Line loss checking and detecting system for power distribution network station area

Abstract

The invention relates to the technical field of electric power systems, in particular to a line loss checking and detecting system for a power distribution network area, and aims to solve the problems that in the prior art, line loss abnormity cannot be detected in real time, problems cannot be found and processed in time, the operation risk of a power grid cannot be effectively reduced, and multi-dimensional data cannot be comprehensively considered. And the comprehensiveness and accuracy of detection cannot be ensured. According to the invention, the anomaly detection module guarantees the scientificity and reliability of a detection result through statistical analysis based on historical data, achieves automatic detection through a machine learning model, remarkably reduces the manual intervention, improves the detection efficiency, can detect line loss anomaly in real time, timely discovers and processes problems, effectively reduces the operation risk of a power grid, and improves the reliability of the power grid. The method can adapt to different scenes and requirements by adjusting the preset threshold value, improves the applicability of the model, and guarantees the comprehensiveness and accuracy of detection by comprehensively considering the multi-dimensional data.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and more specifically to a line loss troubleshooting and detection system for a distribution network substation. Background Art

[0002] Traditional methods for troubleshooting line losses in distribution network substations have problems such as insufficient detection accuracy, unreasonable resource utilization, and delayed maintenance response. With the development of smart grid and big data technology, higher requirements are placed on the accuracy and efficiency of line loss detection.

[0003] The patent application with reference publication number CN116298675A discloses a distribution network line loss anomaly detection method and system based on intelligent algorithm, which relates to the technical field of power system, connects a data interaction device to the distribution network, collects and obtains distribution data of the distribution network, performs line loss calculation to obtain multiple line loss calculation indicators, performs load condition feature extraction to obtain condition feature extraction results, collects and obtains basic information of the distribution network, performs big data screening and extraction, builds an anomaly detection model, inputs multiple line loss calculation indicators and condition feature extraction results into the anomaly detection model, outputs anomaly detection results, and manages the distribution network line loss according to the anomaly detection results; the present invention solves the technical problem that the distribution network line loss detection in the prior art usually requires detailed production line original parameters, resulting in large detection calculation amount and low efficiency, and realizes the construction of anomaly detection model based on intelligent algorithm, so as to achieve the technical effect of improving the detection accuracy of line loss anomalies and reducing the calculation amount and complexity of the detection process;

[0004] However, the above-mentioned reference patent achieves efficient and accurate line loss anomaly detection and management by collecting distribution data and basic information, extracting load condition characteristics and building anomaly detection models, which significantly reduces the amount of calculation and complexity. However, it cannot detect line loss anomalies in real time, cannot discover and deal with problems in a timely manner, cannot effectively reduce the risk of power grid operation, cannot comprehensively consider multi-dimensional data, cannot guarantee the comprehensiveness and accuracy of detection, and cannot accurately locate high line loss areas and time periods. It cannot guarantee the scientificity and reliability of the results, cannot comprehensively consider the two dimensions of region and time period, and cannot guarantee the comprehensiveness and systematicness of the analysis.

[0005] To this end, we propose a line loss inspection and detection system for distribution network substations to address the above problems. Summary of the invention

[0006] The purpose of the present invention is to provide a line loss investigation and detection system for distribution network substations, which solves the problems that the prior art cannot detect line loss anomalies in real time, cannot discover and deal with problems in time, cannot effectively reduce the risk of power grid operation, cannot comprehensively consider multi-dimensional data, cannot guarantee the comprehensiveness and accuracy of detection, and cannot accurately locate high line loss areas and time periods, cannot guarantee the scientificity and reliability of the results, cannot comprehensively consider the two dimensions of region and time period, and cannot guarantee the comprehensiveness and systematicness of the analysis.

