A distribution network fault detection and location analysis system

Abstract

The present invention belongs to the technical field of power anomaly detection, and discloses a distribution network fault detection and positioning analysis system; collects equipment data sets and performs preprocessing, the equipment data sets include voltage data, current data and temperature data; is used to analyze the equipment data sets, obtain equipment state reference values, and classify them to obtain equipment state classification results; processes the equipment data sets, obtains equipment early warning reference values, and performs positioning; processes the equipment state classification results to obtain response decision results. Generally speaking, the present invention has the significant advantages of good auxiliary detection effect, strong fault early warning capability and great auxiliary decision-making function.

Description

Technical Field

[0001] The present invention relates to the technical field of power anomaly detection, and more specifically, to a distribution network fault detection and positioning analysis system. Background Art

[0002] Distribution network refers to the power grid that receives electric energy from the transmission network or regional power plant, distributes it locally or step by step according to voltage to various users through distribution facilities. It is composed of overhead lines, cables, towers, distribution transformers, disconnectors, reactive power compensators and some ancillary facilities. It plays an important role in distributing electric energy in the power grid. With the rapid development of my country's economy, power security has been at the forefront of the world. However, due to the vast territory of my country and the large number of distribution network lines, the base number of the possibility of distribution network failures has inevitably increased. Therefore, how to enhance the efficiency of distribution network fault detection has become the top priority of the power department.

[0003] The patent application publication number CN109459664A discloses a distribution network fault detection and positioning analysis system. The fault location is detected by a fault location detection module, and the fault line or branch line of the line is directly located. At the same time, the specific approximate location of the branch line is given, so that the inspection personnel can quickly go to the scene to check and eliminate the fault. The mobile phone of the line inspection personnel is used, and with the help of the background software of the mobile phone of the inspection personnel, after the line fault is eliminated, only the confirmation information is pressed on the fault elimination confirmation piece, and the background software can obtain the relevant information of the fault elimination, which is convenient for realizing rapid fault recovery and power supply, and realizing real-time, objective and intelligent monitoring, evaluation and fault location of the health status of the main equipment of the distribution network, with high degree of automation and high efficiency.

[0004] However, although the above-mentioned distribution network fault detection and positioning analysis system can detect the fault point through the fault location detection module set therein, it can quickly locate the fault point to a certain extent, but the power department usually needs to be responsible for a large number of distribution network lines, and the range is generally wide. If a fault occurs, it will take a lot of time to rush to the site for repair. Therefore, it is particularly important to warn of distribution network faults and reduce the time cost of maintenance.

