A method and system for detecting leakage faults in a low-voltage AC distribution network

Abstract

The present invention discloses a method and system for detecting leakage faults in a low-voltage AC distribution network, which relates to the technical field of power grid fault detection. It is used to solve the problem that when the leakage current is small but persists, it cannot be detected in time, resulting in detection lag. According to historical data, multiple preselected nodes are preferentially selected in the initialization stage. For each preselected node, the safety distance coefficient and sensor compatibility ratio of each preselected node are collected and synthesized to obtain key points. The key points are obtained and sensors corresponding to the points are connected for power-on detection to obtain drift error and DC component. The sensor screening coefficient is obtained by substituting into the logistic regression algorithm, and the sensor screening threshold is obtained through binary recursion to obtain the retained sensors and deleted sensors. The temperature difference between the environment of each retained sensor and the surface of the line and the total harmonic distortion ratio are detected, and a set of fuzzy rules are formulated for fuzzy reasoning to determine the leakage fault detection result, improving the accuracy of leakage fault detection.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid fault detection. More specifically, the present invention relates to a method and system for detecting leakage faults in a low-voltage AC distribution network. Background Art

[0002] The low-voltage AC distribution network is the end link of the power system, directly supplying power to users. The rated voltages of common low-voltage distribution networks are 220V (single-phase) and 380V (three-phase). The low voltage reduces the insulation requirements of electrical equipment and is used in residential buildings, commercial buildings, industrial enterprises, agricultural production, etc.

[0003] The prior art has the following deficiencies:

[0004] At present, there are various distribution methods for low-voltage AC distribution networks. Usually, there is a three-phase four-wire power supply with the neutral point directly grounded to improve safety and reduce the risk of interphase short circuits. However, no matter which distribution method is used, it will increase line losses, limit the power supply distance. For the detection of leakage faults during short-distance transportation, simply relying on the existing residual current protection device will result in fewer leakage detection mechanisms in the line, and when the leakage current is small but persistent, it cannot be detected in time to trigger the protection mechanism, resulting in detection lag and insufficient detection accuracy. Therefore, a method and system for detecting leakage faults in a low-voltage AC distribution network are proposed.

[0005] The above information disclosed in the background art section is only used to strengthen the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method and system for detecting leakage faults in a low-voltage AC distribution network. By using a screening mechanism to screen out sensors corresponding to preselected nodes and key nodes, the detection accuracy and timeliness are improved to solve the problems raised in the above background art.

[0007] To achieve the above object, the present invention provides the following technical solutions. A method and system for detecting leakage faults in a low-voltage AC distribution network include:

[0008] S1: According to historical data, in the initialization stage, a plurality of preselected nodes are preferentially selected. For each preselected node, the safety distance coefficient and sensor compatibility ratio corresponding to the preselected point positions of each preselected node are collected and synthesized to obtain a point position classification coefficient, which is compared with a preset point position classification coefficient threshold, and the preselected point positions are screened again to obtain key point positions;

[0009] S2: Obtain key points and connect to the sensors at the corresponding points for power-on detection to obtain drift errors and DC components, and substitute them into the logistic regression algorithm to obtain the sensor screening coefficients;

[0010] S3: Obtain the sensor screening coefficients, and obtain the sensor screening thresholds through binary recursion to obtain the retained sensors and the deleted sensors, and select the retained sensors for retention;

[0011] S4: Obtain the retained sensors, detect the environmental characteristic information of each retained sensor and the harmonic characteristics generated by the nonlinear load to obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio;

[0012] S5: Obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio, substitute them into fuzzy logic for fuzzy reasoning to obtain the leakage fault detection result.

[0013] In a preferred embodiment, in the initialization stage, schedule historical data, set the number of preselected nodes to n, and collect the safety distance coefficients and sensor compatibility ratios of each preselected node according to the preselected nodes;

[0014] By obtaining the actual distance between the preselected node and the nearest object and calculating the ratio with the preset safety distance, the safety distance coefficient of the -th preselected node is obtained; where is the index of the -th preselected node and is a positive integer;

[0015] By presetting different types of sensors, count the number of sensor types compatible with each preselected node, and calculate the ratio of the number of compatible sensor types to the total number of sensor types to obtain the sensor compatibility ratio of the -th preselected node.

[0016] In a preferred embodiment, standardize and weighted calculate the safety distance coefficient and sensor compatibility ratio of each preselected node to obtain the point classification coefficient of each preselected node, and the number of point classification coefficients is the same as the number of preselected nodes;

[0017] Compare the point classification coefficient with the preset point classification coefficient threshold. If the point classification coefficient is greater than or equal to the point classification coefficient threshold, mark the current preselected point as a key point. If the point classification coefficient is less than the point classification coefficient threshold, mark the current preselected point as a deleted point and delete the deleted point.