[0007] The purpose of the present invention is achieved through the following technical solutions:

[0008] A line loss investigation and detection system for a distribution network substation, comprising a data acquisition preprocessing module, a line loss calculation module, an anomaly detection module, a positioning analysis module and an intelligent optimization module;

[0009] The data collection and preprocessing module is used to collect the operation data of the distribution network substation and perform preprocessing operations on the collected operation data;

[0010] The line loss calculation module calculates the theoretical line loss and actual line loss of the distribution network area based on the pre-processed operation data;

[0011] The anomaly detection module is used to collect historical line loss data, build an abnormal line loss detection model, and use the model to detect whether there is abnormal line loss in the distribution network area;

[0012] The location analysis module is used to locate and analyze the detected abnormal line loss and identify the high line loss areas and time periods;

[0013] The intelligent optimization module is used to automatically take corresponding measures for optimization according to the line loss investigation results of the distribution network substation.

[0014] As a preferred implementation of the present invention, the specific process of the line loss calculation module calculating the theoretical line loss and actual line loss of the distribution network station area is as follows:

[0015] Acquire pre-processed operation data, the operation data including operation voltage, operation current, operation power, operation quantity and conductor temperature;

[0016] The conductor resistance in the distribution network area will change with temperature. The conductor resistance is calculated by the following formula:

[0017] R(T)=R 20 ×[1+α(T-20)];

[0018] Where R 20 is the resistance of the wire at 20°C, α is the temperature coefficient of the wire, T is the wire temperature, and R(T) is the resistance of the wire at temperature T;

[0019] R is calculated by the following formula 20 :

[0020] Where ρ is the conductor resistivity, L is the conductor length, and A is the conductor cross-sectional area;

[0021] Theoretical line loss is the power line loss caused by the wire resistance and current. The theoretical line loss is calculated by the following formula:

[0022] P 理论 =3×I 2 ×R(T)×t;

[0023] Where I is the operating current, t is the operating time, P 理论 is the theoretical line loss.

[0024] As a preferred embodiment of the present invention, the actual line loss is the difference between the input power and the output power, and the actual line loss is calculated by the following formula:

[0025]

[0026] Where P 实际 is the actual line loss, P is the operating power, E is the operating electricity, V is the operating voltage, and cos(φ) is the power factor;

[0027] The line loss rate is the ratio of actual line loss to input power. The line loss rate is calculated using the following formula:

[0028]

[0029] As a preferred implementation of the present invention, the specific process of the abnormality detection module detecting whether there is abnormal line loss in the distribution network station area is as follows:

[0030] Collect the duration of the distribution network substation in the past and mark it as a time threshold. Divide the time threshold into z sub-time periods, where z is a natural number greater than zero. Obtain the historical line loss data of the distribution network substation within the time threshold. The line loss data includes input power, output power, theoretical line loss, actual line loss, and line loss rate.

[0031] The input power of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the obtained multiple input powers is calculated, and the arithmetic average of the multiple input powers is recorded as the average input power PGR;

[0032] The output power of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the obtained multiple output powers is calculated, and the arithmetic average of the multiple output powers is recorded as the average output power PGC.

[0033] As a preferred embodiment of the present invention, the theoretical line loss of the distribution network area in multiple sub-time periods is obtained, and the arithmetic average of the multiple theoretical line losses is calculated, and the arithmetic average of the multiple theoretical line losses is recorded as the average theoretical line loss PLX;

[0034] The actual line loss of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the multiple actual line losses is calculated, and the arithmetic average of the multiple actual line losses is recorded as the average actual line loss PSX;

[0035] The line loss rate of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic mean of the multiple line loss rates obtained is calculated, and the arithmetic mean of the multiple line loss rates is recorded as the average line loss rate PXS.

[0036] As a preferred embodiment of the present invention, the average input power PGR, the average output power PGC, the average theoretical line loss PLX, the average actual line loss PSX and the average line loss rate PXS are obtained, and the average input power PGR, the average output power PGC, the average theoretical line loss PLX, the average actual line loss PSX and the average line loss rate PXS are combined to construct a feature matrix X;

[0037] The constructed feature matrix X is used as the input of the machine learning model, and the line loss detection assessment value is used as the output of the machine learning model. The line loss detection assessment value is used as the prediction target, and minimizing the sum of prediction errors of all training data is used as the training target. The machine learning model is trained until the sum of prediction errors reaches convergence and the training is stopped to obtain the abnormal line loss detection model.

