diagram of the data analysis unit of the present invention is shown in FIG. Figure 2 ;

[0057] Figure 4 Schematic diagram of the comparison process in the data analysis unit of the present invention. DETAILED DESCRIPTION

[0058] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] Embodiment 1

[0060] See also Figure 1 and Figure 2 As shown, the present invention is a return sorting management system based on digital twin, comprising:

[0061] The sorting and testing unit is used to select an electric energy meter with good performance as a reference electric energy meter, then compare and test each returned electric energy meter with the reference electric energy meter, and collect electric energy metering parameters corresponding to each returned electric energy meter and the reference electric energy meter;

[0062] The comparative detection method is as follows:

[0063] Under the same load conditions, the reference electric energy meter and the returned electric energy meter are connected respectively, and the electric energy measurement parameters of the two electric energy meters within the specified period are recorded;

[0064] Electric energy metering parameters include but are not limited to attribute parameters corresponding to electricity, power, voltage, and current;

[0065] A data analysis unit, used for classifying and analyzing electric energy metering parameters;

[0066] The classification analysis method is as follows:

[0067] Step 1.1, parameter extraction and marking

[0068] Take a returned electric energy meter as an example;

[0069] Extract the electric energy measurement parameters corresponding to the returned electric energy meter and mark them as Ti; and extract the electric energy measurement parameters corresponding to the reference electric energy meter and mark them as Di;

[0070] Wherein, i=1, 2, ... n, n represents the number of attribute parameters in the electric energy metering parameters, and i represents the number of attribute parameters;

[0071] Step 1.2. Calculation of measurement parameter errors

[0072] Let the value of i be 1, 2, ... n;

[0073] And through Ci=|Di-Ti|, the measurement parameter error Ci between the returned electric energy meter and the reference electric energy meter is calculated;

[0074] Step 1.3. Calculation of error analysis value

[0075] The absolute value of the difference between the measurement parameter error Ci between the returned electric energy meter and the reference electric energy meter and the preset allowable error threshold ui is calculated, and then the result is recorded as the error analysis value Wi;

[0076] Wherein, ui represents the allowable error threshold preset according to the attribute parameter number;

[0077] Step 1.4. Calculation of fault assessment value

[0078] The fault assessment value P of the returned electric energy meter is calculated by P=W1×α1+W2×α2+…Wn×αn;

[0079] Among them, αi is the corresponding preset weight value;

[0080] Step 1.5, Fault determination

[0081] Then the fault assessment value of the returned electric energy meter is compared with the preset fault assessment threshold Py:

[0082] When P>Py, the returned electric energy meter is determined to be a faulty electric energy meter;

[0083] When P≤Py, the returned electric energy meter is determined to be a normal electric energy meter;

[0084] Finally, according to the method from Step 1.1 to Step 1.5, determine whether each returned electric energy meter is a faulty electric energy meter;

[0085] The result display unit is used to display the determination result of the data analysis unit to relevant testing personnel.

[0086] The present embodiment provides a return sorting management system based on digital twins. By selecting a reference electric energy meter with good performance for comparative testing with the returned electric energy meter, the electric energy metering parameters of the returned electric energy meter can be accurately collected to provide reliable data for subsequent analysis. The electric energy metering parameters are classified and analyzed, and by calculating the metering parameter error, fault assessment value, etc., it can accurately determine whether the returned electric energy meter is a faulty electric energy meter, providing a scientific basis for sorting the returned electric energy meter and improving the sorting efficiency and accuracy. The judgment results are displayed to relevant testing personnel, so that the testing personnel can timely understand the status of the returned electric energy meter and perform subsequent processing.

[0087] Embodiment 2

[0088] See also Figure 1 , Figure 3 , Figure 4 As shown, as the second embodiment of the present invention, when the present application is implemented, compared with the first embodiment, the technical solution of this embodiment is different from that of the first embodiment only in that the present embodiment further includes:

[0089] A feature collection unit, used to collect feature data of each returned electric energy meter;

[0090] The characteristic data includes the model, current appearance photo, and actual weight of the returned electric energy meter;

[0091] Twin model library, used to store factory photos of all surfaces of all models of electric energy meters and their factory weights;

[0092] The data analysis unit is also used to combine the characteristic data of the returned electric energy meter with the twin model library for secondary classification analysis:

[0093] The secondary classification analysis method is as follows:

[0094] StepV1. Get factory information

[0095] Select a returned energy meter.

