System for monitoring line loss rate abnormity of power distribution network station area

Abstract

The invention discloses a system for monitoring line loss rate abnormity of a power distribution network station area, and relates to the technical field of power distribution network monitoring. The line loss rate abnormity monitoring system for the power distribution network station area comprises a data acquisition module, a data processing module, a line loss rate calculation module, an abnormity identification and alarm module and an on-site troubleshooting and diagnosis module. Through integrating functions of intelligent electric meter data acquisition, power grid data collection and area division, data processing, line loss rate calculation, abnormity identification and alarm, on-site troubleshooting and diagnosis and the like, comprehensive and real-time monitoring of the line loss rate of the power distribution network station area is realized. And a regional line loss calculation model is established according to the characteristics of the power distribution network of each region, and the line loss rate can be monitored in real time, dynamically calculated and predicted, so that the efficiency and accuracy of line loss management are greatly improved. And the abnormity identification and alarm module can timely discover and give an alarm of line loss abnormity, provides probability analysis of abnormity reasons, and is helpful for operation and maintenance personnel to quickly respond and process.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network monitoring, and particularly to a system for abnormally monitoring the line loss rate of a distribution network substation area. Background Art

[0002] The main purpose of abnormally monitoring the line loss rate of a distribution network substation area is to timely discover and handle line loss abnormal problems, reduce power grid losses, and improve the economy and safety of power grid operation. Through monitoring, the line loss situation of the substation area can be mastered, providing a basis for formulating targeted loss reduction measures.

[0003] For example, a method, device, and equipment for monitoring line loss power based on the current data of a gateway meter disclosed in the publication number CN116298501A. The method includes: obtaining the main meter power and the sub-meter current set, calculating the theoretical line loss based on the sub-meter current set; obtaining the sub-meter power set, calculating the statistical line loss based on the main meter power and the sub-meter power set; calculating the management line loss based on the statistical line loss and the theoretical line loss; if the management line loss is greater than a preset line loss threshold, taking the acquisition time of the main meter power as the monitoring start time, performing management line loss monitoring from the monitoring start time until the management line loss is less than or equal to the line loss threshold to obtain the monitoring end time; determining the line loss abnormal time period according to the monitoring start time and the monitoring end time; and performing index verification on a pre-constructed sub-meter set based on the line loss abnormal time period to obtain abnormal sub-meters. It improves the monitoring efficiency of user power consumption data.

[0004] The above existing technologies mainly judge whether there is an abnormality by establishing a threshold, calculating the line loss rate, and comparing the threshold; however, due to differences in power grid structures, load characteristics, etc. in different substation areas or even different regions of the same substation area, using the existing line loss rate calculation model and establishing a unified threshold may not fully adapt to the actual situations of all substation areas, which may lead to a deviation between the calculated line loss rate and the actual value, affecting the accuracy of abnormal identification. And there are also many fault reasons for the abnormal line loss rate. After detecting the abnormal line loss rate, generally, it is directly to go to the site to check and find the reasons, which will undoubtedly greatly reduce the maintenance efficiency. Summary of the Invention

[0005] Aiming at the deficiencies of the existing technology, the present invention provides a system for abnormally monitoring the line loss rate of a distribution network substation area, which solves the above technical problems.

[0006] To achieve the above objectives, the present invention is realized through the following technical solutions: A system for abnormally monitoring the line loss rate of a distribution network substation area, including:

[0007] A data acquisition module: using intelligent meters to collect voltage and current data in real time, and dividing the substation area into multiple sub-regions;

[0008] Data processing module: Clean abnormal and invalid data, convert the format and store it;

[0009] Line loss rate calculation module: Considering the differences in the characteristics of the distribution network in different regions of the substation area, establish a regional line loss calculation model. Based on the power consumption data, calculate the line loss rates of each sub-region and the entire substation area respectively, and conduct real-time monitoring, dynamic calculation and line loss rate prediction;

[0010] Abnormal identification and alarm module: Dynamically set thresholds according to the characteristics of the region, compare the real-time line loss rate with the thresholds, identify abnormalities and automatically alarm, and provide a probability analysis of the cause of the abnormality;

[0011] On-site investigation and diagnosis module: Use the system diagnosis function to initially locate the fault point, and verify it by feeding back data after on-site diagnosis.

[0012] Preferably, the data collected by the data collection module specifically includes:

[0013] Intelligent meter data collection: Use intelligent meters to collect power consumption data in the substation area in real time, including voltage, current, and power factor, and collect data from the main meter of the substation area and each user meter through concentrators and collectors;

[0014] Power grid data collection: Collect data requirements on the power grid structure and load characteristics of the substation area, including power grid topology, equipment parameters, and load curves; Obtain document materials including design drawings, equipment operation records, and load monitoring reports from the power grid management department; Then conduct on-site research and record on-site data such as equipment nameplate information, line routes, and joint conditions;

[0015] Sort out the data, and divide the substation area into multiple sub-regions according to the power grid structure and load characteristics of the substation area, and establish a database in a unified format to store the data.