[0038] As a preferred implementation of the present invention, the abnormal line loss detection model is expressed as follows:

[0039] XJP=λ1·PGR+λ2·PGC+λ3·PLX+λ4·PSX+λ5·PXS+ε;

[0040] Where XJP is the line loss detection assessment value, λ1, λ2, λ3, λ4 and λ5 are all regression coefficients, and ε represents the error term;

[0041] Obtain real-time line loss data, convert it into the corresponding feature matrix and input it into the abnormal line loss detection model. The real-time line loss detection evaluation value XJP is obtained through the abnormal line loss detection model. 实时 , the real-time line loss detection assessment value XJP 实时 Compare with the preset line loss detection assessment value threshold:

[0042] If XJP 实时 If it is less than the preset line loss detection assessment value threshold, it indicates that the line loss in the distribution network substation is normal;

[0043] If XJP实时 If it is greater than or equal to the preset line loss detection assessment value threshold, it indicates that there is abnormal line loss in the distribution network substation.

[0044] As a preferred implementation of the present invention, the specific process of the positioning analysis module performing positioning analysis on the detected abnormal line loss is as follows:

[0045] When abnormal line loss is detected in the distribution network area, the real-time line loss data of the distribution network area is obtained, and the distribution network area is divided into multiple areas based on the operation status of the distribution network area. The line loss difference of each area is calculated by the following formula:

[0046] Where ΔP 区域 is the line loss difference in a certain area, ΔP u is the line loss difference of the uth line in the area, v is the total number of lines in the area;

[0047] The line loss difference of the uth line is calculated by the following formula:

[0048] ΔP u =|P 实际u -P 理论u |;

[0049] Where P 实际u is the actual line loss of the uth line, P 理论u is the theoretical line loss of the uth line;

[0050] Compare the line loss differences in different areas, and use the following expression to represent the high line loss area:

[0051] High line loss area = max(△P 区域1 , △P 区域2 ,…,△P 区域w ), where the high line loss area is the area with the highest line loss difference value, and w is the total number of areas.

[0052] As a preferred implementation of the present invention, the power consumption of the distribution network is divided into multiple time periods based on the power consumption of the distribution network area in one day, and the line loss difference of each time period is calculated by the following formula:

[0053] Where ΔP 时间段 is the line loss difference in a certain period of time, ΔP j is the line loss difference of the jth line in the time period, k is the total number of lines in the time period;

[0054] Compare the line loss differences in different time periods, and use the following expression to represent the high line loss period:

[0055] High line loss time period = max(ΔP 时间段1 , △P时间段2 ,…,△P 时间段o ), where the high line loss time period is the time period with the highest line loss difference value, and o is the total number of time periods.

[0056] As a preferred implementation of the present invention, the specific process of the intelligent optimization module taking corresponding measures to perform optimization processing according to the line loss investigation result is as follows:

[0057] Obtain the line loss investigation results of the distribution network substation area, which include whether there are abnormal line losses, high line loss areas and time periods;

[0058] When it is detected that the line loss in the distribution network area is normal, corresponding measures are automatically taken to optimize the process;

[0059] When abnormal line loss is detected in the distribution network area, corresponding measures are automatically taken to optimize the process;

[0060] When high line loss areas are identified, corresponding measures are automatically taken to optimize the process;

[0061] When high line loss periods are identified, corresponding measures are automatically taken to optimize the process.

[0062] Compared with the prior art, the advantages of the present invention are:

[0063] (1) In the present invention, the anomaly detection module ensures the scientificity and reliability of the detection results through statistical analysis based on historical data, realizes automatic detection through machine learning models, significantly reduces manual intervention, improves detection efficiency, can detect line loss anomalies in real time, discover and handle problems in a timely manner, and effectively reduce the risk of power grid operation. By adjusting the preset threshold, it can adapt to different scenarios and needs, improve the applicability of the model, and ensure the comprehensiveness and accuracy of detection by comprehensively considering multi-dimensional data;