[0018] In a preferred embodiment, set sensors according to the position information of the key points. After the key points are connected to the sensors, perform input tests on their electrical signals;

[0019] Set the sensor to run under constant leakage current for 24 hours, collect the voltage output by the sensor, and calculate the ratio of the absolute value of the subtraction between the maximum output voltage value and the minimum output voltage value measured by the sensor within 24 hours and the rated output value within the theoretical maximum measurement range of the sensor to obtain the first Drift error of each sensor ,in, is the sensor index, which is a positive integer;

[0020] By setting a complete cycle, the sensor is used to perform regular sampling of the sawtooth wave, recording the waveform value at each time point, summing the waveform values of all sampling points within the cycle, and calculating the ratio with the total number of sampling points to obtain the first The DC component of each sensor .

[0021] In a preferred embodiment, the DC component and the drift error are standardized and substituted into a logistic regression calculation to obtain the sensor screening coefficient.

[0022] In a preferred embodiment, by collecting historical data of sensors and sensor screening coefficients, a candidate sensor list is created, a search range is determined by setting minimum and maximum values of sensor screening thresholds, and a recursive function is defined to obtain the sensor screening thresholds;

[0023] If the sensor screening coefficient is greater than or equal to the sensor screening threshold, the sensor corresponding to the current key point is marked as a deleted sensor and an end signal is generated;

[0024] If the sensor screening coefficient is less than the sensor screening threshold, the sensor corresponding to the current key point is marked as a reserved sensor and a reservation signal is generated;

[0025] The sensor corresponding to the current key point marked as the retained sensor is retained, and the sensor corresponding to the current key point marked as the deleted sensor is deleted.

[0026] In a preferred embodiment, a temperature sensor is used to collect the ambient temperature in the distribution network, an infrared temperature sensor is used to measure the line surface temperature, and the ambient temperature is subtracted from the line surface temperature to obtain the temperature difference between the environment and the line surface;

[0027] The current waveform signal is sampled and Fourier transformed to obtain a frequency domain signal. The first frequency component and remaining subharmonics are extracted from the frequency domain signal, and the remaining subharmonics are marked as high-order harmonics. The ratio of the square sum of the current components of all high-order harmonics to the first frequency component is calculated to obtain the total harmonic distortion ratio.

[0028] In a preferred embodiment, the temperature difference between the environment and the line surface and the total harmonic distortion ratio are defined as input variables, and they are respectively divided into different fuzzy sets;

[0029] The leakage fault detection result is defined as an output variable, and it is divided into a fuzzy set;

[0030] Fuzzy rules are formulated to describe the influence of the temperature difference between the environment and the line surface and the total harmonic distortion ratio on the leakage fault detection result;

[0031] Fuzzy reasoning is performed according to the fuzzy rules to determine the leakage fault detection result.

[0032] A leakage fault detection system for a low-voltage AC distribution network includes a point selection module, a sensor matching module, a sensor screening module, and a fault detection module;

[0033] The point selection module is used to preferentially select multiple preselected nodes in the initialization stage based on historical data. For each preselected node, the safety distance coefficient and the sensor compatibility ratio corresponding to the preselected point of each preselected node are collected and integrated to obtain a point classification coefficient, which is compared with a preset point threshold, and the preselected points are secondarily screened to obtain key points, and then sent to the sensor matching module;

[0034] The sensor matching module is used to obtain the key points and connect the sensors corresponding to the points, perform power-on detection, obtain the drift error and the DC component, substitute them into the logistic regression algorithm to obtain the sensor screening coefficient, and then send it to the sensor screening module;

[0035] The sensor screening module is used to obtain the sensor screening coefficient, obtain the sensor screening threshold through binary recursion, obtain the retained sensors and the deleted sensors, select the retained sensors for retention, and send the retained sensors to the fault detection module;

[0036] The fault detection module is used to obtain the retained sensors, detect the environmental characteristic information of each retained sensor and the harmonic characteristics generated by the nonlinear load, obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio, and use fuzzy logic to determine the leakage fault detection result.

