A low-voltage line fault detection and location method

Abstract

The present invention discloses a method for detecting and locating low-voltage line faults, comprising the following steps: S1, the collected current and voltage data are sent to the data scheduling center through the intranet after preprocessing; S2, the positioning data are encrypted and sent to the data scheduling center through the extranet; S3, the data scheduling center fuses the data sent through the intranet and the extranet and establishes a joint data model; S4, after the fused data are verified by the joint data model, they are sent to a number of mobile terminals with the highest matching degree. The current and voltage data and the positioning data are collected and transmitted respectively through the intranet and the extranet in an encrypted manner, and a joint data model is established according to the data sent through the intranet and the extranet; the method of the present invention can effectively monitor the low-voltage line, and through the comprehensive transmission method of encrypting the intranet and the extranet, while ensuring the integrity and error-free transmission, the data security is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of low-voltage line detection, and in particular to a method for detecting and locating low-voltage line faults. Background Art

[0002] With the continuous development of power technology, the number of power users is increasing. People's requirements for the stability and reliability of power supply are constantly improving, and higher requirements are also put forward for the speed and accuracy of power grid fault repair. Low-voltage lines are easily affected by factors such as climate change, insulation aging, and natural damage, with a high failure rate and great difficulty in fault inspection. When a low-voltage line fails, it will have a certain impact on people's production and life. Therefore, it is necessary to effectively detect low-voltage lines, be able to discover faults in a timely manner when they occur, and perform quick repair. Sometimes, there are errors in the fault information, resulting in misjudgment.

[0003] Disclosed in the Chinese patent document "A Fault Location System and Method for a Distribution Line", with the publication number CN113759207A, a fault location system and method for a distribution line. The fault location system includes a plurality of fault location devices, which are respectively installed on the low-voltage side of the distribution transformer at the end of the branch line to monitor the live condition of the distribution area. The plurality of fault location devices actively transmit the monitoring data to the remote monitoring background through a wireless network. The remote monitoring background receives the abnormal data monitored by the distributed fault location devices, performs distributed fault judgment, and actively alarms the power outage range and fault location. However, the specific location and data processing methods disclosed in CN113759207A are relatively simple. Summary of the Invention

[0004] The present invention solves the problem that existing low-voltage lines lack effective detection and location means, and proposes a method for detecting and locating low-voltage line faults. It collects current and voltage data as well as location data, and transmits them through encryption methods on the intranet and the extranet respectively. According to the data sent through the intranet and the extranet, a joint data model is established. The joint data model verifies the fault data, and the complete fault data is sent to the mobile terminal for dispatching and repair. The method of the present invention can effectively monitor low-voltage lines, and through the comprehensive transmission method of encryption on the intranet and the extranet, while ensuring the integrity and error-free transmission, it ensures the security of the data.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A method for detecting and locating low-voltage line faults, including the following steps:

[0006] S1, the collected current and voltage data are sent to the data dispatching center through the intranet after preprocessing;

[0007] S2. Encrypt and send the positioning data to the data scheduling center via the external network;

[0008] S3. The data scheduling center fuses the data sent via the internal network and the external network and establishes a joint data model;

[0009] S4. After the fused data is verified by the joint data model, it is sent to several mobile terminals with the highest matching degree.

[0010] In the present invention, the collected current and voltage data are waveform signals, and this data is directly sent through the internal network of the power grid. Due to the security of the internal network itself, there is no need to encrypt the current and voltage data; while the positioning data is collected by the positioning module in the fault collection point, and the positioning module itself uses the external network. Therefore, in the present invention, the positioning data transmission directly uses the external network for transmission. However, due to the insecurity of the external network, which does not conform to the data security of the power grid, the positioning data is encrypted; the current and voltage data and the positioning data will then be fused at the data scheduling center and correspond one by one, and the established joint data model is used for the verification of faults; finally, the fused data will be transmitted to the mobile terminal, and the mobile terminal will perform dispatching and maintenance operations after receiving the instruction. The method of the present invention can achieve good detection and positioning effects.