[0016] Preferably, the content of the line loss rate calculation module specifically includes:

[0017] Line loss rate calculation:

[0018] Divide the substation area into N sub-regions according to factors such as the geographical distribution, load density, and line route of the power grid in the substation area;

[0019] For each sub-region, set the power supply as O i and the power sold as S i ;

[0020] The formula for the line loss power of sub-region i is: ΔO i = O i - S i ;

[0021] Line loss rate calculation formula:

[0022] The total bus loss rate of the entire substation area is calculated by the weighted average method:

[0023] Loss calculation:

[0024] Considering the comprehensive influence of active and reactive power losses, the calculation formulas for active power loss and reactive power loss are:

[0025]

[0026]

[0027] Among them, ΔP i is the active power loss of sub-region i, I i is the line current, R i is the line resistance, ω is the angular frequency, R i is the line resistance, L i is the line inductance, cosφ i is the load power factor, φ is the power factor angle; Q i is the reactive power transmitted by the line, U i is the rated voltage of the line, X L is the line reactance.

[0028] Preferably, in the loss calculation step, since it is relatively difficult to directly measure the current and power factor of each sub-region, the load data estimation method is used to obtain them. The current estimation method is:

[0029] Estimation based on load power and voltage: If the load power of the sub-region is known, that is, the active power P and reactive power Q, and the average voltage V of the sub-region, then the following formula is used to estimate the current (I):

[0030]

[0031] Among them, the current I, active power P, reactive power Q, and voltage V satisfy P = IVcosφ and Q = IVsinφ;

[0032] Estimation based on the load curve: The load data is given in the form of a time series, and the estimation is carried out by calculating the average value or peak value of the load;

[0033] Statistical estimation based on historical data: If historical load data is available, statistical methods are used to estimate the current or future current value.

[0034] Preferably, the power factor estimation method is:

[0035] Estimation based on load power: When the active power P and reactive power Q of the sub-region are known, the power factor cosφ is directly calculated:

[0036]

[0037] Among them, the power factor cosφ is the ratio of the active power P to the apparent power S = P 2 +Q 2 ;

[0038] Machine learning estimation based on historical data: If historical load data and corresponding power factor data are available, use machine learning algorithms to train a model, and the model estimates the power factor based on new load data;

[0039] Result fusion estimation: Combine the estimation value based on load power and the machine learning estimation value to improve the estimation accuracy; The fusion methods include average, weighted average, or rules specifically set according to the actual situation;

[0040] Result verification and adjustment: Compare the fused power factor estimation value with the actual measurement value to verify the accuracy and reliability of the estimation result; According to the verification result, adjust and optimize the fusion method.

[0041] Preferably, in the calculation formula of the line loss rate calculation module, considering external influence factors and actual power grid influence factors, adjust the load characteristics of the model and optimize the power grid structure, including:

[0042] Load characteristic adjustment:

[0043] Considering the volatility of the load, use time series analysis or machine learning algorithms to predict load changes; Let the load prediction model be f(t), then the future load prediction value is P pred (t) = f(t);

[0044] According to the prediction result, dynamically adjust the parameters including resistance and inductance in the calculation model, and set corresponding adjustment coefficients for each parameter to be adjusted;

[0045] Considering seasonal changes, establish a seasonal load model: Where P base is the base load, A j , T j , P j are the amplitude, phase and period of the j-th seasonal component respectively, and n is the total number of seasons; Then dynamically adjust the parameters including resistance and inductance according to seasonal changes;

[0046] Impact of load change on line loss: Update the load current I a (t) and power factor cosφ a (t) in real time, and calculate the loss: Where, I a(t) represents the load current of the a-th line at time t, and cosφ a (t) represents the power factor of the a-th line at time t, and R a represents the resistance of the a-th line, and X a represents the reactance corresponding to the inductance of the a-th line, and A is the total number of lines;

[0047] Load curve fitting: The load data is fitted into a continuous function P(t) by using polynomial fitting or exponential fitting methods.

[0048] Preferably, the power grid structure optimization includes:

[0049] Refinement based on equipment parameters: According to the actual parameters of transformers, lines, and switchgear in the power grid, the calculation model is refined;

[0050] Influence of reactive power compensation and harmonic filters:

[0051] Power factor adjustment: The influence of the reactive power compensation device is reflected by adjusting the power factor cosφ;

[0052] Harmonic content adjustment: The influence of the harmonic filter is reflected by adjusting the harmonic content;

[0053] Influence of power grid structure changes on line losses: The power grid structure parameters are updated in real time to ensure the accuracy of the calculation model;

[0054] Power grid simulation: The power grid structure is modeled and analyzed by using power grid simulation software, and the loss P sim_loss = the calculation result of the simulation software, and P sim_loss is the loss value calculated by using the power grid simulation software.