[0037] The technical effects and advantages of the present invention:

[0038] 1. The present invention preferentially selects multiple preselected nodes in the initialization stage based on historical data. For each preselected node, the safety distance coefficient and sensor compatibility ratio of each preselected node are collected and synthesized to obtain key points. The key points are obtained and the sensors corresponding to the corresponding points are accessed for power-on detection to obtain drift errors and DC components, which are substituted into the logistic regression algorithm to obtain sensor screening coefficients. The sensor screening threshold is obtained through binary recursion, and the retained sensors and deleted sensors are obtained. The retained sensors are selected for retention, improving the accuracy of leakage detection, ensuring the long-term stability of sensors, improving detection efficiency, and reducing resource waste;

[0039] 2. The present invention obtains the retained sensors, detects the environmental characteristic information of each retained sensor and the harmonic characteristics generated by nonlinear loads, obtains the temperature difference between the environment and the line surface and the total harmonic distortion ratio, formulates a set of fuzzy rules for fuzzy inference, and determines the leakage fault detection result, improving the accuracy of leakage fault detection, the early fault warning ability, and reducing the probability of false alarms and missed alarms of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a method flow chart of a leakage fault detection method for a low-voltage AC distribution network according to the present invention.

[0041] Figure 2 It is a module schematic diagram of a leakage fault detection system for a low-voltage AC distribution network according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0043] Embodiment 1

[0044] Please refer to Figure 1 , a leakage fault detection method for a low-voltage AC distribution network, and the specific operation process is as follows:

[0045] S1: Based on historical data, multiple preselected nodes are preferentially selected in the initialization stage. For each preselected node, the safety distance coefficient and sensor compatibility ratio corresponding to the preselected points of each preselected node are collected and synthesized to obtain a point classification coefficient, which is compared with a preset point classification coefficient threshold, and the preselected points are screened again to obtain key points;

[0046] In the initialization phase, schedule historical data and ensure that the selected nodes can accurately reflect the leakage risk in the low-voltage distribution network according to the set data scheduling and analysis mechanism. The following are the specific methods:

[0047] Obtain and organize historical data, call the low-voltage distribution network operation and maintenance system, substation database, and fault logs, and count the number of leakage faults that occurred on the line, the duration of each fault, the fault type (instantaneous leakage, continuous leakage, intermittent leakage, etc.), and the proportion of the affected area;

[0048] Determine the specific location and quantity of the preselected nodes based on the number of leakage faults that occurred on the line, the duration of each fault, the fault type, and the proportion of the affected area;

[0049] Collect the safety distance coefficient and sensor compatibility ratio for each preselected node according to the preselected nodes;

[0050] The acquisition logic of the safety distance coefficient is to calculate the ratio by obtaining the actual distance between the preselected node and the nearest object and the preset safety distance, and obtain the safety distance coefficient of the th preselected node ; where is the preselected node index and is a positive integer;

[0051] Specifically, determine the nearest object around the preselected node. The nearest object refers to any object that is stationary relative to the preselected node, including but not limited to fixed facilities such as building structures, distribution cabinets, equipment casings, metal conductors, trees, and pipes. For the distance measurement between the preselected node and the nearest object, the straight-line distance between the installation point and the nearest object can be measured based on a laser rangefinder sensor or the distance from the installation point to the nearest object can be obtained using an ultrasonic sensor. The specific sensor for measurement is not limited and will not be elaborated here;

[0052] It should be noted that the preset safety distance is obtained through the comprehensive analysis of historical data and the statistical proportion of the possibility of historical leakage faults by the experimenters and will not be elaborated here;

[0053] Among them, when the safety distance coefficient is greater than 1, it indicates that the installation point meets the safety requirements, that is, the distance between the installation point and the surrounding objects is greater than the preset safety distance and meets the safety specifications; when the safety distance coefficient is less than or equal to 1, it indicates that there may be potential safety hazards at this installation point, and the installation position needs to be adjusted or additional safety protection measures need to be taken;

[0054] The sensor compatibility ratio refers to the ratio of each preselected node's compatibility with sensors. Its acquisition logic is to preset different types of sensors, count the number of sensor types that each preselected node can be compatible with, and calculate the ratio of the number of compatible sensor types to the total number of sensor types to obtain the Sensor compatibility ratio of a preselected node ;

[0055] Among them, the number of different types of sensors preset is not limited. Each type of sensor can be used for specific environments or specific types of leakage detection, including but not limited to: residual current sensors, voltage sensors, current transformers, electromagnetic field induction sensors. Further, the experimenter can classify the sensors by size or corresponding function, or by different sensor versions. The specific number of different types of sensors is not limited and will not be elaborated here;

[0056] It should be noted that the higher the sensor compatibility ratio, the better the adaptability of the preselected node, which can install and support multiple types of sensors, making the node highly available under different detection schemes;