[0011] Preferably, the step S1 includes the following steps:

[0012] S11. Set fault collection points every K kilometers on the low-voltage line, and adjacent fault collection points jointly detect the current and voltage data on the unit line;

[0013] S12. The fault collection point automatically inspects and checks the collected current and voltage data, and extracts the current and voltage data that are not within the preset normal threshold range;

[0014] S13. Perform digital-to-analog conversion on the current and voltage data and transmit the data through the internal network of the power grid, and finally transmit it to the internal network data processing unit of the data scheduling center. In the present invention, the distance K between the fault collection points can be adjusted according to the actual situation. Specifically, in implementation, the fault collection points can also directly skip multiple intermediate fault collection points. At this time, the distance between the fault collection points can be 2K, 3K, or even nK; the fault collection points have a threshold judgment function. Every once in a while, the fault collection points will automatically perform the threshold judgment on the current and voltage data collected during this period, automatically eliminate the data within the threshold range, and save and transmit the data exceeding the threshold range through the internal network.

[0015] Preferably, the step S2 includes the following steps:

[0016] S21. Immediately obtain the positioning data at adjacent fault acquisition points, and at the same time retrieve the positioning data corresponding to the current and voltage data that are not within the preset normal threshold range, and preliminarily determine the unit line with line anomalies.

[0017] S22. Let the position data of adjacent fault acquisition points with line anomalies be represented as Ln and Ln+1 respectively. There is a line positioning initial coordinate set. According to the linear distance and angular relationship between each fault acquisition point, gradually calculate the three-dimensional coordinates of Ln and Ln+1; S23. Encrypt the retrieved positioning data and transmit it from the external network of the power grid to the data dispatching center. In the present invention, the positioning data corresponds to the current and voltage data. The Ln coordinate is specifically:

[0018] (L1 cos α1 ± L2 cos α2... ± Lncos αn, L1 sin α1 ± L2 sin α2... ± Lnsin αn, h)

[0019] The Ln+1 coordinate is specifically:

[0020] (L1 cos α1 ± L2 cos α2... ± L(n+1)cos α(n+1), L1 sin α1 ± L2 sin α2... ± L(n+1)sin α(n+1), h)

[0021] Among them, L1 is the distance between the first fault acquisition point and the line positioning initial coordinate, Ln is the distance between the nth fault acquisition point and the previous fault acquisition point, αn is the angle between the nth fault acquisition point and the previous acquisition point, ± specifically uses + or - depending on the specific position of the nth fault acquisition point, h is the height of the fault acquisition point, and h is usually a fixed value.

[0022] Preferably, the data encryption includes the following steps:

[0023] S231. Perform dimensionless processing on the three-dimensional coordinate data of the positioning data. Under the condition of unified data units, through proportional amplification or reduction, convert the three-dimensional coordinate data into data between 0-255. The specific multiple of proportional amplification or reduction is sent to the data dispatching center through the internal network;

[0024] S232. Successively use the converted three-dimensional coordinates to represent the RGB values respectively to form several color images;

[0025] S233. A number of color images form an encrypted transmission image of X*Y in a specific order. X is the number of rows of the transmission image, and Y is the number of columns of the transmission image. The encrypted transmission image is encrypted data and is finally sent to the external network data processing unit of the data scheduling center. In the present invention, first, the three-dimensional coordinate data is processed. The three-dimensional coordinate data is represented by data from 0 to 255, and this data is represented as RGB values, thereby obtaining a color image. Each color image represents a three-dimensional coordinate. During transmission, the transmission image composed of color images is transmitted, and the arrangement of the transmission image has a specific order. The encrypted transmission image is directly sent to the data scheduling center, and in the external network data processing unit, it is decrypted in combination with the magnification or reduction factor sent through the internal network.

[0026] Preferably, the specific order is specifically related to the fault type. If the fault type is a short circuit, a number of color images are arranged in ascending order of the number of rows of the transmission image, and in ascending order of the number of columns in the same row. If the fault type is an open circuit, a number of color images are arranged in ascending order of the number of columns of the transmission image, and in ascending order of the number of rows in the same column. In the present invention, the characteristic order is divided into two different arrangement methods according to the fault type. Specifically, it is arranged by row and arranged by column. When arranged by row, the first row is filled in ascending order first, and then the second row is filled in ascending order in the same way, and so on until the last row X is arranged. If the last row X is not filled, 0 is filled for complement. At the same time, when arranged by column, it is arranged in ascending order from the first column to the Y column. If the last column Y is not filled, 0 is filled for complement.

[0027] Preferably, the step S3 includes the following steps:

[0028] S31. Decrypt the transmission image into positioning data, fuse the positioning data with the current and voltage data in a one-to-one correspondence to form fusion data.