[0005] In view of this, the present invention proposes a distribution network fault detection and location analysis system to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: comprising: The data acquisition module is used to collect and pre-process the equipment data set, which includes voltage data, current data and temperature data; Furthermore, the method of collecting device data sets includes: By installing smart meters, the voltage readings and current readings of the power grid in the specified area are collected to obtain voltage data and current data; By installing non-contact infrared thermometers, the temperature readings of power grid lines in the specified area are collected to obtain temperature data; The preprocessing methods include data cleaning, data denoising and data normalization; The data processing module is used to analyze the equipment data set, obtain the equipment status reference value, and classify it to obtain the equipment status classification result; Further, the methods for analyzing the device data set include: Based on the device data set, by substituting into the calculation formula: Get the real-time device reference value, where is the voltage data, is the current data, is the temperature data, , and are the corresponding weight factors respectively; Collect M groups of historical real-time equipment reference values ​​as sample sets, and divide the sample sets into 70%M training sets, 15%M test sets, and 15%M validation sets; A primary classification model is established based on the training set, and the historical real-time equipment reference values ​​in the training set are obtained. The low value cluster center R, the normal value cluster center T, and the high value cluster center Y are preset. Three values ​​are selected in the training set as the original cluster centers, which represent the low value cluster center R, the normal value cluster center T, and the high value cluster center Y respectively. By substituting into the calculation formula: Get the distance between the data item and the original cluster center, calculate the distance between the data item in the training set and the original cluster center, and assign the data item to the nearest original cluster center, get three initial cluster centers, calculate the mean of the original cluster centers respectively, as the new original cluster center, where is the number of original cluster centers, For the The set of data items of the original cluster centers, Data items, For the The centroid of the original cluster centers; After reaching the preset number of iterations, a state classification model is obtained; The real-time equipment reference value is input into the state classification model, and the equipment state reference value is output; Further, the classification methods include: A device state threshold interval (D1, D2) is preset. When the device state reference value is less than D1, low value information is generated. When the device state reference value is greater than D1 and less than D2, normal value information is generated. When the device state reference value is greater than D2, high value information is generated. The low value information includes a group of fields with low device status data, the normal value information includes a group of fields with normal device status data, and the high value information includes a group of fields with high device status data; Pack low value information, normal value information and high value information to obtain the equipment status classification result; The early warning positioning module is used to process the equipment data set, obtain the equipment early warning reference value, and perform positioning; Furthermore, the device data set is processed in the following ways: By substituting into the calculation group: The voltage change rate is , Current change rate and temperature change rate ,in is the derivative, For time; Collect L sets of historical equipment data sets as sample sets, and divide the sample sets into 70%L training sets, 15%L test sets, and 15%L validation sets; Extract feature vectors based on the sample set, use keras to define the input layer and output layer, use feature vectors as input layer data, and use equipment warning reference values ​​as output layer data. Feature vectors are historical voltage change rates, historical current change rates, and historical temperature change rates. Building an early warning model based on the sample set, training the early warning model to obtain a first early warning model, testing and verifying the first early warning model using a test set and a validation set, and outputting a first early warning model that meets a preset accuracy, wherein the early warning model is one of a logistic regression model, a naive Bayes model, or a support vector machine model; Inputting the equipment data set into the first equipment early warning model to obtain an equipment early warning reference value; Furthermore, the positioning method includes connecting a signal generator to a distribution network line, injecting a periodic and low-frequency known signal into the line, using a detector to detect the strength, frequency and phase information of the known signal at different positions of the line, inputting the data collected by the detector into a host, calculating the propagation of the known signal according to a preset algorithm, outputting the calculation result, and obtaining the location of the fault point; The emergency response module is used to process the equipment status classification results and obtain the response decision results; Furthermore, the method for processing the device status classification result includes: When the equipment status classification result is low value information, a low value alarm SMS is sent to the staff receiving end through the communication unit; when the equipment status classification result is normal value information, a normal value SMS is sent to the staff receiving end through the communication unit; when the equipment status classification result is high value information, a high value alarm SMS is sent to the staff receiving end through the communication unit; The low value alarm message includes an explanation that the value of the distribution network line status is too low, the fault line location has poor insulation, poor grounding or low line temperature, and requires the staff to immediately locate the fault point, replace the insulation material of the fault point line, tighten the wiring or increase the line current; The normal value SMS includes information indicating that the distribution network line status values ​​are normal and the operating status is good, and reminds the staff to clean the lines, check the insulation materials and pay attention to the temperature changes in the area according to the planned cycle; The high value alarm SMS includes an explanation that the value of the distribution network line is too high, there is an overload, short circuit or high line temperature at the fault line location, and requires the staff to immediately locate the fault point, cut off the power supply, replace the damaged equipment at the fault point line, add overload protection equipment or reduce the line current; Pack low value alarm SMS, normal value SMS and high value alarm SMS to get response decision results; Further, S1: collecting and preprocessing a device data set, the device data set includes voltage data, current data and temperature data; S2: used to analyze the equipment data set, obtain the equipment status reference value, and classify it to obtain the equipment status classification result; S3: Process the equipment data set to obtain the equipment warning reference value and perform positioning; S4: Process the equipment status classification results to obtain a response decision result.

[0007] The technical effects and advantages of a distribution network fault detection and location analysis system of the present invention are as follows: The present invention analyzes the equipment data set to obtain the equipment status reference value, which allows the staff to very intuitively view the current operating status of the distribution network line. Compared with traditional manual inspections and manual tests, the work intensity of the staff is greatly reduced. By processing the equipment data set, the equipment warning reference value obtained can assist the staff to monitor the abnormal values ​​in the distribution network line in real time, thereby effectively reducing the time cost and economic loss caused by the occurrence of failures. By responding to the decision results, the staff can be assisted to take corresponding treatment measures quickly when a failure occurs, greatly reducing the decision-making time cost required for maintenance, thereby further reducing property losses. Generally speaking, the present invention has the significant advantages of good auxiliary detection effect, strong fault warning capability and large auxiliary decision-making function. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A schematic diagram of a distribution network fault detection and location analysis system of the present invention; Figure 2 It is a schematic diagram of a distribution network fault detection and location analysis method of the present invention. DETAILED DESCRIPTION

[0009] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not 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.