[0057] Standardize the safety distance coefficient and sensor compatibility ratio of each preselected node according to the Max-Min normalization algorithm. The specific formula is:

[0058] In the formula, is the safety distance coefficient, is the standardized safety distance coefficient, is the minimum value of the safety distance coefficient, is the maximum value of the safety distance coefficient;

[0059] Among them, the above formula is also used for standardizing the sensor compatibility ratio and will not be elaborated here;

[0060] Perform weighted calculation on the safety distance coefficient and sensor compatibility ratio of each preselected node to obtain the point classification coefficient of each preselected node, and the number of point classification coefficients is the same as the number of preselected points;

[0061] It should be noted that the preselected point is the specific point obtained by the preselected node through a series of parameter processing and calculations. The preselected point also includes the position information of the preselected node. Specifically, one node corresponds to one point, that is, there are preselected points with the same number as the preselected nodes;

[0062] Compare the point classification coefficient with the preset point classification coefficient threshold. If the point classification coefficient is greater than or equal to the point classification coefficient threshold, mark the current preselected point as a key point. If the point classification coefficient is less than the point classification coefficient threshold, mark the current preselected point as a deleted point and delete the deleted point;

[0063] S2: Obtain the key points and connect the sensors corresponding to the points, conduct power-on detection to obtain the drift error and DC component, and substitute them into the logistic regression algorithm to obtain the sensor screening coefficient;

[0064] Among them, the key points include the position information of the key points. Sensors of corresponding categories are set according to the position information of the key points. Specifically, the categories of sensors connected to each key point are not limited, but are specifically implemented by the experimenters according to the actual application scenarios, which will not be elaborated here;

[0065] Conducting power-on detection means that after the key points are connected to the sensors, an electrical signal input test is performed on them to evaluate their performance in the actual working state, including measuring the DC component and drift error, to ensure that the selected sensors have high accuracy, stability and adaptability. The specific operation process is as follows:

[0066] Step A1: Install sensors of corresponding categories at the key points and make electrical connections with the low-voltage power distribution network;

[0067] Step A2: Calibrate the zero point of the sensor to ensure that the output voltage / current without external interference is the reference benchmark value;

[0068] Step A3: Set the data sampling frequency (such as 10 kHz) to ensure sufficient sampling accuracy;

[0069] Through the above steps, the corresponding sawtooth wave and constant input of the sensor can be collected;

[0070] Among them, the sawtooth wave refers to a periodic waveform whose shape is similar to the edge of a sawtooth. Its characteristic is that the signal rises (or falls) linearly within one cycle, then suddenly returns to the starting value, and repeats this process. It is used to measure the dynamic response characteristics of the sensor, observe the ability of the sensor to follow gradually changing signals, detect whether the sensor can accurately capture high-frequency components, and analyze whether there are abnormal jitters or distortions in the sawtooth wave signal;

[0071] The Fourier series expansion of the sawtooth wave signal shows that it contains multiple high-order harmonics. The specific mathematical formula is expressed as:

[0072] ;

[0073] In the formula, is the signal amplitude, is the fundamental frequency, is the sawtooth wave signal in the time domain, that is, the voltage or current that changes with time, is the harmonic order (i.e., the index of the sine component in the Fourier expansion), is the alternating positive and negative sign, indicating the alternating phase change of odd harmonics, is a time variable, is the frequency sine wave;

[0074] Furthermore, the sawtooth wave is a non-sinusoidal periodic signal, which can be decomposed into the superposition of multiple sine waves and contains all odd harmonics. The amplitude of each harmonic decreases. Due to the slow attenuation of high-order harmonics, the sawtooth wave has a strong high-frequency component, which poses a high requirement for the frequency response of the sensor;

[0075] A constant input means giving the sensor a stable and unchanged DC or AC signal, which can further evaluate the sensitivity, linearity, frequency response, and error characteristics of the sensor;

[0076] The drift error refers to measuring the measurement stability of the sensor during long-term operation or environmental changes (temperature, electromagnetic interference). Its acquisition logic is to set the sensor to operate at a constant leakage current for 24 hours, collect the voltage output by the sensor, and subtract the absolute value of the difference between the maximum output voltage value and the minimum output voltage value measured by the sensor within 24 hours from the rated output value within the theoretical maximum measurement range of the sensor, and then calculate the ratio to obtain the drift error of the th sensor, where is the sensor index and is a positive integer;