[0029] S32. In the fusion data, the time-domain transformation is used to process the current and voltage data, calculate the fundamental wave to higher harmonic phasors. The phasors include the amplitude, phase angle, frequency, active power, and reactive power of the current signal and voltage signal. Calculate the current and voltage characteristics according to the phasors. At the same time, extract the coordinate characteristics in the positioning data to form the mapping relationship between the current and voltage characteristics, coordinate characteristics, and fault type as the training sample.

[0030] S33. Based on the training samples, a joint data model is established by means of deep learning. In the present invention, for the decryption method of the encrypted transmission image, decryption is performed according to the magnification or reduction ratio and a specific order. In the present invention, the magnification or reduction ratio is transmitted separately from the positioning data transmission to ensure the security of the positioning data. In addition, for the established joint data model, it is essentially an inspection model, and the data sources of the training samples are all data obtained from historical collections and have all been verified.

[0031] Preferably, the step S4 includes the following steps:

[0032] S41. Input the latest fusion data sent to the data scheduling center and formed into the joint data model for verification, re-extract the current and voltage data and convert them into waveform signals;

[0033] S42. The waveform signal is fitted and compared with the simulated waveform of the joint data model. Specifically, the amplitude, phase angle, and frequency of the waveform signal are compared. If the similarity is less than 1%, the comparison is successful;

[0034] S43. After the comparison is successful, the data scheduling center sends the fusion data to several mobile terminals that are closest to the fault location and are not in the maintenance state. In the present invention, for the verification of the waveform signal, specifically, the waveform signal to be verified is fitted with the closest simulated waveform in the joint data model in the same coordinate. Through simulation calculation, it is compared whether the amplitude, phase angle, and frequency of the waveform signals of the two are all less than 1%. If they are all less than 1%, the comparison is successful. If any one of them is greater than or equal to 1%, the comparison fails.

[0035] Preferably, the fault collection points are communicatively connected to each other, the fault collection points are communicatively connected to the data scheduling center, and a positioning module based on the fault collection point distribution map is provided inside the fault collection points. In the present invention, a single fault collection point may be provided with multiple collection devices. The fault collection point distribution map in the positioning module inside the fault collection point is obtained by being sent down by the data scheduling center, and the positioning module performs positioning and positioning judgment according to the fault collection point distribution map.

[0036] The beneficial effects of the present invention are as follows: A low-voltage line fault detection and location method of the present invention collects current and voltage data as well as location data, and transmits them through encryption via the intranet and the extranet respectively. According to the data sent through the intranet and the extranet, a combined data model is established. The combined data model verifies the fault data, and the completed fault data is sent to the mobile terminal for scheduling and maintenance; the method of the present invention can effectively monitor the low-voltage line, and through the comprehensive transmission method of encryption via the intranet and the extranet, while ensuring the integrity and error-free transmission of the data, it ensures the security of the data; the established combined data model can perform fault verification, ensuring the accuracy of detection and preventing misjudgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flowchart of a low-voltage line fault detection and location method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] Embodiment:

[0039] This embodiment proposes a low-voltage line fault detection and location method, referring to Figure 1 , which mainly includes the following multiple steps.

[0040] Step S1, the collected current and voltage data are sent to the data scheduling center through the intranet after preprocessing; specifically, in the present invention, the data scheduling center is the general control center, which is responsible for data transmission, data storage, and data processing.

[0041] Step S1 further includes the following multiple steps. Step S11, fault collection points are set on the low-voltage line every K kilometers, and adjacent fault collection points jointly detect the current and voltage data on the unit line; specifically, the interval of the fault collection points can be set according to actual needs.

[0042] After the fault collection point collects the current and voltage data, it conducts an automatic inspection and troubleshooting, and then extracts the current and voltage data that are not within the preset normal threshold range; specifically, the current and voltage data that are not within the preset normal threshold range are the preliminary suspected line abnormal data.

[0043] The current and voltage data are subjected to digital-to-analog conversion, and then the data is transmitted through the intranet of the power grid, specifically to the intranet data processing unit of the data scheduling center. Specifically, the data in the intranet data processing unit does not need to be decrypted and can be flexibly called.

[0044] In the present invention, the distance K between fault acquisition points can be adjusted according to the actual situation. During specific implementation, the fault acquisition points can also directly skip multiple intermediate fault acquisition points. At this time, the distance between the fault acquisition points can be 2K, 3K, or even nK. The fault acquisition points have a threshold judgment function. Every once in a while, the fault acquisition points will automatically perform threshold judgment on the current and voltage data collected during this period, automatically eliminating the data within the threshold range, and saving and transmitting the data exceeding the threshold range through the intranet.