[0096] According to the model of the returned electric energy meter, the factory photos and factory weight of the electric energy meter of the same model are obtained from the twin model library;

[0097] StepV2, compare the coefficients

[0098] Then, the current appearance photo and actual weight of the returned electric energy meter are compared with the factory photo and factory weight of the electric energy meter of the same model, and the appearance damage coefficient and weight damage coefficient of the returned electric energy meter are obtained;

[0099] The comparison is processed as follows:

[0100] Step 0.1. Extract similar photos

[0101] A current appearance photo, combined with a picture similarity algorithm, and then extracting a factory photo with the highest similarity to the current appearance photo from all factory photos;

[0102] Step 0.2. Adjust photo parameters

[0103] The main body size and resolution of the extracted factory photo and the current appearance photo are adjusted to be consistent, and the total number of pixels of the main body of the electric energy meter in the factory photo and the current appearance photo after adjustment is consistent;

[0104] Step 0.3, Convert to grayscale image

[0105] Convert the factory photos and current appearance photos corresponding to the electric energy meter into grayscale images;

[0106] Step 0.4. Calculate pixel difference

[0107] Select to obtain pixel values ​​of all pixel coordinate points from two grayscale images respectively, then select the pixel value of the same pixel coordinate point from the two grayscale images, and calculate the pixel difference value of the same pixel coordinate point;

[0108] Step 0.5: Mark pixel differences

[0109] Get the pixel difference of all the same pixel coordinates in the two grayscale images and mark them as Xj, j = 1, 2, ... m, m represents the total number of pixels of the subject in the photo;

[0110] Step 0.6. Calculate the mean and standard deviation

[0111] pass: , calculate the average value XP of the pixel differences corresponding to all the same pixel coordinate points;

[0112] pass: , calculate the standard deviation XB of the pixel differences corresponding to all the same pixel coordinate points;

[0113] Step 0.7. Determine the number of abnormal pixels

[0114] By XH=XP+XB, the sum of the mean value and the standard deviation XH is calculated, and then the pixel difference values ​​of each pixel coordinate point with the same pixel coordinate point are compared with XH one by one, and the number of pixel differences of the same pixel coordinate point exceeding XH is obtained, and then it is marked as the abnormal pixel quantity v;

[0115] Step 0.8. Calculate the appearance damage coefficient

[0116] By K0=v / m, calculate the proportion of v in m, K0, and record K0 as the appearance damage coefficient;

[0117] Step 0.9. Calculate weight loss coefficient

[0118] Extract the factory weight and actual weight of the same model of electric energy meter;

[0119] Then, the weight loss coefficient F0 is calculated through F0= (E0-E1) / E0;

[0120] In the formula, E0 is the factory weight of the same model of electric energy meter, and E1 is the actual weight of the same model of electric energy meter;

[0121] Step V3. Calculate the damage assessment coefficient

[0122] The damage assessment coefficient H of the returned electric energy meter is calculated by H=F0×γ1+K0×γ2;

[0123] In the formula, F0 is the weight damage coefficient, K0 is the appearance damage coefficient, γ1 and γ2 are the corresponding preset weight coefficients;

[0124] Step V4, damage determination

[0125] Compare the damage assessment coefficient with the preset damage assessment threshold Hy:

[0126] When H>Hy, the returned electric energy meter is determined to be a damaged electric energy meter;

[0127] When H≤Hy, the returned electric energy meter is determined to be a normal electric energy meter.

[0128] In this embodiment, this embodiment can directly evaluate and judge all returned electric energy meters, or, based on the first embodiment, perform a secondary evaluation and judgement on all normal electric energy meters obtained in the first embodiment.