[0077] Among them, the rated output value within the theoretical maximum measurement range of the sensor refers to the standard output voltage or current value when the sensor is at full scale within its designed operating range. It is determined by the design specifications and calibration of the sensor and is not affected by the actual measurement conditions. Even if the current measurement value of the sensor is small, it still has a fixed theoretical maximum output value. The specific rated output value within the theoretical maximum measurement range of the sensor is not limited and will not be elaborated here;

[0078] The DC component refers to the average value of all instantaneous values of the signal waveform (sawtooth wave) within a complete cycle in a periodic signal. Its acquisition logic is to set a complete cycle, use the sensor to sample the sawtooth wave at regular intervals, record the waveform values at each time point, sum up the waveform values of all sampling points within the cycle, and calculate the ratio with the total number of sampling points to obtain the DC component of the th sensor;

[0079] It should be noted that the calculation of the DC component involves the periodic sampling of the sawtooth wave signal, while the calculation of the drift error involves the output fluctuation of the sensor during long-term operation. Setting the complete cycle needs to be consistent with the 24 hours set for the drift error to ensure that within the same cycle time period, the accuracy, stability, and comparability of the measurement data are improved, so as to more scientifically evaluate the performance of the sensor;

[0080] Standardize the DC component and the drift error. The specific processing method can be based on the above Max-Min standardization algorithm, which will not be elaborated here;

[0081] Substitute the DC component and the drift error into the logistic regression calculation. The specific formula is as follows:

[0082] ;

[0083] In the formula, is the result of the logistic regression calculation, that is, the sensor screening coefficient. e is the natural base, and y is the linear combination term of the logistic regression model. Specifically, y can be set as:

[0084] ;

[0085] In the formula, is the bias term, and are the regression coefficients of the DC component and the drift error respectively;

[0086] S3: Obtain the sensor screening coefficient, and obtain the sensor screening threshold through binary recursion, obtain the retained sensors and the deleted sensors, and select the retained sensors for retention;

[0087] The logic for obtaining the sensor screening threshold is to collect the historical data of the sensors, including the sensor readings and the corresponding key point data, determine the sensor screening coefficient to evaluate the screening results of the sensors, select the sensors strongly related to the key points according to the sensor screening coefficient, create a list of candidate sensors, and then determine the search range by setting the minimum value low and the maximum value high of the sensor screening threshold. Define a recursive function find_threshold(low, high), and its implementation steps are as follows:

[0088] Step B1: Calculate the middle value of the current range mid = (low + high) / 2;

[0089] Step B2: Use the current middle value mid to evaluate the performance of the sensor and judge whether the evaluation result meets the expectation;

[0090] Step B3: If it meets, it is used as a candidate threshold. If it does not meet and the performance is too low, adjust the search range to low = mid + 1 (that is, greater than the middle value of the current range). If it does not meet and the performance meets the requirements, continue to search in the left half and adjust to high = mid - 1;

[0091] Step B4: When low is greater than high, the recursion ends at this time, and return the best threshold found;

[0092] It should be noted that in step B3, the right half can be preferentially selected for searching. The specific rules for selecting the left and right halves are not limited, but are determined by the experimenter based on the specific actual candidate sensor list, which will not be elaborated here;

[0093] After obtaining the sensor screening coefficient, compare and analyze the sensor screening coefficient with the sensor screening threshold obtained by binary recursion;

[0094] If the sensor screening coefficient is greater than or equal to the sensor screening threshold, mark the sensor corresponding to the current key point as a deleted sensor and generate an end signal;

[0095] If the sensor screening coefficient is less than the sensor screening threshold, mark the sensor corresponding to the current key point as a retained sensor and generate a retention signal;

[0096] Retain the sensor corresponding to the current key point marked as a retained sensor, and delete the sensor corresponding to the current key point marked as a deleted sensor;

[0097] In this invention, by relying on historical data, multiple preselected nodes are preferentially selected in the initialization stage. For each preselected node, collect and synthesize the safety distance coefficient and sensor compatibility ratio of each preselected node to obtain key points, obtain the key points and connect the sensors corresponding to the corresponding points, conduct power-on detection to obtain drift error and DC component, substitute them into the logistic regression algorithm to obtain the sensor screening coefficient, obtain the sensor screening threshold through binary recursion, obtain the retained sensors and deleted sensors, and select the retained sensors for retention, which improves the accuracy of leakage detection, ensures the long-term stability of the sensors, improves the detection efficiency, and reduces resource waste.