[0045] Step S2: Encrypt and send the positioning data to the data scheduling center through the external network. Specifically, this step is transmitted through the external network and requires encryption.

[0046] Step S2 includes the following three steps. Step S21: Adjacent fault acquisition points immediately obtain the positioning data, and then retrieve the positioning data corresponding to the current and voltage data not within the preset normal threshold range to preliminarily determine the unit line with abnormal line. Specifically, the retrieved positioning data corresponds to the current and voltage data sent through the intranet.

[0047] Step S22: The position data of adjacent fault acquisition points with abnormal lines are respectively represented as Ln and Ln+1. There is also a line positioning initial coordinate. According to the linear distance and angular relationship between each fault acquisition point, the three-dimensional coordinates of Ln and Ln+1 are gradually obtained. Specifically, the line positioning initial coordinate is (0, 0, 0). Generally, the z coordinate of Ln is usually customized and is all h.

[0048] After the retrieved positioning data is obtained in Step S23, data encryption is performed, and then it is transmitted to the data scheduling center through the external network of the power grid.

[0049] In the present invention, the positioning data corresponds to the current and voltage data. The specific coordinates of Ln are:

[0050] (L1 cos α1 ± L2 cos α2... ± Lncos αn, L1 sin α1 ± L2 sin α2... ± Lnsin αn, h)

[0051] The specific coordinates of Ln+1 are:

[0052] (L1 cos α1 ± L2 cos α2... ± L(n+1)cos α(n+1), L1 sin α1 ± L2 sin α2... ± L(n+1)sin α(n+1), h)

[0053] Among them, L1 is the distance between the first fault acquisition point and the initial coordinate of the line positioning, Ln is the distance between the nth fault acquisition point and the previous fault acquisition point, an is the angle between the nth fault acquisition point and the previous acquisition point, and ± is specifically determined by using + or - according to the specific position of the nth fault acquisition point. h is the height of the fault acquisition point, and h is usually a fixed value.

[0054] Specifically, for data encryption, it further includes step S231 of dimensionless processing of the three-dimensional coordinate data of the positioning data. Under the condition of unified data units, the positioning data is proportionally enlarged or reduced, and then the three-dimensional coordinate data is converted into data between 0 and 255. The specific multiple of proportional enlargement or reduction is sent to the data scheduling center through the intranet. Specifically, the multiple of proportional enlargement or reduction is sent separately from the positioning data.

[0055] Step S232 is to sequentially represent the values of RGB with the converted three-dimensional coordinates to form multiple color images. In the present invention, one three-dimensional coordinate corresponds to one color image.

[0056] Step S233 is that multiple color images are combined in a specific order to form an encrypted transmission image of X*Y. X and Y are respectively the number of rows and columns of the transmission image. The encrypted transmission image is the encrypted data and is finally sent to the external network data processing unit of the data scheduling center.

[0057] In the present invention, first, the three-dimensional coordinate data is processed. The three-dimensional coordinate data is represented by data between 0 and 255, and this data is represented as RGB values, thereby obtaining color images. Each color image represents a three-dimensional coordinate. During transmission, it is transmitted with a transmission image composed of color images, and the arrangement method of the transmission image has a specific order. The encrypted transmission image is directly sent to the data scheduling center, and in the external network data processing unit, it is decrypted in combination with the multiple of proportional enlargement or reduction sent through the intranet.

[0058] For the aforementioned specific order, it is related to the specific fault type. If the fault type is a short circuit, then several color images are arranged in ascending order of the number of rows of the transmission image, and in the same row, they are arranged in ascending order of the number of columns. If the fault type is an open circuit, then several color images are arranged in ascending order of the number of columns of the transmission image, and in the same column, they are arranged in ascending order of the number of rows.

[0059] In the present invention, the order of features is divided into two different permutation methods according to the fault type. Specifically, they are arranged row by row and column by column. When arranged row by row, the first row is filled in ascending order first, and then the second row is also filled in ascending order, and so on until the last row X. If the last row X is not filled, 0 is filled in for padding. At the same time, when arranged column by column, it is arranged in ascending order from the first column to the Y column. If the last column Y is not filled, 0 is filled in for padding.

[0060] Step S3: The data scheduling center fuses the data sent through the intranet and the extranet and establishes a joint data model. Specifically, the establishment process and storage location of the joint data model are both in the data scheduling center.