[0055] Preferably, in combination with the load characteristics and the changes in the power grid structure, the calculation model is dynamically adjusted and optimized, and the comprehensive loss formula is obtained by adding the losses caused by load changes, the line itself, and the power grid structure changes: P total_loss = P loss (t) + P line_loss + P sim_loss ;

[0056] The machine learning algorithm is used to train and validate the model to improve the accuracy and generalization ability of the model. Let the trained model be g(x), where x is the input feature, and the output is the predicted loss P pred_loss = g(x); The model parameters are optimized by minimizing the error between the predicted loss and the actual loss: min θ ∥P pred_loss - P actual_loss ∥ 2 , where θ is the model parameter, and P actual_lossIndicates the actual loss.

[0057] Preferably, the dynamic threshold setting function of the abnormal identification and alarm module specifically includes:

[0058] Data collection and preprocessing: Collect historical power consumption data and corresponding line loss rate data of the substation area, including data under normal conditions and data of possible abnormal conditions; perform preprocessing on the data, including removing outliers and filling missing values.

[0059] Calculate the basic threshold: Based on historical data, calculate a basic threshold, which is the average value of the historical line loss rate plus one standard deviation to cover the line loss rate under most normal conditions; the basic threshold T base Is expressed as:

[0060] T base = μ + h·σ;

[0061] Where μ is the average value of the historical line loss rate, σ is the standard deviation of the historical line loss rate, and h is an adjustment coefficient to cover most normal data.

[0062] Consider the actual situation of the substation area: According to the power grid structure, load characteristics, and seasonal change factors of the substation area, adjust the basic threshold by introducing some adjustment factors. The adjusted threshold T^ is expressed as:

[0063] Y^ = Y base ·(1 + ∑ i b i ·F i );

[0064] Where b i Is the coefficient of the adjustment factor, and F i Is the corresponding adjustment factor, which is selected and calculated according to the actual situation of the substation area.

[0065] Real-time monitoring and dynamic adjustment: During the real-time monitoring process, dynamically adjust the threshold according to the current power consumption data and line loss rate data; achieve this by calculating the statistics of the current data and comparing it with historical data. The dynamically adjusted threshold T′ is expressed as:

[0066]

[0067] Where β and γ are coefficients used to adjust the difference between the current data and historical data; Δμ and Δσ are the differences between the current data and historical data in terms of the average value and standard deviation, respectively.

[0068] Preferably, the content of abnormal identification by the abnormal identification and alarm module specifically includes:

[0069] Integrate abnormal factors: Based on historical data, classify the identified abnormal factors and establish a set D = (D 1 , D 2 ,..., D Z ) that contains all the occurred abnormal factors. And for the real-time grid abnormal data detected when any one of the abnormal factors D z occurs, establish a set d = (d 1 , d 2 ,..., d y );

[0070] Match performance characteristics: During actual monitoring, match the actually monitored performance characteristics with the abnormal factors and their subsets d in set D, and preliminarily judge the possible abnormal factors according to the matching degree;

[0071] Calculate the probability of abnormal factors: According to the quantity and degree of the matched performance characteristics, as well as the frequencies of each abnormal factor in historical data, calculate the probability of occurrence of each abnormal factor;

[0072] Sorting and screening: Sort the calculated probabilities of abnormal factors, and screen out the most likely abnormal factors; According to the sorting result, preferentially conduct further investigation and confirmation on the abnormal factors with higher probabilities.

[0073] The present invention provides a system for monitoring abnormal line loss rate in a distribution network substation area. Compared with the prior art, it has the following beneficial effects:

[0074] 1. The system for monitoring abnormal line loss rate in a distribution network substation area realizes comprehensive and real-time monitoring of the line loss rate in the distribution network substation area by integrating functions such as intelligent meter data acquisition, power grid data collection and regional division, data processing, line loss rate calculation, abnormal identification and alarm, and on-site investigation and diagnosis. By establishing a regional line loss calculation model according to the characteristics of the distribution network in each region, it can monitor the line loss rate in real time, calculate it dynamically, and make predictions, which greatly improves the efficiency and accuracy of line loss management. The abnormal identification and alarm module can timely detect and alarm line loss abnormalities, and provide a probability analysis of the abnormal reasons, which helps the operation and maintenance personnel to respond and handle quickly.

[0075] 2. The system for abnormal monitoring of line loss rate in distribution network substations can more accurately calculate the line loss rate of distribution network substations by comprehensively considering the dynamic changes of load characteristics and grid structure, as well as the influence of reactive power compensation and harmonic filters. By using time series analysis and machine learning algorithms for load forecasting and equipment parameter adjustment, and grid simulation software for modeling and analyzing the grid structure, it ensures that the calculation model can reflect the actual operation status of the grid in real time. At the same time, this module also improves the estimation accuracy of key parameters such as current and power factor by integrating multiple estimation methods and result verification and adjustment steps, thus further enhancing the accuracy of line loss rate calculation. In addition, using machine learning algorithms to train and verify the model not only improves the accuracy and generalization ability of the model, but also provides strong data support for future grid optimization and decision-making.