[0098] Embodiment 2

[0099] In Embodiment 1 of this invention, it mainly illustrates by example that based on historical data, multiple preselected nodes are preferentially selected in the initialization stage. For each preselected node, collect and synthesize the safety distance coefficient and sensor compatibility ratio of each preselected node to obtain key points, obtain the key points and connect the sensors corresponding to the corresponding points, conduct power-on detection to obtain drift error and DC component, substitute them into the logistic regression algorithm to obtain the sensor screening coefficient, obtain the sensor screening threshold through binary recursion, obtain the retained sensors and deleted sensors, and select the retained sensors for retention; however, in Embodiment 1, only the accuracy of leakage fault detection in low-voltage AC distribution networks is calibrated and screened, and how to detect leakage faults is not analyzed, lacking a specific leakage fault determination mechanism. For the above problems, Embodiment 2 of this invention is further refined;

[0100] S4: Obtain reserved sensors, detect the environmental characteristic information of each reserved sensor and the harmonic characteristics generated by non-linear loads, and obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio;

[0101] Among them, the environmental characteristic information refers to finding the electrical fire risk by detecting the temperature characteristics to avoid causing the temperature of electrical equipment to rise and trigger a fire. Usually, due to aging of the line, the insulation layer is damaged, resulting in leakage current, the equipment wiring is loose, resulting in local high temperature, and overload operation causes the wire to heat up, accelerating insulation aging, etc.;

[0102] Among them, the harmonic characteristics generated by non-linear loads refer to when non-linear devices such as rectifiers, frequency converters, switching power supplies, and arc furnaces are operating, the current waveform is distorted, resulting in high-order harmonic currents other than the fundamental wave being injected into the power grid, thus affecting the power quality. Since the current waveform of non-linear loads is different from the standard sine wave, its Fourier series expansion will contain integer multiple harmonic components other than the fundamental wave frequency (power frequency, such as 50Hz or 60Hz). Therefore, harmonic currents cause additional losses, causing the temperature of equipment such as motors, transformers, and cables to rise, increasing the electrical fire risk;

[0103] The acquisition logic of the temperature difference between the environment and the line surface is to use a temperature sensor to collect the environmental temperature in the distribution network, use an infrared temperature sensor to measure the line surface temperature, and subtract the environmental temperature from the line surface temperature to obtain the temperature difference between the environment and the line surface;

[0104] The acquisition logic of the total harmonic distortion ratio is to sample the current waveform signal, perform Fourier transform on the current waveform signal to obtain the frequency domain signal, then extract the first frequency component and the remaining sub-harmonics from the frequency domain signal, mark the remaining sub-harmonics as high-order harmonics, and calculate the ratio of the sum of the squares of the current components of all high-order harmonics to the first frequency component to obtain the total harmonic distortion ratio;

[0105] It should be noted that the total harmonic distortion ratio represents the intensity of the harmonic components in the current and reflects the distortion degree of the current waveform. The higher the value of the total harmonic distortion ratio, the greater the distortion degree of the current waveform and the greater the electrical fire risk;

[0106] S5: Obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio, substitute them into fuzzy logic for fuzzy reasoning, and obtain the leakage fault detection result;

[0107] For example, "High", "Low", "Medium" for the temperature difference between the environment and the line surface, and "Elevated", "Slow", "Moderate" for the total harmonic distortion ratio;

[0108] Formulate a set of fuzzy rules to describe the influence of different input variables on the output variable. The definition of the rules can be based on professional knowledge or obtained through data analysis and experiments. For example:

[0109] Mark the temperature difference between the environment and the line surface as X, the total harmonic distortion ratio as U, and the leakage fault detection result as C_results;

[0110] Then it can be defined as:

[0111] Rule 1: IF (X is High) AND (U is Elevated) THEN (C_results is High)

[0112] Rule 2: IF (X is Low) AND (U is Slow) THEN (C_results is Low) ...

[0113] Perform fuzzy reasoning according to the fuzzy rules to determine the leakage fault detection result;

[0114] It should be noted that the division of the fuzzy set can be adjusted according to the actual situation. For example, although three fuzzy sets are used as examples in this embodiment, in fact, the temperature difference between the environment and the line surface and the total harmonic distortion ratio can be divided into more than three sets to facilitate more accurate adjustment according to different occupancy rates and change rates;

[0115] Furthermore, for the judgment of high, medium, and low of the temperature difference between the environment and the line surface and the total harmonic distortion ratio, thresholds can be set according to the actual situation for judgment. For example, when the temperature difference between the environment and the line surface exceeds 67%, it is marked as "High", and when the total harmonic distortion ratio is higher than 72%, it is marked as "Elevated", etc., which will not be elaborated here;

[0116] Optionally, return the current leakage fault detection result as historical data to step S1. When calling the preselected points in the next round, use it as the selection rule in the initialization stage to screen out multiple preselected nodes to complete the method closed-loop;