[0064] (2) In the present invention, the positioning and analysis module can accurately locate high line loss areas and time periods by dividing the areas and time periods, and perform calculations and analysis based on real-time line loss data to ensure the scientificity and reliability of the results. It comprehensively considers the two dimensions of area and time period to ensure the comprehensiveness and systematicness of the analysis. Through automated calculation and comparison, it significantly improves the analysis efficiency and reduces manual intervention. It can adjust the division of areas and time periods according to actual needs to adapt to different application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a system block diagram of Embodiment 1 of the present invention;

[0066] Figure 2 This is a system block diagram of Embodiment 2 of the present invention;

[0067] Figure 3 This is a logical flow diagram of the first embodiment of the present invention. DETAILED DESCRIPTION

[0068] The following will combine 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 the embodiments. All other embodiments obtained by ordinary technicians in this field without creative work based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0069] Embodiment 1: Figure 1 and Figure 3 As shown, the present invention proposes a line loss investigation and detection system for a distribution network substation, including a data acquisition preprocessing module, a line loss calculation module, an anomaly detection module, a positioning analysis module and an intelligent optimization module;

[0070] The data collection and preprocessing module is used to collect the operation data of the distribution network substation and perform preprocessing operations on the collected operation data. The preprocessing operations include data cleaning, format conversion and normalization processing;

[0071] The data collection preprocessing module significantly improves system performance through data cleaning, format conversion and normalization. Data cleaning removes noise and outliers, improves data quality, and reduces misjudgments. Format conversion unifies multi-source heterogeneous data, improves consistency and processing efficiency. Normalization eliminates dimensional effects and optimizes model performance and stability. These operations provide high-quality data support for subsequent modules, reduce operation and maintenance costs, reduce manual intervention, and lay the foundation for intelligent analysis, ensuring efficient and stable operation of the system.

[0072] The line loss calculation module calculates the theoretical line loss and actual line loss of the distribution network area based on the pre-processed operation data;

[0073] The specific process of calculating the theoretical line loss and actual line loss of the distribution network area by the line loss calculation module is as follows:

[0074] Acquire pre-processed operation data, the operation data including operation voltage, operation current, operation power, operation quantity and conductor temperature;

[0075] The conductor resistance in the distribution network area will change with temperature. The conductor resistance is calculated by the following formula:

[0076] R(T)=R 20 ×[1+α(T-20)];

[0077] Where R 20is the resistance of the wire at 20°C, α is the temperature coefficient of the wire, T is the wire temperature, and R(T) is the resistance of the wire at temperature T;

[0078] R is calculated by the following formula 20 :

[0079] Where ρ is the conductor resistivity, L is the conductor length, and A is the conductor cross-sectional area;

[0080] Theoretical line loss is the power line loss caused by the wire resistance and current. The theoretical line loss is calculated by the following formula:

[0081] P 理论 =3×I 2 ×R(T)×t;

[0082] Where I is the operating current, t is the operating time, P 理论 is the theoretical line loss;

[0083] The actual line loss is the difference between the input power and the output power. The actual line loss is calculated using the following formula:

[0084]

[0085] Where P 实际 is the actual line loss, P is the operating power, E is the operating electricity, V is the operating voltage, and cos(φ) is the power factor;

[0086] The line loss rate is the ratio of actual line loss to input power. The line loss rate is calculated using the following formula:

[0087]

[0088] The line loss calculation module significantly improves the accuracy of line loss calculation by considering the change of conductor resistance with temperature; comprehensively considers multiple factors such as operating voltage, current, power, and electricity to ensure the comprehensiveness of calculation; the calculation of actual line loss and line loss rate provides a scientific basis for grid efficiency evaluation and optimization; the automated calculation based on preprocessed data improves efficiency, reduces manual intervention, and lays the foundation for intelligent grid management.

[0089] The anomaly detection module is used to collect historical line loss data, build an abnormal line loss detection model, and use the model to detect whether there is abnormal line loss in the distribution network area;

[0090] The specific process of the abnormal detection module detecting whether there is abnormal line loss in the distribution network area is as follows:

[0091] Collect the duration of the distribution network substation in the past and mark it as a time threshold. Divide the time threshold into z sub-time periods, where z is a natural number greater than zero. Obtain the historical line loss data of the distribution network substation within the time threshold. The line loss data includes input power, output power, theoretical line loss, actual line loss, and line loss rate.