[0129] Based on the first embodiment, this embodiment adds a feature acquisition unit and a twin model library, which can collect feature data of returned electricity meters, and perform secondary classification analysis in combination with the twin model library, thereby improving the comprehensiveness and accuracy of the judgment on the status of returned electricity meters; by comparing the current appearance photos and actual weight of the returned electricity meters with the factory photos and factory weights of the electricity meters of the same model, the appearance damage coefficient and weight damage coefficient are obtained, and then the damage assessment coefficient is calculated, which can accurately determine whether the returned electricity meter is a damaged electricity meter, and provide support for the refined sorting of returned electricity meters.

[0130] Embodiment 3

[0131] As the third embodiment of the present invention, when the present application is implemented, compared with the first and second embodiments, the technical solution of the present embodiment is to combine the solutions of the first and second embodiments, and the difference between the technical solution of the present embodiment and the first and second embodiments is that: on the basis of the second embodiment, the present embodiment removes the abnormal value of the abnormal pixel amount;

[0132] The method is as follows:

[0133] Among them, each pixel is arranged in the form of a 9-square grid;

[0134] When the pixel difference of the same pixel coordinate point exceeds the sum of the mean and the standard deviation, and the pixel difference of its adjacent 8 pixel coordinates does not exceed the sum of the mean and the standard deviation, the v value is reduced by one; otherwise, the v value remains unchanged.

[0135] Based on the second embodiment, this embodiment removes abnormal values ​​from the abnormal pixel quantity, thereby improving the accuracy of the calculation of the appearance damage coefficient, and further improving the accuracy of the damage assessment coefficient, making the judgment of the status of the returned electric energy meter more reliable.

[0136] Embodiment 4

[0137] As the fourth embodiment of the present invention, when the present application is specifically implemented, compared with the first, second and third embodiments, the technical solution of this embodiment is to combine and implement the solutions of the above-mentioned first, second and third embodiments.

[0138] This embodiment combines the solutions of Embodiment 1, Embodiment 2 and Embodiment 3, and can comprehensively and accurately sort and manage returned electric energy meters, thereby improving sorting efficiency and accuracy, and providing a scientific basis for the reuse or treatment of returned electric energy meters.

[0139] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0140] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. The return sorting management system based on digital twin is characterized by: include: The sorting and testing unit is used to select an electric energy meter with good performance as a reference electric energy meter, and then compare and test each returned electric energy meter with the reference electric energy meter, and collect electric energy metering parameters corresponding to each returned electric energy meter and the reference electric energy meter obtained by the comparison test; A feature collection unit, used to collect feature data of each returned electric energy meter; Twin model library, used to store factory feature information of all types of electric energy meters; The data analysis unit is used to classify and analyze the electric energy metering parameters obtained by the sorting and detection unit, and determine whether the returned electric energy meter is faulty based on the classification and analysis results. It is also used to combine the feature data of the returned electric energy meter with the twin model library for secondary classification analysis, and determine whether the returned electric energy meter is damaged based on the secondary classification analysis results.

2. The return sorting management system based on digital twin according to claim 1 is characterized in that: The classification analysis method is as follows: Step 1.1, select a returned electric energy meter as an example; Extract the electric energy measurement parameters corresponding to the returned electric energy meter and mark them as Ti; and extract the electric energy measurement parameters corresponding to the reference electric energy meter and mark them as Di; Wherein, i=1, 2, ... n, n represents the number of attribute parameters in the electric energy metering parameters, and i represents the number of attribute parameters; Step 1.2, calculate the absolute value of the difference between the corresponding electric energy metering parameters of the returned electric energy meter and the reference electric energy meter, and mark it as the metering parameter error Ci; Step 1.3, calculate the absolute value of the difference between the metering parameter error Ci between the returned electric energy meter and the reference electric energy meter and the preset allowable error threshold ui, and then mark the result as the error analysis value Wi; Wherein, ui represents the allowable error threshold preset according to the attribute parameter number; Step 1.4, calculate the fault assessment value P of the returned electric energy meter through P=W1×α1+W2×α2+…Wn×αn; Among them, αi is the corresponding preset weight value; Step 1.5, calculate the fault assessment value of each returned electric energy meter according to the method from Step 1.1 to Step 1.4; Then, the fault assessment value of each returned electric energy meter is compared with a preset fault assessment threshold, and the returned electric energy meter exceeding the fault assessment threshold is determined as a faulty electric energy meter; otherwise, it is determined as a normal electric energy meter.

