Power grid theoretical line loss measuring and calculating method based on automatic data acquisition

Abstract

The invention relates to the technical field of power grid theoretical line loss measurement and calculation based on automatic data acquisition, and discloses a power grid theoretical line loss measurement and calculation method based on automatic data acquisition. Power grid topology and measurement data are obtained, and branch equivalent resistance is automatically collected and calculated; constructing a de-weighting and descending current threshold sequence; generating a sub-graph according to a threshold value; constructing a generation forest for each sub-graph, and extracting a base cycle formed by a branch and a unique path; performing deduplication and merging to obtain a whole network unique base cycle set; calculating an average energy loss on the basis of each cycle; the cycle durability is measured in a normalized mode according to the occurrence frequency; and carrying out weighted accumulation on the cycle energy metric values of all cycle energy output networks as a theoretical line loss estimation result. Through automatic acquisition and graph theory modeling, full-process structured measurement and calculation are realized, and multi-dimensional cycle energy and persistence indexes are introduced, so that evaluation precision and real-time performance are remarkably improved, experience dependence is reduced, and operation scheduling optimization and abnormal energy consumption identification are facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid theoretical line loss calculation based on automatic data collection, and in particular to a power grid theoretical line loss calculation method based on automatic data collection. Background Art

[0002] Grid line loss estimation is fundamental to power system operation and management, and its accuracy directly impacts energy metering, operational economics, and grid dispatch optimization. Traditional line loss estimation methods rely on grid state estimation or empirical formulas based on static models. A typical approach involves comprehensively analyzing transmission line parameter presets, power flow calculation results, and load curves, combined with system-level static power flow calculations, to estimate the current and voltage distribution of each branch. Losses are then calculated based on impedance data. These methods rely primarily on standardized assumptions about line resistance, line length, and load distribution, ignoring the dynamic impact of current fluctuations and line topology on energy losses during actual operation, which can easily lead to significant errors.

[0003] With the advancement of measurement technology, online line loss monitoring and dynamic estimation methods based on real-time data are gaining increasing application. Leveraging current, voltage, and active power data collected by SCADA systems, improved state estimation algorithms and data fusion techniques enable real-time assessment of overall network line loss. However, existing state estimation-based technologies often rely on accurate network parameter models, complex iterative algorithms, and continuous adjustments by professional operations and maintenance personnel. They also struggle to adapt promptly to network segmentation, topology changes, or equipment failures, and are unable to fully utilize high-frequency, large-scale measurement data to reveal the multi-scale and multi-modal energy loss characteristics of power grid operation. In recent years, graph theory methods have garnered attention in power grid analysis. Researchers have attempted to model power grids as node-edge structures and analyze energy transmission paths using graph connectivity, spanning trees, and loop detection. However, most work focuses solely on reachability analysis, shortest paths, or network reliability assessment, lacking a quantitative characterization of loop energy loss. Furthermore, these methods typically analyze a single network snapshot, failing to consider the impact of current threshold variations on loop structure or to incorporate the persistence of loops at different operating levels to measure the depth and stability of energy loss. Another type of method uses signal processing techniques such as spectral analysis or wavelet decomposition to decouple and analyze the current and voltage time series signals in the frequency domain or time-frequency domain to extract the loss characteristics corresponding to the high-frequency and low-frequency components. However, this type of method requires a preset window function or filtering threshold for the signal, and the analysis results are often based on energy spectrum density or frequency band power, which is difficult to directly map to the physical loop structure of the actual network, limiting its application in large-scale power grids. In addition, the access of distributed power sources, flow reversal, and switching of various operating conditions in modern power grids make the distribution of current between branches present complex spatiotemporal characteristics. When dealing with such dynamic, multi-level network circulation energy characteristics, existing technologies often lack a unified multi-threshold, multi-resolution analysis framework, and do not deeply combine measurement data with network topology to systematically identify and quantify the contribution of various loops to line losses at different operating levels.