[0092] The input power of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the obtained multiple input powers is calculated, and the arithmetic average of the multiple input powers is recorded as the average input power PGR;

[0093] The output power of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the obtained multiple output powers is calculated, and the arithmetic average of the multiple output powers is recorded as the average output power PGC;

[0094] Obtain theoretical line losses in a distribution network substation in multiple sub-time periods, calculate an arithmetic average of the multiple theoretical line losses, and record the arithmetic average of the multiple theoretical line losses as an average theoretical line loss PLX;

[0095] The actual line loss of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the multiple actual line losses is calculated, and the arithmetic average of the multiple actual line losses is recorded as the average actual line loss PSX;

[0096] Obtain the line loss rate of the distribution network substation in multiple sub-time periods, calculate the arithmetic average of the multiple line loss rates obtained, and record the arithmetic average of the multiple line loss rates as the average line loss rate PXS;

[0097] The average input power PGR, the average output power PGC, the average theoretical line loss PLX, the average actual line loss PSX and the average line loss rate PXS are obtained, and the average input power PGR, the average output power PGC, the average theoretical line loss PLX, the average actual line loss PSX and the average line loss rate PXS are combined to construct a feature matrix X;

[0098] The constructed feature matrix X is used as the input of the machine learning model, and the line loss detection assessment value is used as the output of the machine learning model. The line loss detection assessment value is used as the prediction target, and the sum of the prediction errors of all training data is minimized as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the training is stopped to obtain the abnormal line loss detection model.

[0099] The expression formula of abnormal line loss detection model is as follows:

[0100] XJP=λ1·PGR+λ2·PGC+λ3·PLX+λ4·PSX+λ5·PXS+ε;

[0101] Where XJP is the line loss detection assessment value, λ1, λ2, λ3, λ4 and λ5 are all regression coefficients, and ε represents the error term;

[0102] Obtain real-time line loss data, convert it into the corresponding feature matrix and input it into the abnormal line loss detection model. The real-time line loss detection evaluation value XJP is obtained through the abnormal line loss detection model. 实时 , the real-time line loss detection assessment value XJP 实时 Compare with the preset line loss detection assessment value threshold:

[0103] If XJP 实时 If it is less than the preset line loss detection assessment value threshold, it indicates that the line loss in the distribution network substation is normal;

[0104] If XJP 实时 If it is greater than or equal to the preset line loss detection assessment value threshold, it indicates that there is abnormal line loss in the distribution network substation area;

[0105] The anomaly detection module ensures the scientificity and reliability of the detection results through statistical analysis based on historical data. It realizes automated detection through machine learning models, significantly reduces manual intervention and improves detection efficiency. It can detect line loss anomalies in real time, discover and handle problems in a timely manner, and effectively reduce the risk of power grid operation. By adjusting the preset thresholds, it can adapt to different scenarios and needs, improve the applicability of the model, and ensure the comprehensiveness and accuracy of detection by comprehensively considering multi-dimensional data.

[0106] Embodiment 2: The technical solution of the embodiment of the present invention is different from that of embodiment 1 in that:

[0107] like Figure 2 As shown, the positioning analysis module is used to perform positioning analysis on the detected abnormal line loss and identify the high line loss area and time period;

[0108] The specific process of the location analysis module for locating and analyzing the detected abnormal line loss is as follows:

[0109] When abnormal line loss is detected in the distribution network area, the real-time line loss data of the distribution network area is obtained, and the distribution network area is divided into multiple areas based on the operation status of the distribution network area. The line loss difference of each area is calculated by the following formula:

[0110] Where ΔP 区域 is the line loss difference in a certain area, ΔP u is the line loss difference of the uth line in the area, v is the total number of lines in the area;

[0111] The line loss difference of the uth line is calculated by the following formula:

[0112] ΔP u =|P实际u -P 理论u |;

[0113] Where P 实际u is the actual line loss of the uth line, P 理论u is the theoretical line loss of the uth line;