[0076] 3. The system for abnormal monitoring of line loss rate in distribution network substations calculates the basic threshold based on historical data and introduces an adjustment factor to consider the actual situation of the substation area, making the threshold closer to the actual operation condition and effectively improving the accuracy of abnormal identification. Moreover, the real-time monitoring and dynamic adjustment function enables the threshold to adaptively adjust with the change of the grid operation status, avoiding false alarms or missed alarms caused by fixed thresholds, and further enhancing the practicability and stability of the system. This enables the abnormal identification and alarm module to more accurately identify abnormal line loss rate situations and send alarm signals in a timely manner, providing strong guarantee for the safe operation of the distribution network.

[0077] 4. The system for abnormal monitoring of line loss rate in distribution network substations can quickly match the performance characteristics in actual monitoring and initially locate the abnormality by integrating historical abnormal factors and establishing a detailed data set. By combining the occurrence frequency and matching degree of abnormal factors, calculating the abnormal probability and sorting and screening, it effectively improves the accuracy and efficiency of identification. This method not only reduces false alarms and missed alarms, but also provides clear guidance for on-site investigation, ensuring the safe and stable operation of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 is the system principle block diagram of the present invention;

[0079] Figure 2 is the schematic diagram of the system framework steps of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0080] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying 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. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0081] Refer to Figure 1 - Figure 2 , the present invention provides the following six technical solutions:

[0082] The first implementation method: A system for abnormally monitoring the line loss rate of a distribution network substation area, comprising:

[0083] Data acquisition module: Utilize smart meters to collect voltage and current data in real time, and divide the substation area into multiple sub-regions; specifically including:

[0084] Smart meter data acquisition: Utilize smart meters to collect power consumption data in the substation area in real time, including voltage, current, and power factor, and collect data of the main meter of the substation area and each user meter through concentrators and collectors;

[0085] Power grid data acquisition: Acquire data requirements of the power grid structure and load characteristics of the substation area, including power grid topology, equipment parameters (such as transformer capacity, line resistance, inductance, etc.), and load curves; Obtain document materials including design drawings, equipment operation records, and load monitoring reports from the power grid management department; Then conduct on-site research and record on-site data such as equipment nameplate information, line routing, and joint conditions;

[0086] Sort out the data, and divide the substation area into multiple sub-regions according to the power grid structure and load characteristics of the substation area, and establish a database with a unified format to store the data;

[0087] Data processing module: Clean abnormal and invalid data, convert the format and store it;

[0088] Line loss rate calculation module: Consider the differences in the characteristics of the distribution network in different regions of the substation area, establish a regional line loss calculation model, and based on the power consumption data, calculate the line loss rates of each sub-region and the entire substation area respectively, and conduct real-time monitoring, dynamic calculation, and line loss rate prediction;

[0089] Abnormality identification and alarm module: Dynamically set thresholds according to the characteristics of the region, compare the real-time line loss rate with the thresholds, identify abnormalities and automatically alarm, and provide a probability analysis of the reasons for the abnormalities;

[0090] On-site investigation and diagnosis module: Use the system diagnosis function to initially locate the fault point, and verify it by feeding back data after on-site diagnosis.

[0091] The system realizes the comprehensive and real-time monitoring of the line loss rate of the distribution network substation area by integrating functions such as intelligent electricity meter data acquisition, power grid data collection and regional division, data processing, line loss rate calculation, abnormal identification and alarm, and on-site investigation and diagnosis. By establishing a regional line loss calculation model according to the characteristics of the distribution network in each region, it can monitor the line loss rate in real time, calculate it dynamically, and make predictions, which greatly improves the efficiency and accuracy of line loss management. The abnormal identification and alarm module can detect and alarm line loss abnormalities in a timely manner, and provide a probability analysis of the causes of abnormalities, which helps the operation and maintenance personnel to respond and handle quickly.

[0092] The second implementation method: In this embodiment, the content of the line loss rate calculation module specifically includes:

[0093] Line loss rate calculation:

[0094] According to factors such as the geographical distribution, load density, and line orientation of the substation area power grid, the substation area is divided into N sub-regions;

[0095] For each sub-region, the power supply is set to O i , and the electricity sold is set to S i ;

[0096] The formula for the line loss electricity of sub-region i is: ΔO i = O i - S i ;

[0097] Line loss rate calculation formula:

[0098] The total line loss rate of the entire substation area is calculated by the weighted average method:

[0099] Loss calculation:

[0100] Considering the comprehensive influence of active and reactive power losses, the calculation formulas for active and reactive power losses are:

[0101] ΔP i = I i 2 R i + ωL i I i 2 i sin 2 φ i ;

[0102]

[0103] Among them, ΔP i is the active power loss of sub-region i, I i is the line current, R iis the line resistance, ω is the angular frequency, R i is the line resistance, L i is the line inductance, cosφ i is the load power factor, φ is the power factor angle; Q i is the reactive power transmitted by the line, U i is the rated voltage of the line, X L is the line reactance.