[0117] The present invention obtains and retains sensors, detects the environmental characteristic information of each retained sensor and the harmonic characteristics generated by non-linear loads, obtains the temperature difference between the environment and the line surface and the total harmonic distortion ratio, formulates a set of fuzzy rules for fuzzy reasoning, determines the leakage fault detection result, improves the accuracy of leakage fault detection, the early fault warning ability, and reduces the probability of false alarms and missed alarms of the system;

[0118] Embodiment 3

[0119] Please refer to Figure 2 , a leakage fault detection system for a low-voltage AC distribution network, including a point selection module, a sensor matching module, a sensor screening module, and a fault detection module;

[0120] The point selection module is used to preferentially select multiple preselected nodes in the initialization stage based on historical data. For each preselected node, collect and synthesize the safety distance coefficient and sensor compatibility ratio corresponding to the preselected points of each preselected node, obtain the point classification coefficient, compare it with the preset point threshold, perform secondary screening on the preselected points, obtain the key points, and send them to the sensor matching module;

[0121] The sensor matching module is used to obtain the key points and connect the sensors corresponding to the points, perform power-on detection, obtain the drift error and DC component, substitute them into the logistic regression algorithm to obtain the sensor screening coefficient, and send it to the sensor screening module;

[0122] The sensor screening module is used to obtain the sensor screening coefficient, obtain the sensor screening threshold through binary recursion, obtain the retained sensors and deleted sensors, select the retained sensors for retention, and send the retained sensors to the fault detection module;

[0123] The fault detection module is used to obtain the retained sensors, detect the environmental characteristic information of each retained sensor and the harmonic characteristics generated by the non-linear load, obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio, and use fuzzy logic to determine the leakage fault detection result.

[0124] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0125] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0126] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context.

[0127] In this application, "at least one" means one or more, and "multiple" means two or more. "At least one of the following" or its similar expressions refer to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0128] It should be understood that in various embodiments of the present application, the magnitudes of the sequence numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0129] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by the combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0130] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0131] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0132] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0133] In addition, the functional units in each embodiment of this application can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0134] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0135] As described above, the above are only specific implementation manners of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application shall be subject to the protection scope of the claims.

Claims

1. A method for detecting leakage faults in a low-voltage AC distribution network, characterized by: include: S1: Based on historical data, multiple pre-selected nodes are selected in the initialization phase. For each pre-selected node, the safety distance coefficient and sensor compatibility ratio corresponding to the pre-selected points of each pre-selected node are collected and integrated to obtain the point classification coefficient. The coefficient is then compared with the preset point classification coefficient threshold, and the pre-selected points are screened twice to obtain the key points. In the initialization phase, historical data is dispatched to collect the safety distance coefficient and sensor compatibility ratio of each pre-selected node according to the pre-selected nodes; The actual distance between the nearest object and the pre-selected node is obtained by calculating the ratio with the preset safety distance. Safety distance coefficient of pre-selected nodes ;in is the pre-selected node index, which is a positive integer; By presetting the existence of different categories of sensors, the number of sensor categories that each pre-selected node is compatible with is counted, and the ratio of the number of compatible sensor categories to the total number of sensor categories is calculated to obtain the first Sensor compatibility ratio of pre-selected nodes ; The safety distance coefficient and sensor compatibility ratio of each pre-selected node are standardized and weighted to obtain the point classification coefficient of each pre-selected node. The number of point classification coefficients is consistent with the number of pre-selected nodes. Compare the point classification coefficient with the preset point classification coefficient threshold. If the point classification coefficient is greater than or equal to the point classification coefficient threshold, the current pre-selected point is marked as a key point. If the point classification coefficient is less than the point classification coefficient threshold, the current pre-selected point is marked as a deleted point and the deleted point is deleted. S2: Obtain key points and connect sensors at corresponding points to perform power-on detection to obtain drift error and DC component, which are then substituted into the logistic regression algorithm to obtain the sensor screening coefficient. Substituting the DC component and drift error into the logistic regression calculation, the specific formula is expressed as follows: ； Where, is the result of logistic regression calculation, that is, the sensor screening coefficient, e is the natural base, and y is the linear combination term of the logistic regression model. Specifically, y can be set as: ； Where, is the bias term, and are the regression coefficients of the DC component and the drift error respectively; S3: Obtain the sensor screening coefficient and obtain the sensor screening threshold through binary recursion, obtain the retained sensors and deleted sensors, and select the retained sensors for retention; S4: Obtaining the retained sensors, detecting the environmental characteristic information of each retained sensor and the harmonic characteristics generated by the nonlinear load, and obtaining the temperature difference between the environment and the line surface and the total harmonic distortion ratio; S5: Obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio, substitute them into fuzzy logic for fuzzy reasoning, and obtain the leakage fault detection result.