[0061] Step S3 includes the following multiple steps. Step S31: Decrypt the transmitted image that has been transmitted. Specifically, it is decrypted into positioning data, and the positioning data is fused with the current and voltage data in a one-to-one correspondence to obtain fused data. In the present invention, the decrypted positioning data is exactly the same as the positioning data before encrypted transmission.

[0062] Step S32: Process the data among the fused data. Use time-domain transformation to process the current and voltage data, calculate the fundamental wave to higher harmonic phasors. The phasors include the amplitude, phase angle, frequency, active power, and reactive power of the current signal and voltage signal. Calculate the current and voltage characteristics from the phasors. In addition, extract the coordinate characteristics in the positioning data to obtain the mapping relationship between the current and voltage characteristics, coordinate characteristics, and fault type as the training samples. Specifically, the training data of the samples in this step is the data collected historically.

[0063] Step S33: Based on the training samples, adopt the deep learning method to establish a joint data model. Specifically, the joint data model can continuously update and learn.

[0064] In the present invention, for the decryption method of the encrypted transmitted image, it is decrypted according to the magnification or reduction ratio and a specific order. In the present invention, the magnification or reduction ratio is transmitted separately from the positioning data to ensure the security of the positioning data. In addition, for the established joint data model, it is essentially a verification model. The data sources of the training samples are all the data collected historically and have all been verified.

[0065] Subsequently, step S4 is carried out. After the fused data is verified by the joint data model, it is finally transmitted to multiple mobile terminals with the highest matching degree. Specifically, the fused data verified by the joint data model is the finally confirmed fault data.

[0066] Step S41: Input the latest integrated data sent to the data scheduling center into the joint data model for verification, re-extract the current and voltage data from the integrated data, and convert them into waveform signals. Specifically, the waveform signals only support the conversion of current and voltage data.

[0067] Step S42: Fit and compare the waveform signals with the simulated waveforms of the joint data model. Specifically, compare the amplitude, phase angle, and frequency of the waveform signals. If the similarity is less than 1%, the comparison is successful. Specifically, if any one of the similarities is greater than or equal to 1%, the comparison fails.

[0068] Step S43: After the comparison is successful, the data scheduling center sends the integrated data to several mobile terminals that are closest to the fault location and not in the maintenance state. Specifically, select multiple mobile terminals that are closest to the fault location data. At the same time, the mobile terminals need to be not in the maintenance state to ensure the schedulability of the mobile terminals.

[0069] In the present invention, for the verification of the waveform signals, specifically, the waveform signals to be verified are fitted with the closest simulated waveforms in the joint data model in the same coordinate. Through simulation calculation, compare whether the amplitude, phase angle, and frequency of the waveform signals of both are less than 1%. If all are less than 1%, the comparison is successful. If any one of them is greater than or equal to 1%, the comparison fails. For the mobile terminals, they can be installed on maintenance vehicles or carried by maintenance personnel.

[0070] In the present invention, the fault collection points are communicatively connected to each other. The fault collection points are communicatively connected to the data scheduling center. A positioning module based on the fault collection point distribution map is provided inside the fault collection points. In this embodiment, a single fault collection point can be provided with multiple collection devices. The fault collection point distribution map in the positioning module inside the fault collection points is obtained by the data scheduling center. The positioning module then performs positioning and positioning judgment according to the fault collection point distribution map.

[0071] In the present invention, the collected current and voltage data are waveform signals, and this data is directly sent through the internal network of the power grid. Due to the security of the internal network itself, there is no need to encrypt the current and voltage data. The positioning data is collected by the positioning module in the fault collection points. The positioning module itself uses the external network. Therefore, in the present invention, the positioning data is directly transmitted using the external network. However, due to the insecurity of the external network, it does not conform to the data security of the power grid. Therefore, the positioning data is encrypted. The current and voltage data and the positioning data will then be integrated at the data scheduling center and correspond one by one. The established joint data model is used for fault verification. Finally, the integrated data will be transmitted to the mobile terminals, and the mobile terminals perform scheduling and maintenance operations after receiving the instructions. The method of the present invention can achieve good detection and positioning effects.

[0072] For the substantial effects that the present invention can bring, specifically: collect current and voltage data as well as positioning data, and transmit them through the encryption methods of the intranet and the extranet respectively. According to the data sent through the intranet and the extranet, establish a joint data model, use the joint data model to verify the fault data, and send the confirmed completed fault data to the mobile terminal for dispatching and maintenance; the method of the present invention can effectively monitor the low-voltage line, and through the comprehensive transmission method of encrypting the intranet and the extranet, ensure the security of the data while ensuring the integrity of the transmission; the established joint data model can verify the fault and ensure the accuracy of the detection.