[0104] In the loss calculation steps, since it is difficult to directly measure the current and power factor of each sub-region, they are obtained through the load data estimation method. The current estimation method is as follows:

[0105] Estimation based on load power and voltage: If the load power of the sub-region is known, that is, the active power P and the reactive power Q, and the average voltage V of the sub-region, then the following formula is used to estimate the current (I):

[0106]

[0107] Among them, the relationship between the current I, the active power P, the reactive power Q, and the voltage V satisfies P = IVcosφ and Q = IVsinφ;

[0108] Estimation based on the load curve: The load data is given in the form of a time series (for example, the load value per hour or per minute), and the estimation is carried out by calculating the average value or peak value of the load;

[0109] Statistical estimation based on historical data: If historical load data is available, statistical methods (such as regression analysis, time series analysis, etc.) are used to estimate the current or future current value.

[0110] The power factor estimation method is as follows:

[0111] Estimation based on load power: When the active power P and the reactive power Q of the sub-region are known, the power factor cosφ is directly calculated:

[0112]

[0113] Among them, the power factor cosφ is the ratio of the active power P to the apparent power S = P 2 +Q 2 ;

[0114] Machine learning estimation based on historical data: If historical load data and the corresponding power factor data are available, machine learning algorithms (such as neural networks, support vector machines, etc.) are used to train a model, and the model estimates the power factor according to the new load data;

[0115] Result fusion estimation: Fuse the estimation value based on load power and the machine learning estimation value to improve the estimation accuracy; the fusion methods are average, weighted average or rules specifically set according to the actual situation.

[0116] Result verification and adjustment: Compare the fused power factor estimation value with the actual measurement value to verify the accuracy and reliability of the estimation result; according to the verification result, adjust and optimize the fusion method to improve the future estimation accuracy.

[0117] In the calculation formula of the line loss rate calculation module, considering external influence factors and actual power grid influence factors, adjust the load characteristics and optimize the power grid structure of the model, including:

[0118] Load characteristic adjustment:

[0119] Considering the volatility of the load, use time series analysis or machine learning algorithms to predict the load change; algorithm selection: Select time series analysis algorithms such as LSTM (Long Short-Term Memory Network) or ARIMA (Autoregressive Integrated Moving Average Model), or machine learning algorithms such as random forest and support vector machine for load prediction; assume the load prediction model is f(t), then the future load prediction value is P pred (t) = f(t);

[0120] According to the prediction result, dynamically adjust the parameters including resistance and inductance in the calculation model, and set corresponding adjustment coefficients for each parameter to be adjusted; such as R adj (t) = R base ×k(P pred (t)), where k is the adjustment coefficient;

[0121] Considering seasonal changes, establish a seasonal load model: where P base is the base load, A j , T j , P j are the amplitude, phase and period of the j-th seasonal component respectively, and n is the total number of seasons; then dynamically adjust the parameters including resistance and inductance according to seasonal changes, such as L adj (t) = L base ×(1 + α·P seasonal (t)), where α is the adjustment ratio;

[0122] Influence of load change on line loss: Update the load current I a (t) and power factor cosφ a (t) in real time, and calculate the loss: where, I a (t) represents the load current of the a-th line at time t, cosφ a(t) represents the power factor of the a-th line at time t, and R a represents the resistance of the a-th line, and X a represents the reactance corresponding to the inductance of the a-th line (in an AC circuit, inductance generates reactance, and its magnitude is related to the inductance value and frequency). A is the total number of lines;

[0123] Load curve fitting: The load data is fitted into a continuous function P(t) using polynomial fitting or exponential fitting methods;

[0124] Power grid structure optimization includes:

[0125] Refinement based on equipment parameters: According to the actual parameters of transformers, lines, and switchgear in the power grid (such as resistance R, inductance L, capacity S, etc.), the calculation model is refined; The line loss calculation formula is expressed as: P line_loss = I 2 ·R, where I is the line current and R is the line resistance;

[0126] Influence of reactive power compensation and harmonic filters:

[0127] Power factor adjustment: The influence of the reactive power compensation device is reflected by adjusting the power factor cosφ, such as Q comp = P·(tanφ before - tanφ after ), where Q comp represents the reactive power compensation amount provided by the reactive power compensation device, and tanφ before and tanφ after represent the tangent values of the power factor before and after reactive power compensation, respectively;

[0128] Harmonic content adjustment: The influence of the harmonic filter is reflected by adjusting the harmonic content, such as where I k is the k-th harmonic current;

[0129] Influence of power grid structure changes on line loss: The power grid structure parameters, such as line resistance R, inductance L, etc., are updated in real time to ensure the accuracy of the calculation model.