[0004] To this end, this project aims to propose a method for calculating theoretical power line losses in power grids based on automated data collection. Through graph modeling, current data analysis, multi-threshold subgraph decomposition, and cycle energy calculation, this method achieves efficient and accurate assessment of theoretical line losses in power grids. Based on real-time data such as current and voltage, a grid topology model is automatically constructed. Subgraphs are filtered and base cycle sets are extracted at different current thresholds. Using average loss and loop persistence as quantitative indicators, an estimated network-wide cycle energy value is ultimately derived to characterize the theoretical line loss level. Summary of the Invention

[0005] The present invention provides a method for calculating theoretical line loss of a power grid based on automatic data collection, which helps solve the problems mentioned in the above background technology.

[0006] The present invention provides the following technical solution: a method for calculating theoretical line loss of a power grid based on automatic data collection, comprising:

[0007] Obtain the grid topology and related measurement data, including the set of measurement points at each node in the grid and the set of connected branches, automatically collect the absolute value of the branch current and the voltage difference between the two ends of the branch, and calculate the equivalent resistance of the branch based on this;

[0008] A current threshold sequence is constructed based on the collected current data, and the current set is deduplicated and arranged in descending order to obtain multiple current thresholds;

[0009] For each current threshold, a corresponding threshold subgraph is constructed, which contains all branches with current greater than the threshold;

[0010] Construct a spanning forest for each threshold subgraph and extract the corresponding base cycle set based on the spanning forest. The base cycle is a simple loop formed by a branch and a unique path connecting the two end nodes of the branch in the spanning forest.

[0011] The base cycle sets extracted from each threshold subgraph are deduplicated and merged to obtain a unique base cycle set for the entire network.

[0012] Calculate the average loss energy based on each basis cycle;

[0013] The persistence of each base cycle is measured and normalized based on its frequency of occurrence at different thresholds to obtain the corresponding persistence ratio;

[0014] The energy indicators of all base cycles are weighted and accumulated to output the overall cycle energy measurement value of the power grid as the estimation result of the theoretical line loss.

[0015] Optionally, the obtaining of the grid topology and related measurement data thereof, including a set of measurement points of each node in the grid and a set of connected branches, automatically collecting the absolute value of the branch current and the voltage difference across the branch, and calculating the equivalent resistance of the branch based on the absolute value, specifically includes:

[0016] Assume that the node set of all measurement points in the network is V = {v1,v2,...,v N}; where v i is the i-th node; N is the total number of nodes;

[0017] Suppose the branch set containing all network connection edges is E={e1,e2,...,e M}; where e k is the kth branch; M is the total number of branches;

[0018] Automatic collection branch e k The real-time current absolute value Ik and branch e k Voltage difference ΔU k ;

[0019] Calculate branch e k The equivalent resistance

[0020] Optionally, constructing a current threshold sequence based on the collected current data, removing duplicates from the current set and arranging them in descending order to obtain multiple current thresholds specifically includes:

[0021] Construct current set S = {I k |k=1,2,...,M};

[0022] Remove duplicates from the current set S and sort them in descending order, and get the threshold sequence τ1>τ2>…>τ L ; Among them, {τ j}=unique(S), which is the threshold sequence; L≤M, which is the total number of thresholds; unique(·) is the set deduplication function, which removes duplicate elements.

[0023] Optionally, constructing a corresponding threshold subgraph for each current threshold specifically includes:

[0024] For each threshold τ j Construct a subgraph:

[0025] G j =(V,E j ), E j ={e k ∈E|I k ≥τ j}; Among them, G j is the threshold τ j The subgraph below; E j To include all k ≥τ j branch road.