[0073] The above embodiments are further elaborations and explanations of the present invention for the convenience of understanding, and are not any limitations to the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for detecting and locating low-voltage line faults, characterized in that, it includes the following steps: S1. After the collected current and voltage data are preprocessed, they are sent to the data scheduling center through the intranet; including: S11. Fault collection points are set every K kilometers on the low-voltage line, and adjacent fault collection points jointly detect the current and voltage data on the unit line; S2. The positioning data is encrypted and sent to the data scheduling center through the extranet; including: S21. Adjacent fault collection points immediately obtain the positioning data, retrieve the positioning data corresponding to the current and voltage data not within the preset normal threshold range, and initially determine the unit line with abnormal line conditions; S22. Set the position data of adjacent fault collection points with abnormal line conditions as Ln and Ln+1 respectively, and set the initial coordinates of line positioning. According to the straight-line distance and angular relationship between each fault collection point, gradually calculate the three-dimensional coordinates of Ln and Ln+1; S23. Encrypt the retrieved positioning data and transmit it to the data scheduling center through the extranet of the power grid; S3. The data scheduling center fuses the data sent through the intranet and extranet and establishes a joint data model; S4. After the fused data is verified by the joint data model, it is sent to several mobile terminals with the highest matching degree.

2. A method for detecting and locating low-voltage line faults according to claim 1, characterized in that, the step S1 further includes the following steps: S12. The fault collection point automatically inspects and checks the collected current and voltage data, and retrieves the current and voltage data not within the preset normal threshold range; S13. Perform digital-to-analog conversion on the current and voltage data, and transmit the data through the intranet of the power grid, and finally transmit it to the intranet data processing unit of the data scheduling center.

3. A method for detecting and locating low-voltage line faults according to claim 1, characterized in that, the data encryption includes the following steps: S231. Perform dimensionless processing on the three-dimensional coordinate data of the positioning data. Under the condition of unified data units, after proportional amplification or reduction, convert the three-dimensional coordinate data into data between 0 and 255. The specific multiple of proportional amplification or reduction is sent to the data scheduling center through the intranet; S232. Represent the RGB values with the converted three-dimensional coordinates in turn to form several color images; S233. Several color images are combined into an encrypted transmission image of X*Y in a specific order, where X is the number of rows of the transmission image and Y is the number of columns of the transmission image. The encrypted transmission image is encrypted data and is finally sent to the extranet data processing unit of the data scheduling center.

4. A method for detecting and locating low-voltage line faults according to claim 3, characterized in that, the specific order is specifically related to the fault type. If the fault type is a short circuit, several color images are arranged in ascending order of the number of rows of the transmission image, and in ascending order of the number of columns in the same row; if the fault type is an open circuit, several color images are arranged in ascending order of the number of columns of the transmission image, and in ascending order of the number of rows in the same column.

5. A method for detecting and locating low-voltage line faults according to claim 3, characterized in that, Step S3 includes the following steps: S31, decrypt the transmitted image into positioning data, fuse the positioning data with the current and voltage data in a one-to-one correspondence to form fused data; S32, in the fused data, process the current and voltage data using time-domain transformation, calculate the fundamental to higher harmonic phasors. The phasors include the amplitude, phase angle, frequency, active power, and reactive power of the current signal and voltage signal. Calculate the current and voltage characteristics based on the phasors; at the same time, extract the coordinate characteristics from the positioning data to form the mapping relationship between the current and voltage characteristics, coordinate characteristics, and fault types as training samples; S33, based on the training samples, establish a joint data model using deep learning methods.

6. A low-voltage line fault detection and location method according to claim 5, characterized in that, Step S4 includes the following steps: S41, input the fused data newly sent to the data scheduling center and formed into the joint data model for verification, re-extract the current and voltage data and convert them into waveform signals; S42, perform a fitting comparison between the waveform signal and the simulated waveform of the joint data model. Specifically, compare the amplitude, phase angle, and frequency of the waveform signal. If the similarity is less than 1%, the comparison is successful; S43, after the comparison is successful, the data scheduling center sends the fused data to several mobile terminals that are closest to the fault location and are not in the maintenance state.

7. A low-voltage line fault detection and location method according to claim 1, characterized in that, The fault collection points are communicatively connected to each other, the fault collection points are communicatively connected to the data scheduling center, and a positioning module based on the fault collection point distribution map is provided inside the fault collection points.

Images