[0010] Example 1 See also Figure 1 As shown, a distribution network fault detection and location analysis system described in this embodiment includes: The data acquisition module is used to collect and pre-process the equipment data set, which includes voltage data, current data and temperature data; Furthermore, the method of collecting device data sets includes: By installing smart meters, the voltage readings and current readings of the power grid in the specified area are collected to obtain voltage data and current data; By installing non-contact infrared thermometers, the temperature readings of power grid lines in the specified area are collected to obtain temperature data; The preprocessing methods include data cleaning, data denoising and data normalization; The data processing module is used to analyze the equipment data set, obtain the equipment status reference value, and classify it to obtain the equipment status classification result; Further, the methods for analyzing the device data set include: Based on the device data set, by substituting into the calculation formula: Get the real-time device reference value, where is the voltage data, is the current data, is the temperature data, , and are the corresponding weight factors respectively; Collect M groups of historical real-time equipment reference values ​​as sample sets, and divide the sample sets into 70%M training sets, 15%M test sets, and 15%M validation sets; A primary classification model is established based on the training set, and the historical real-time equipment reference values ​​in the training set are obtained. The low value cluster center R, the normal value cluster center T, and the high value cluster center Y are preset. Three values ​​are selected in the training set as the original cluster centers, which represent the low value cluster center R, the normal value cluster center T, and the high value cluster center Y respectively. By substituting into the calculation formula: Get the distance between the data item and the original cluster center, calculate the distance between the data item in the training set and the original cluster center, and assign the data item to the nearest original cluster center, get three initial cluster centers, calculate the mean of the original cluster centers respectively, as the new original cluster center, where is the number of original cluster centers, For the The set of data items of the original cluster centers, Data items, For the The centroid of the original cluster centers; It should be explained that the centroid refers to the central position of a cluster center, which is the average value of the coordinates of all data lines in the cluster; After reaching the preset number of iterations, a state classification model is obtained; The real-time equipment reference value is input into the state classification model, and the equipment state reference value is output; Further classification methods include: A device state threshold interval (D1, D2) is preset. When the device state reference value is less than D1, low value information is generated. When the device state reference value is greater than D1 and less than D2, normal value information is generated. When the device state reference value is greater than D2, high value information is generated. The low value information includes a group of fields with low device status data, the normal value information includes a group of fields with normal device status data, and the high value information includes a group of fields with high device status data; Pack low value information, normal value information and high value information to obtain the equipment status classification result; The early warning positioning module is used to process the equipment data set, obtain the equipment early warning reference value, and perform positioning; Furthermore, the device data set is processed in the following ways: By substituting into the calculation group: The voltage change rate is , Current change rate and temperature change rate ,in is the derivative, For time; It needs to be explained that the derivative is used to indicate a small change or increment in a variable; Collect L sets of historical equipment data sets as sample sets, and divide the sample sets into 70%L training sets, 15%L test sets, and 15%L validation sets; Extract feature vectors based on the sample set, use keras to define the input layer and output layer, use feature vectors as input layer data, and use equipment warning reference values ​​as output layer data. Feature vectors are historical voltage change rates, historical current change rates, and historical temperature change rates. Building an early warning model based on the sample set, training the early warning model to obtain a first early warning model, testing and verifying the first early warning model using a test set and a validation set, and outputting a first early warning model that meets a preset accuracy, wherein the early warning model is one of a logistic regression model, a naive Bayes model, or a support vector machine model; Inputting the equipment data set into the first equipment early warning model to obtain an equipment early warning reference value; Furthermore, the positioning method includes connecting a signal generator to a power distribution network line, injecting a periodic and low-frequency known signal into the line, using a detector to detect the strength, frequency and phase information of the known signal at different positions of the line, inputting the data collected by the detector into a host, calculating the propagation of the known signal according to a preset algorithm, outputting the calculation result, and obtaining the location of the fault point; The emergency response module is used to process the equipment status classification results and obtain the response decision results; Furthermore, the method for processing the device status classification result includes: When the equipment status classification result is low value information, a low value alarm SMS is sent to the staff receiving end through the communication unit; when the equipment status classification result is normal value information, a normal value SMS is sent to the staff receiving end through the communication unit; when the equipment status classification result is high value information, a high value alarm SMS is sent to the staff receiving end through the communication unit; The low value alarm message includes an explanation that the value of the distribution network line status is too low, the fault line location has poor insulation, poor grounding or low line temperature, and requires the staff to immediately locate the fault point, replace the insulation material of the fault point line, tighten the wiring or increase the line current; The normal value SMS includes information indicating that the distribution network line status values ​​are normal and the operating status is good, and reminds the staff to clean the lines, check the insulation materials and pay attention to the temperature changes in the area according to the planned cycle; The high value alarm SMS includes an explanation that the value of the distribution network line is too high, there is an overload, short circuit or high line temperature at the fault line location, and requires the staff to immediately locate the fault point, cut off the power supply, replace the damaged equipment at the fault point line, add overload protection equipment or reduce the line current; Pack low value alarm SMS, normal value SMS and high value alarm SMS to get response decision results; The beneficial effects of this embodiment are as follows: by analyzing the equipment data set, the equipment status reference value obtained can enable the staff to view the current operating status of the distribution network line very intuitively. Compared with the traditional manual inspection and manual testing, the work intensity of the staff is greatly reduced. By processing the equipment data set, the equipment warning reference value obtained can assist the staff to monitor the abnormal values ​​in the distribution network line in real time, thereby effectively reducing the time cost and economic loss caused by the occurrence of faults. By responding to the decision results, the staff can be assisted to take corresponding treatment measures quickly when a fault occurs, greatly reducing the decision-making time cost required for maintenance, thereby further reducing property losses. In general, the present invention has the significant advantages of good auxiliary detection effect, strong fault warning capability and large auxiliary decision-making function.