3. The return sorting management system based on digital twin according to claim 1 is characterized in that: The secondary classification analysis method is as follows: Step V1. Select a returned electric energy meter, and obtain the factory photos and factory weight of the electric energy meter of the same model from the twin model library according to the model of the returned electric energy meter; Step V2, then compare the current appearance photo and actual weight of the returned electric energy meter with the factory photo and factory weight of the same model electric energy meter, and obtain the appearance damage coefficient and weight damage coefficient of the returned electric energy meter; Step V3, multiply the weight damage coefficient and the appearance damage coefficient by the corresponding preset weight coefficient, then add the results to obtain the damage assessment coefficient; Step V4, compare the damage assessment coefficient with the preset damage assessment threshold. When the damage assessment coefficient exceeds the damage assessment threshold, the returned electricity meter is determined to be a damaged electricity meter. Otherwise, the returned electricity meter is determined to be a normal electricity meter.

4. The return sorting management system based on digital twin according to claim 3 is characterized in that: The comparison process in StepV2 is as follows: Step 0.1, a current appearance photo, combined with the image similarity algorithm, and then extracting a factory photo with the highest similarity to the current appearance photo from all factory photos; Step 0.2, adjust the subject size and resolution of the extracted factory photo and the current appearance photo to be consistent; Step 0.3, convert the factory photos and current appearance photos corresponding to the electric energy meter into grayscale images; Step 0.4, select the pixel values ​​of all pixel coordinate points from the two grayscale images, then select the pixel value of the same pixel coordinate point from the two grayscale images, and calculate the pixel difference of the same pixel coordinate point; Step 0.5, obtain the pixel difference of all the same pixel coordinate points in the two grayscale images; Step 0.6, calculate the average value of the pixel differences corresponding to all the same pixel coordinate points; Then, the standard deviation of the pixel differences corresponding to all the same pixel coordinate points is calculated based on the average value; Step 0.7, calculate the sum of the mean value and the standard deviation, then compare the pixel difference values ​​of each pixel coordinate point with the sum of the mean value and the standard deviation one by one, and obtain the number of pixel differences of the same pixel coordinate point that exceed the sum of the mean value and the standard deviation, and then mark it as the abnormal pixel quantity v; Step 0.8, calculate the proportion of v in m and record it as the appearance damage coefficient; Among them, m represents the total number of pixels of the subject in the photo; Step 0.

9. Extract the factory weight and actual weight of the same model of electricity meter, calculate the absolute value of the difference between them, and then divide the result by the factory weight, and then mark the result as the weight loss coefficient.

5. The return sorting management system based on digital twin according to claim 1 is characterized in that: The comparative detection method is as follows: Under the same load conditions, the reference electric energy meter and the returned electric energy meter are connected respectively, and the electric energy measurement parameters of the two electric energy meters within the specified period are recorded.

6. The return sorting management system based on digital twin according to claim 1 is characterized in that: Electric energy metering parameters include but are not limited to attribute parameters corresponding to electricity, power, voltage, and current.

7. The return sorting management system based on digital twin according to claim 1 is characterized in that: The characteristic data includes the model, current appearance photo, and actual weight of the returned electric energy meter; The factory characteristic information includes factory photos of each surface of the electric energy meter and its factory weight.

8. The return sorting management system based on digital twin according to claim 4 is characterized in that: The factory photo and the current appearance photo after adjustment in Step 0.2 have the same total number of pixels on the main body of the energy meter.

9. The return sorting management system based on digital twin according to claim 4 is characterized in that: In Step 0.7, each pixel is arranged in a 9-square grid; When the pixel difference of the same pixel coordinate point exceeds the sum of the mean and the standard deviation, and the pixel difference of its adjacent 8 pixel coordinates does not exceed the sum of the mean and the standard deviation, the v value is reduced by one; otherwise, the v value remains unchanged.

10. The return sorting management system based on digital twin according to claim 1, characterized in that: Also includes: The result display unit is used to display the determination result of the data analysis unit to relevant testing personnel.

Images