[0114] Compare the line loss differences in different areas, and use the following expression to represent the high line loss area:

[0115] High line loss area = max(△P 区域1 , △P 区域2 ,…,△P 区域w ), where the high line loss area is the area with the highest line loss difference value, and w is the total number of areas;

[0116] The power consumption of the distribution network is divided into multiple time periods based on the power consumption of the distribution network area in one day. The line loss difference in each time period is calculated by the following formula:

[0117] Where ΔP 时间段 is the line loss difference in a certain period of time, ΔP j is the line loss difference of the jth line in the time period, k is the total number of lines in the time period;

[0118] Compare the line loss differences in different time periods, and use the following expression to represent the high line loss period:

[0119] High line loss time period = max(ΔP 时间段1 , △P 时间段2 ,…,△P 时间段o ), where the high line loss time period is the time period with the highest line loss difference value, and o is the total number of time periods;

[0120] The positioning and analysis module can accurately locate high line loss areas and time periods by dividing them into regions and time periods, and perform calculations and analyses based on real-time line loss data to ensure the scientific nature and reliability of the results. It comprehensively considers both the region and time period dimensions to ensure the comprehensiveness and systematic nature of the analysis. Through automated calculation and comparison, it significantly improves analysis efficiency and reduces manual intervention. It can adjust the division of regions and time periods according to actual needs to adapt to different application scenarios.

[0121] Intelligent optimization module, which is used to automatically take corresponding measures for optimization according to the line loss investigation results of the distribution network substation;

[0122] The specific process of the intelligent optimization module taking corresponding measures to optimize the processing according to the line loss investigation results is as follows:

[0123] Obtain the line loss investigation results of the distribution network substation area, which include whether there are abnormal line losses, high line loss areas and time periods;

[0124] When it is detected that the line loss in the distribution network area is normal, corresponding measures are automatically taken for optimization. The specific contents of the optimization measures are as follows:

[0125] Regularly inspect the equipment in the distribution network area, adjust the operation mode of the transformer according to the load conditions, gradually replace inefficient equipment and adopt high-efficiency energy-saving equipment;

[0126] When abnormal line loss is detected in the distribution network area, corresponding measures are automatically taken to optimize the process. The specific contents of the optimization measures are as follows:

[0127] Investigate electricity theft in areas with abnormal line loss, repair or replace faulty equipment, and check and calibrate metering equipment in areas with abnormal line loss;

[0128] When a high line loss area is identified, corresponding measures are automatically taken to optimize the process. The specific contents of the optimization measures are as follows:

[0129] Replace the conductors in high-loss areas, use low-resistivity materials, install reactive power compensation devices in high-loss areas, adjust the load distribution in high-loss areas, and replace aging transformers, switches or connectors in high-loss areas;

[0130] When a high line loss period is identified, corresponding measures are automatically taken to optimize the process. The specific contents of the optimization measures are as follows:

[0131] Implement load peak shaving and valley filling strategies during high-loss periods, dynamically adjust the output of reactive power compensation devices during high-loss periods, and adjust the operation mode of transformers during high-loss periods;

[0132] The intelligent optimization module significantly improves the efficiency and stability of power grid operation by automatically obtaining line loss investigation results and taking targeted measures. When line loss is normal, the module ensures efficient operation of the power grid by inspection, adjusting transformer operation mode and replacing inefficient equipment. When line loss is abnormal, the module quickly investigates power theft, repairs faulty equipment and calibrates metering equipment to reduce line loss risks. For areas with high line loss, the module reduces regional line loss by replacing conductors, installing reactive power compensation devices and adjusting load distribution. For periods of high line loss, the module implements a load peak shaving and valley filling strategy and optimizes transformer operation mode to reduce period line loss.