[0130] Power grid simulation: The power grid structure is modeled and analyzed using power grid simulation software (such as ETAP, PSASP, etc.), and the loss P sim_loss = simulation software calculation result, and P sim_loss is the loss value calculated using the power grid simulation software;

[0131] Combined with the changes in load characteristics and power grid structure, the calculation model is dynamically adjusted and optimized, and the comprehensive loss formula is obtained by adding the losses caused by load changes, the line itself, and power grid structure changes: P total_loss = Ploss (t) + P line_loss +P sim_loss ;

[0132] Using machine learning algorithms to train and validate the model to improve the accuracy and generalization ability of the model. Let the trained model be g(x), where x is the input feature (such as load, power grid structure parameters, etc.), and the output is the predicted loss P pred_loss = g(x); Optimize the model parameters by minimizing the error between the predicted loss and the actual loss: min θ ∥P pred_loss - P actual_loss ∥ 2 , where θ is the model parameter, and P actual_loss represents the actual loss.

[0133] By comprehensively considering the dynamic changes of load characteristics and power grid structure, as well as the influence of reactive power compensation and harmonic filters, this line loss rate calculation module can calculate the line loss rate of the distribution network substation area more accurately. Using time series analysis and machine learning algorithms for load forecasting and equipment parameter adjustment, and using power grid simulation software to model and analyze the power grid structure, it ensures that the calculation model can reflect the actual operation state of the power grid in real time. At the same time, this module also improves the estimation accuracy of key parameters such as current and power factor by integrating multiple estimation methods and result verification and adjustment steps, thereby further improving the accuracy of line loss rate calculation. In addition, using machine learning algorithms to train and validate the model not only improves the accuracy and generalization ability of the model, but also provides strong data support for future power grid optimization and decision-making.

[0134] The third implementation method: In this embodiment, the dynamic setting threshold function of the abnormal recognition and alarm module specifically includes:

[0135] Data collection and preprocessing: Collect the historical power consumption data and corresponding line loss rate data of the substation area, including data under normal conditions and possible abnormal situation data; Perform preprocessing on the data including removing outliers and filling missing values;

[0136] Calculate the basic threshold: Based on historical data, calculate a basic threshold, which is the average value of the historical line loss rate plus a standard deviation (or other statistic) to cover the line loss rate under most normal conditions; The basic threshold T base is expressed as:

[0137] T base = μ + h·σ;

[0138] where μ is the average value of the historical line loss rate, σ is the standard deviation of the historical line loss rate, and h is an adjustment coefficient to cover most normal data;

[0139] Consider the actual situation of the substation area: According to the power grid structure, load characteristics, and seasonal change factors of the substation area, some adjustment factors are introduced to adjust the basic threshold. The adjusted threshold T^ is expressed as:

[0140] Y^ = Y base ·(1 + ∑ i b i ·F i );

[0141] Among them, b i is the coefficient of the adjustment factor, and F i is the corresponding adjustment factor (such as load change rate, seasonal coefficient, etc.), which is selected and calculated according to the actual situation of the substation area;

[0142] Real-time monitoring and dynamic adjustment: During the real-time monitoring process, the threshold is dynamically adjusted according to the current power data and line loss rate data; it is achieved by calculating the statistical quantities (such as average value, standard deviation, etc.) of the current data and comparing them with historical data. The dynamically adjusted threshold T′ is expressed as:

[0143]

[0144] Among them, β and γ are coefficients used to adjust the difference between the current data and historical data; Δμ and Δσ are the differences between the current data and historical data in terms of average value and standard deviation respectively.

[0145] Calculating the basic threshold based on historical data and introducing adjustment factors to consider the actual situation of the substation area makes the threshold closer to the actual operating conditions, effectively improving the accuracy of abnormal identification. Moreover, the real-time monitoring and dynamic adjustment functions enable the threshold to be adaptively adjusted with the change of the power grid operating state, avoiding false alarms or missed alarms caused by fixed thresholds, and further enhancing the practicality and stability of the system. It enables the abnormal identification and alarm module to more accurately identify the abnormal situation of the line loss rate and send out alarm signals in a timely manner, providing a strong guarantee for the safe operation of the distribution network.

[0146] The fourth implementation method: In this embodiment, the content of abnormal identification by the abnormal identification and alarm module specifically includes:

[0147] Integrating abnormal factors: Based on historical data, the identified abnormal factors are classified, and a set D = (D 1 , D 2 ,..., D Z ) containing all occurred abnormal factors is established, and for any abnormal factor D z , a set d = (d 1 , d 2 ,..., dy );

[0148] Matching performance characteristics: In actual monitoring, match the actually monitored performance characteristics with the abnormal factors and their subsets d in set D, and preliminarily judge the possible abnormal factors according to the matching degree.

[0149] Calculating the probability of abnormal factors: Calculate the probability of each abnormal factor according to the quantity and degree of the matched performance characteristics, as well as the frequency of each abnormal factor in historical data.

[0150] Sorting and screening: Sort the calculated probabilities of abnormal factors and screen out the most likely abnormal factors; according to the sorting results, prioritize further investigation and confirmation of the abnormal factors with higher probabilities.