[0011] Example 2 See also Figure 2 As shown, the part not described in detail in this embodiment is described in the description of embodiment 1, and a method for detecting and locating faults in a distribution network is provided, comprising: S1: collecting a device data set and performing preprocessing, the device data set comprising voltage data, current data and temperature data; S2: used to analyze the equipment data set, obtain the equipment status reference value, and classify it to obtain the equipment status classification result; S3: Process the equipment data set to obtain the equipment warning reference value and perform positioning; S4: Process the equipment status classification result to obtain a response decision result; Example 3 This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the distribution network fault detection and location analysis method provided above is implemented.

[0012] Since the electronic device introduced in this embodiment is an electronic device used to implement a distribution network fault detection and location analysis method in the embodiment of the present application, based on the distribution network fault detection and location analysis method introduced in the embodiment of the present application, the technical personnel of this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is not described in detail here. As long as the technical personnel of this field implement the electronic device used in the distribution network fault detection and location analysis method in the embodiment of the present application, it belongs to the scope of protection of this application.

[0013] 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.

[0014] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A distribution network fault detection and location analysis system, characterized in that: include: The data acquisition module is used to collect and pre-process the equipment data set, which includes voltage data, current data and temperature data; The data processing module is used to analyze the equipment data set, obtain the equipment status reference value, and classify it to obtain the equipment status classification result; The early warning positioning module is used to process the equipment data set, obtain the equipment early warning reference value, and perform positioning; The emergency response module is used to process the equipment status classification results and obtain the response decision results.

2. A distribution network fault detection and location analysis system according to claim 1, characterized in that: Ways to collect device data sets include: By installing smart meters, the voltage readings and current readings of the power grid in the specified area are collected to obtain voltage data and current data; By installing non-contact infrared thermometers, the temperature readings of power grid lines in the specified area are collected to obtain temperature data; The preprocessing methods include data cleaning, data denoising and data normalization.