[0133] Working principle of the present invention: When in use, in the present invention, the anomaly detection module ensures the scientificity and reliability of the detection results through statistical analysis based on historical data, realizes automatic detection through machine learning models, significantly reduces manual intervention, improves detection efficiency, can detect line loss anomalies in real time, discover and deal with problems in time, and effectively reduce the risk of power grid operation. By adjusting the preset threshold, it can adapt to different scenarios and needs, and improve the applicability of the model. By comprehensively considering multi-dimensional data, the comprehensiveness and accuracy of the detection are guaranteed, and the positioning analysis module can accurately locate high line loss areas and time periods through the division of regions and time periods, and calculate and analyze based on real-time line loss data to ensure the scientificity and reliability of the results. The two dimensions of region and time period are comprehensively considered to ensure the comprehensiveness and systematicness of the analysis. Through automatic calculation and comparison, the analysis efficiency is significantly improved, and manual intervention is reduced. The division of regions and time periods can be adjusted according to actual needs to adapt to different application scenarios.

[0134] The above are only preferred specific implementation modes of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solutions and improved concepts of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A line loss detection system for distribution network substations, characterized in that: It includes data acquisition preprocessing module, line loss calculation module, anomaly detection module, positioning analysis module and intelligent optimization module; The data collection and preprocessing module is used to collect the operation data of the distribution network substation and perform preprocessing operations on the collected operation data; The line loss calculation module calculates the theoretical line loss and actual line loss of the distribution network area based on the pre-processed operation data; The anomaly detection module is used to collect historical line loss data, build an abnormal line loss detection model, and use the model to detect whether there is abnormal line loss in the distribution network area; The location analysis module is used to locate and analyze the detected abnormal line loss and identify the high line loss areas and time periods; The intelligent optimization module is used to automatically take corresponding measures for optimization according to the line loss investigation results of the distribution network substation.

2. A line loss detection system for distribution network area according to claim 1, characterized in that: The specific process of calculating the theoretical line loss and actual line loss of the distribution network area by the line loss calculation module is as follows: Acquire pre-processed operation data, the operation data including operation voltage, operation current, operation power, operation quantity and conductor temperature; The conductor resistance in the distribution network area will change with temperature. The conductor resistance is calculated using the following formula: R(T)=R 20 ×[1+α(T-20)]; Where R 20 is the resistance of the wire at 20°C, α is the temperature coefficient of the wire, T is the wire temperature, and R(T) is the resistance of the wire at temperature T; R is calculated by the following formula 20 : Where ρ is the conductor resistivity, L is the conductor length, and A is the conductor cross-sectional area; Theoretical line loss is the power line loss caused by the wire resistance and current. The theoretical line loss is calculated by the following formula: P 理论 =3×I 2 ×R(T)×t; Where I is the operating current, t is the operating time, P 理论 is the theoretical line loss.

3. A line loss detection system for distribution network substations according to claim 2, characterized in that: The actual line loss is the difference between the input power and the output power. The actual line loss is calculated using the following formula: Where P 实际 is the actual line loss, P is the operating power, E is the operating electricity, V is the operating voltage, and cos(φ) is the power factor; The line loss rate is the ratio of actual line loss to input power. The line loss rate is calculated using the following formula:

4. A line loss detection system for distribution network substations according to claim 1, characterized in that: The specific process of the abnormal detection module detecting whether there is abnormal line loss in the distribution network area is as follows: Collect the duration of the distribution network substation in the past and mark it as a time threshold. Divide the time threshold into z sub-time periods, where z is a natural number greater than zero. Obtain the historical line loss data of the distribution network substation within the time threshold. The line loss data includes input power, output power, theoretical line loss, actual line loss, and line loss rate. The input power of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the obtained multiple input powers is calculated, and the arithmetic average of the multiple input powers is recorded as the average input power PGR; The output power of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the obtained multiple output powers is calculated, and the arithmetic average of the multiple output powers is recorded as the average output power PGC.

5. A line loss detection system for distribution network substations according to claim 4, characterized in that: Obtain theoretical line losses in a distribution network substation in multiple sub-time periods, calculate an arithmetic average of the multiple theoretical line losses, and record the arithmetic average of the multiple theoretical line losses as an average theoretical line loss PLX; The actual line loss of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic average of the multiple actual line losses is calculated, and the arithmetic average of the multiple actual line losses is recorded as the average actual line loss PSX; The line loss rate of the distribution network substation in multiple sub-time periods is obtained, and the arithmetic mean of the multiple line loss rates obtained is calculated, and the arithmetic mean of the multiple line loss rates is recorded as the average line loss rate PXS.