[0151] By integrating historical abnormal factors and establishing a detailed data set, the module can quickly match performance characteristics in actual monitoring and initially locate abnormalities. Combining the occurrence frequency and matching degree of abnormal factors, calculating the probability of abnormalities and sorting for screening effectively improves the accuracy and efficiency of identification. This method not only reduces false alarms and missed alarms but also provides clear guidance for on-site investigation, ensuring the safe and stable operation of the distribution network.

[0152] Meanwhile, the content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0153] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device.

[0154] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A line loss rate abnormality monitoring system for a distribution network area, characterized in that: include: Data acquisition module: Use smart meters to collect voltage and current data in real time, and divide the area into multiple sub-areas; Data processing module: clean abnormal and invalid data, convert format and store; Line loss rate calculation module: Considering the differences in distribution network characteristics in different areas of the substation, a regional line loss calculation model is established. Based on the power data, the line loss rate of each sub-area and the entire substation is calculated separately, and real-time monitoring, dynamic calculation and line loss rate prediction are carried out; Abnormal identification and alarm module: dynamically set thresholds based on regional characteristics, compare the real-time line loss rate with the threshold, identify abnormalities and automatically alarm, and provide probability analysis of abnormal causes; On-site troubleshooting and diagnosis module: Use the system diagnosis function to preliminarily locate the fault point, and feedback data for verification after on-site diagnosis.

2. A line loss rate abnormality monitoring system for distribution network area according to claim 1, characterized in that: The data acquisition module specifically collects data including: Smart meter data collection: Use smart meters to collect real-time electricity data in the area, including voltage, current, and power factor. Collect data from the area's total meter and each user's meter through concentrators and collectors; Grid data collection: Collect data on the grid structure and load characteristics of the substation, including grid topology, equipment parameters, and load curves; obtain documents and materials including design drawings, equipment operation records, and load monitoring reports from the grid management department; then conduct field surveys and record on-site data on equipment nameplate information, line directions, and joint conditions; Organize the data, divide the substation into multiple sub-areas according to the grid structure and load characteristics of the substation, and establish a database in a unified format to store data.

3. The line loss rate abnormality monitoring system for distribution network area according to claim 1 is characterized by: The content of the line loss rate calculation module specifically includes: Line loss rate calculation: The substation area is divided into N sub-areas according to the geographical distribution, load density and line direction of the substation area power grid; For each sub-area, set the power supply to 0 i , the electricity sold is S i ; The line loss calculation formula for sub-area i is: ΔO i =O i -S i ; Line loss rate calculation formula: The total bus loss rate of the entire substation area is calculated by the weighted average method: Loss calculation: Considering the comprehensive impact of active and reactive power losses, the calculation formulas for active and reactive power losses are: ΔP i =I i 2 R i +ωL i I i 2 i sin 2 φ i ; Among them, ΔP i is the active power loss of sub-area i, I i is the line current, R i is the line resistance, ω is the angular frequency, R i is the line resistance, L i is the line inductance, cosφ i is the load power factor, φ is the power factor angle; Q i is the reactive power transmitted by the line, U i is the line rated voltage, X L It is the line inductive reactance.

4. A line loss rate abnormality monitoring system for distribution network area according to claim 3, characterized in that: In the loss calculation step, since it is difficult to directly measure the current and power factor of each sub-area, they are obtained through the load data estimation method. The current estimation method is: Estimation based on load power and voltage: If the load power of the sub-area, i.e., active power P and reactive power Q, and the average voltage V of the sub-area are known, the current (I) can be estimated using the following formula: Among them, the current I, active power P, reactive power Q and voltage V satisfy P = IVcosφ and Q = IVsinφ; Estimation based on load curve: Load data is given in the form of time series, and estimation is performed by calculating the average or peak value of the load; Statistical estimation based on historical data: If historical load data is available, use statistical methods to estimate current or future current values.

5. A line loss rate abnormality monitoring system for distribution network area according to claim 4, characterized in that: The power factor estimation method is: Estimation based on load power: When the active power P and reactive power Q of the sub-area are known, the power factor cosφ can be directly calculated: Among them, the power factor cosφ is the active power P and the apparent power S = P 2 +Q 2 ratio; Machine learning estimation based on historical data: If historical load data and corresponding power factor data are available, a machine learning algorithm is used to train a model that estimates the power factor based on the new load data; Result fusion estimation: The estimated value based on load power and the estimated value based on machine learning are integrated to improve the estimation accuracy; the fusion method is average, weighted average or rules set according to actual conditions; Result verification and adjustment: Compare the fused power factor estimation value with the actual measured value to verify the accuracy and reliability of the estimation result; adjust and optimize the fusion method based on the verification result.