3. A distribution network fault detection and location analysis system according to claim 2, characterized in that: Ways to analyze device data sets include: Based on the device data set, by substituting into the calculation formula: Get the real-time device reference value, where is the voltage data, is the current data, is the temperature data, , and are the corresponding weight factors respectively; Collect M groups of historical real-time equipment reference values ​​as sample sets, and divide the sample sets into 70%M training sets, 15%M test sets, and 15%M validation sets; A primary classification model is established based on the training set, and the historical real-time equipment reference values ​​in the training set are obtained. The low value cluster center R, the normal value cluster center T, and the high value cluster center Y are preset. Three values ​​are selected in the training set as the original cluster centers, which represent the low value cluster center R, the normal value cluster center T, and the high value cluster center Y respectively. By substituting into the calculation formula: Get the distance between the data item and the original cluster center, calculate the distance between the data item in the training set and the original cluster center, and assign the data item to the nearest original cluster center, get three initial cluster centers, calculate the mean of the original cluster centers respectively, as the new original cluster center, where is the number of original cluster centers, For the The set of data items of the original cluster centers, Data items, For the The centroid of the original cluster centers; After reaching the preset number of iterations, a state classification model is obtained; The real-time device reference value is input into the state classification model, and the device state reference value is output.

4. A distribution network fault detection and location analysis system according to claim 3, characterized in that: The classification methods include: A device state threshold interval (D1, D2) is preset. When the device state reference value is less than D1, low value information is generated. When the device state reference value is greater than D1 and less than D2, normal value information is generated. When the device state reference value is greater than D2, high value information is generated. The low value information includes a group of fields with low device status data, the normal value information includes a group of fields with normal device status data, and the high value information includes a group of fields with high device status data; Pack low value information, normal value information and high value information to obtain the device status classification result.

5. A distribution network fault detection and location analysis system according to claim 4, characterized in that: The device data set is processed in the following ways: By substituting into the calculation group: The voltage change rate is , Current change rate and temperature change rate ,in is the derivative, For time; Collect L sets of historical equipment data sets as sample sets, and divide the sample sets into 70%L training sets, 15%L test sets, and 15%L validation sets; Extract feature vectors based on the sample set, use keras to define the input layer and output layer, use feature vectors as input layer data, and use equipment warning reference values ​​as output layer data. Feature vectors are historical voltage change rates, historical current change rates, and historical temperature change rates. Building an early warning model based on the sample set, training the early warning model to obtain a first early warning model, testing and verifying the first early warning model using a test set and a validation set, and outputting a first early warning model that meets a preset accuracy, wherein the early warning model is one of a logistic regression model, a naive Bayes model, or a support vector machine model; The equipment data set is input into the first equipment early warning model to obtain the equipment early warning reference value.

6. A distribution network fault detection and location analysis system according to claim 5, characterized in that: The positioning method includes connecting the signal generator to the distribution network line, injecting periodic and low-frequency known signals into the line, using detectors to detect the strength, frequency and phase information of the known signals at different positions of the line, inputting the data collected by the detectors into the host, calculating the propagation of the known signals according to the preset algorithm, outputting the calculation results, and obtaining the location of the fault point.

7. A distribution network fault detection and location analysis system according to claim 6, characterized in that: The methods for processing the equipment status classification results include: When the equipment status classification result is low value information, a low value alarm SMS is sent to the staff receiving end through the communication unit; when the equipment status classification result is normal value information, a normal value SMS is sent to the staff receiving end through the communication unit; when the equipment status classification result is high value information, a high value alarm SMS is sent to the staff receiving end through the communication unit; The low value alarm message includes an explanation that the value of the distribution network line status is too low, the fault line location has poor insulation, poor grounding or low line temperature, and requires the staff to immediately locate the fault point, replace the insulation material of the fault point line, tighten the wiring or increase the line current; The normal value SMS includes information indicating that the distribution network line status values ​​are normal and the operating status is good, and reminds the staff to clean the lines, check the insulation materials and pay attention to the temperature changes in the area according to the planned cycle; The high value alarm SMS includes an explanation that the value of the distribution network line is too high, there is an overload, short circuit or high line temperature at the fault line location, and requires the staff to immediately locate the fault point, cut off the power supply, replace the damaged equipment at the fault point line, add overload protection equipment or reduce the line current; The low value alarm SMS, normal value SMS and high value alarm SMS are packaged to obtain the response decision result.

8. A method for fault detection and location analysis in a distribution network, implemented by the analysis system according to any one of claims 1 to 7, characterized in that: S1: Collect and preprocess equipment data sets, including voltage data, current data, and temperature data; S2: used to analyze the equipment data set, obtain the equipment status reference value, and classify it to obtain the equipment status classification result; S3: Process the equipment data set to obtain the equipment warning reference value and perform positioning; S4: Process the equipment status classification results to obtain a response decision result.

Images