6. A line loss detection system for distribution network substations according to claim 5, characterized in that: The average input power PGR, the average output power PGC, the average theoretical line loss PLX, the average actual line loss PSX and the average line loss rate PXS are obtained, and the average input power PGR, the average output power PGC, the average theoretical line loss PLX, the average actual line loss PSX and the average line loss rate PXS are combined to construct a feature matrix X; The constructed feature matrix X is used as the input of the machine learning model, and the line loss detection assessment value is used as the output of the machine learning model. The line loss detection assessment value is used as the prediction target, and minimizing the sum of prediction errors of all training data is used as the training target. The machine learning model is trained until the sum of prediction errors reaches convergence and the training is stopped to obtain the abnormal line loss detection model.

7. A line loss detection system for distribution network substations according to claim 6, characterized in that: The expression formula of abnormal line loss detection model is as follows: XJP=λ1·PGR+λ2·PGC+λ3·PLX+λ4·PSX+λ5·PXS+ε; Where XJP is the line loss detection assessment value, λ1, λ2, λ3, λ4 and λ5 are all regression coefficients, and ε represents the error term; Obtain real-time line loss data, convert it into the corresponding feature matrix and input it into the abnormal line loss detection model. The real-time line loss detection evaluation value XJP is obtained through the abnormal line loss detection model. 实时 , the real-time line loss detection assessment value XJP 实时 Compare with the preset line loss detection assessment value threshold: If XJP 实时 If it is less than the preset line loss detection assessment value threshold, it indicates that the line loss in the distribution network substation is normal; If XJP 实时 If it is greater than or equal to the preset line loss detection assessment value threshold, it indicates that there is abnormal line loss in the distribution network substation.

8. The line loss detection system for distribution network area according to claim 1 is characterized in that: The specific process of the positioning analysis module performing positioning analysis on the detected abnormal line loss is as follows: When abnormal line loss is detected in the distribution network area, the real-time line loss data of the distribution network area is obtained, and the distribution network area is divided into multiple areas based on the operation status of the distribution network area. The line loss difference of each area is calculated by the following formula: Where ΔP 区域 is the line loss difference in a certain area, ΔP u is the line loss difference of the uth line in the area, v is the total number of lines in the area; The line loss difference of the uth line is calculated by the following formula: ΔP u =|P 实际u -P 理论u |; Where P 实际u is the actual line loss of the uth line, P 理论u is the theoretical line loss of the uth line; Compare the line loss differences in different areas, and use the following expression to represent the high line loss area: High line loss area = max(△P 区域1 , △P 区域2 ,…,△P 区域w ), where the high line loss area is the area with the highest line loss difference value, and w is the total number of areas.

9. A line loss detection system for distribution network substations according to claim 8, characterized in that: The power consumption of the distribution network is divided into multiple time periods based on the power consumption of the distribution network area in one day. The line loss difference in each time period is calculated by the following formula: Where ΔP 时间段 is the line loss difference in a certain period of time, ΔP j is the line loss difference of the jth line in the time period, k is the total number of lines in the time period; Compare the line loss differences in different time periods, and use the following expression to represent the high line loss period: High line loss time period = max(ΔP 时间段1 , △P 时间段2 ,…,△P 时间段o ), where the high line loss time period is the time period with the highest line loss difference value, and o is the total number of time periods.

10. A line loss detection system for distribution network substations according to claim 1, characterized in that: The specific process of the intelligent optimization module taking corresponding measures to optimize the processing according to the line loss investigation results is as follows: Obtain the line loss investigation results of the distribution network substation area, which include whether there are abnormal line losses, high line loss areas and time periods; When it is detected that the line loss in the distribution network area is normal, corresponding measures are automatically taken to optimize the process; When abnormal line loss is detected in the distribution network area, corresponding measures are automatically taken to optimize the process; When high line loss areas are identified, corresponding measures are automatically taken to optimize the process; When high line loss periods are identified, corresponding measures are automatically taken to optimize the process.

Images