6. The line loss rate abnormality monitoring system for distribution network area according to claim 3 is characterized by: In the calculation formula of the line loss rate calculation module, external influencing factors and actual influencing factors of the power grid are considered to adjust the load characteristics of the model and optimize the power grid structure, including: Load characteristic adjustment: Considering the volatility of load, time series analysis or machine learning algorithm is used to predict load changes; assuming the load forecasting model is f(t), the future load forecast value is P pred (t) = f(t); According to the prediction results, dynamically adjust the parameters including resistance and inductance in the calculation model, and set the corresponding adjustment coefficient for each parameter to be adjusted; Considering seasonal changes, a seasonal load model is established: Where P base is the basic load, A j , T j , P j are the amplitude, phase and period of the jth seasonal component, respectively, and n is the total number of seasons; then the parameters including resistance and inductance are dynamically adjusted according to seasonal changes; Impact of load changes on line losses: real-time update of load current I a (t) and power factor cosφ a (t), calculate the loss: Among them, I a (t) represents the load current of line a at time t, cosφ a (t) represents the power factor of line a at time t, R a represents the resistance of the ath line, X a represents the inductive reactance corresponding to the inductance of the ath line, and A is the total number of lines; Load curve fitting: Use polynomial fitting or exponential fitting method to fit the load data into a continuous function P(t).

7. A line loss rate abnormality monitoring system for distribution network area according to claim 6, characterized in that: The grid structure optimization includes: Refinement based on equipment parameters: Refine the calculation model according to the actual parameters of transformers, lines, and switchgear in the power grid; Effects of reactive power compensation and harmonic filter: Power factor adjustment: The influence of reactive power compensation device is reflected by adjusting the power factor cosφ; Harmonic content adjustment: reflect the influence of harmonic filters by adjusting the harmonic content; Impact of grid structure changes on line losses: Update grid structure parameters in real time to ensure the accuracy of the calculation model; Grid simulation: Use grid simulation software to model and analyze the grid structure and calculate the loss P sim_loss = simulation software calculation results, P sim_loss is the loss value calculated using power grid simulation software.

8. The line loss rate abnormality monitoring system for distribution network area according to claim 7 is characterized by: Combined with the changes in load characteristics and grid structure, the calculation model is dynamically adjusted and optimized, and the losses caused by load changes, the line itself, and grid structure changes are added together to obtain the comprehensive loss formula: P total_loss =P loss (t)+P line_loss +P sim_loss ; The machine learning algorithm is used to train and verify the model to improve the accuracy and generalization ability of the model. Suppose the training model is g(x), where x is the input feature, and the output is the prediction loss P pred_loss = g(x); optimize the model parameters by minimizing the error between the predicted loss and the actual loss: min θ ∥P pred_loss -P actual_loss ∥ 2 , where θ is the model parameter, P actual_loss Indicates actual loss.

9. The line loss rate abnormality monitoring system for distribution network area according to claim 1 is characterized by: The dynamic threshold setting function of the abnormality identification and alarm module specifically includes: Data collection and preprocessing: Collect historical power data and corresponding line loss rate data of the substation, including data under normal conditions and possible abnormal conditions; preprocess the data including removing outliers and filling missing values; Calculate the basic threshold: Based on historical data, calculate a basic threshold, which is the average value of the historical line loss rate plus a standard deviation to cover the line loss rate under most normal circumstances; the basic threshold T base It is expressed as: T base =μ+h·σ; Among them, μ is the average value of historical line loss rate, σ is the standard deviation of historical line loss rate, and h is an adjustment coefficient to cover most normal data; Considering the actual situation of the substation: According to the grid structure, load characteristics, and seasonal variation factors of the substation, some adjustment factors are introduced to adjust the basic threshold. The adjusted threshold T^ is expressed as: Y^=Y base ·(1+∑ i b i ·F i ); Among them, b i is the coefficient of the adjustment factor, F i is the corresponding adjustment factor, which is selected and calculated according to the actual situation of the substation; Real-time monitoring and dynamic adjustment: In the real-time monitoring process, the threshold is dynamically adjusted according to the current power data and line loss rate data. This is achieved by calculating the statistics of the current data and comparing it with the historical data. The dynamically adjusted threshold T′ is expressed as: Among them, β and γ are coefficients used to adjust the difference between current data and historical data; Δμ and Δσ are the differences between current data and historical data in terms of mean and standard deviation, respectively.

10. The line loss rate abnormality monitoring system for distribution network area according to claim 1 is characterized by: The abnormality identification and alarm module specifically identifies abnormalities including: Integrate abnormal factors: Based on historical data, classify the identified abnormal factors and establish a set D = (D1, D2, ..., D Z ), and for any abnormal factor D z The real-time abnormal data of the power grid detected when it occurs establishes a set d = (d1, d2, ..., d y ); Matching performance characteristics: In actual monitoring, the performance characteristics actually monitored are matched with the abnormal factors in set D and its subset d, and the possible abnormal factors are preliminarily judged according to the matching degree; Calculate the probability of abnormal factors: Calculate the probability of each abnormal factor based on the number and degree of matched performance characteristics and the frequency of each abnormal factor in historical data; Sorting and screening: Sort the calculated abnormal factor probabilities and screen out the most likely abnormal factors; based on the sorting results, prioritize the abnormal factors with higher probabilities for further investigation and confirmation.

Images