2. A method for detecting leakage faults in a low-voltage AC distribution network according to claim 1, characterized in that: Sensors are set up according to the location information of key points. After the key points are connected to the sensors, the electrical signal input is tested. Set the sensor to run under constant leakage current for 24 hours, collect the voltage output by the sensor, and calculate the ratio of the absolute value of the subtraction between the maximum output voltage value and the minimum output voltage value measured by the sensor within 24 hours and the rated output value within the theoretical maximum measurement range of the sensor to obtain the first Drift error of each sensor ,in, is the sensor index, which is a positive integer; By setting a complete cycle, the sensor is used to perform regular sampling of the sawtooth wave, recording the waveform value at each time point, summing the waveform values of all sampling points within the cycle, and calculating the ratio with the total number of sampling points to obtain the first The DC component of each sensor .

3. A method for detecting leakage faults in a low-voltage AC distribution network according to claim 2, characterized in that: The DC component and drift error are standardized and substituted into the logistic regression calculation to obtain the sensor screening coefficient.

4. A method for detecting leakage faults in a low-voltage AC distribution network according to claim 3, characterized in that: By collecting historical data of sensors and sensor screening coefficients, a candidate sensor list is created. The search range is determined by setting the minimum and maximum values of the sensor screening threshold. A recursive function is defined to obtain the sensor screening threshold. If the sensor screening coefficient is greater than or equal to the sensor screening threshold, the sensor corresponding to the current key point is marked as a deleted sensor and an end signal is generated; If the sensor screening coefficient is less than the sensor screening threshold, the sensor corresponding to the current key point is marked as a reserved sensor and a reservation signal is generated; The sensor corresponding to the current key point marked as the retained sensor is retained, and the sensor corresponding to the current key point marked as the deleted sensor is deleted.

5. A method for detecting leakage faults in a low-voltage AC distribution network according to claim 4, characterized in that: Use a temperature sensor to collect the ambient temperature in the distribution network, use an infrared temperature sensor to measure the line surface temperature, subtract the ambient temperature from the line surface temperature to get the temperature difference between the environment and the line surface; The current waveform signal is sampled and Fourier transformed to obtain a frequency domain signal. The first frequency component and remaining subharmonics are extracted from the frequency domain signal, and the remaining subharmonics are marked as high-order harmonics. The ratio of the square sum of the current components of all high-order harmonics to the first frequency component is calculated to obtain the total harmonic distortion ratio.

6. A method for detecting leakage faults in a low-voltage AC distribution network according to claim 5, characterized in that: The temperature difference between the environment and the line surface and the total harmonic distortion ratio are defined as input variables, which are divided into different fuzzy sets respectively. The leakage fault detection result is defined as the output variable and divided into fuzzy sets; Formulate fuzzy rules to describe the impact of the temperature difference between the environment and the line surface and the total harmonic distortion ratio on the leakage fault detection results; Fuzzy reasoning is performed based on fuzzy rules to determine the leakage fault detection results.

7. A low-voltage AC distribution network leakage fault detection system, applied to a low-voltage AC distribution network leakage fault detection method according to any one of claims 1 to 6, characterized in that: Including point screening module, sensor matching module, sensor screening module, fault detection module; The point screening module is used to select multiple pre-selected nodes based on historical data during the initialization phase. For each pre-selected node, the safety distance coefficient and sensor compatibility ratio corresponding to the pre-selected points of each pre-selected node are collected and integrated to obtain the point classification coefficient. The coefficient is then compared with the preset point threshold, and the pre-selected points are screened again to obtain key points, which are then sent to the sensor matching module. The sensor matching module is used to obtain key points and connect sensors at corresponding points, perform power-on detection, obtain drift error and DC component, substitute them into the logistic regression algorithm to obtain sensor screening coefficients, and send them to the sensor screening module; The sensor screening module is used to obtain the sensor screening coefficient and obtain the sensor screening threshold through binary recursion, obtain the retained sensors and deleted sensors, select the retained sensors for retention, and send the retained sensors to the fault detection module; The fault detection module is used to obtain the retained sensors, detect the environmental characteristic information of each retained sensor and the harmonic characteristics generated by the nonlinear load, obtain the temperature difference between the environment and the line surface and the total harmonic distortion ratio, and use fuzzy logic to determine the leakage fault detection result.

Images