[0026] Optionally, constructing a spanning forest for each threshold subgraph and extracting a corresponding basis cycle set based on the spanning forest specifically includes:

[0027] For each subgraph G j Construct the generated forest T as follows: j And extract the base cycle set C j :

[0028] S110, E j For the middle branch, press I k Sort from large to small and record it as a sequence Among them, e i,h G jThe branch with the hth largest current; I(e i,h ) is its corresponding current amplitude;

[0029] S120, initialize the generated forest

[0030] S130, add branches one by one, for h=1 to |E j |Execute steps S131 to S132:

[0031] S131, branch e i,h Connecting nodes (u i,h ,v i,h ); where u i,h ,v i,h Branch e i,h The starting point and end point;

[0032] S132, if the edge e i,h Join T j After, u i,h With v i,h In T j is still not connected, then T j ←T j ∪{e i,h};

[0033] S140, let the base cycle edge set be B j =E j \T j ;

[0034] S150, base loop structure:

[0035] For each e∈B j : Among them, e is the connection node u e With v e edge; c j (e) is the forest T generated by branch e j A simple circuit consisting of a unique path connecting its endpoints; For the forest j Connecting node u e With v e The only path;

[0036] S160, forming a base cycle set C j ={c j (e)|e∈B j}.

[0037] Optionally, deduplication and merging of the base cycle sets extracted from the threshold subgraphs to obtain a unique base cycle set for the entire network may include:

[0038] Union all base loops

[0039] For any If the edge sets are the same, that is, {e|e∈c}={e|e∈c'}, then only one copy is retained, and the final base cycle set is obtained

[0040] Optionally, the calculating of the average energy loss based on each base cycle specifically includes:

[0041] For each loop c∈C, perform the following steps:

[0042] S210, calculate the number of loop edges |c|=#{e|e∈c}; where #{·} is a set cardinality operator used for counting;

[0043] S220, calculate average energy loss Among them, I e is the current amplitude of branch e; R e is the equivalent resistance of branch e.

[0044] Optionally, the persistence of each base cycle is measured, and normalized based on its frequency of occurrence under different thresholds to obtain a corresponding persistence ratio, specifically including:

[0045] The threshold level of the statistical loop c n(c) = #{j|c∈C j}, 1≤n(c)≤L;

[0046] Normalized persistence ratio

[0047] Optionally, the weighted accumulation of energy indicators of all base cycles and output of a cycle energy measurement value of the entire power grid as an estimation result of theoretical line loss specifically includes:

[0048] The weighted accumulation of all base cycles gives the overall cycle energy metric of the network:

[0049] P cycle =∑ c∈C E(c)×r(c).

[0050] The present invention has the following beneficial effects:

[0051] 1. By automatically collecting the current and voltage differences of each branch of the power grid and then combining them with a collection of nodes and branches for topological modeling, a seamless integration of the physical structure and state of the power grid is achieved. Traditional line loss measurement often relies on manually constructed network models and designated measurement points, which is highly subjective, inefficient, and lacks support for large-scale and complex power grids. In contrast, this method significantly improves the real-time and accuracy of data acquisition, while providing a solid data foundation for subsequent algorithms. It is highly automated and adaptable, making it particularly suitable for the automated monitoring and high-precision analysis needs of modern smart grids.

[0052] 2. We propose an automatic current threshold generation mechanism based on deduplication and descending sorting of the absolute values ​​of current across the entire network to guide subsequent graph partitioning and loop extraction. This method requires no manual parameter setting and is versatile and scalable. Traditional methods, which mostly rely on static single thresholds or specific partitioning strategies, often struggle to reflect the evolution of network topology under different current levels. The dynamic multi-threshold construction employed in this solution not only enhances the multi-level modeling capabilities of line loss estimation but also increases sensitivity to weak or abnormal loops, laying a solid foundation for subsequent base cycle extraction.

[0053] 3. Valid branches under different current thresholds are screened to form multi-level subgraphs, achieving "multi-scale decomposition" of the power grid in the graph structure. This processing method is more hierarchical and recognizable than a unified analysis of the entire graph, and can effectively discover energy consumption paths that are only active at specific energy flow levels. Traditional methods often uniformly model the entire power grid and lack targeted analysis of the energy flow characteristics of each layer. This solution uses "threshold" as a scale to achieve quantification and stratification of power grid load levels, which has high practical value in energy consumption analysis and fault identification.

[0054] 4. A spanning forest algorithm and graph traversal strategy are introduced to automatically extract physically meaningful base cycles from acyclic structures. Specifically, the algorithm prioritizes branches with larger current values ​​for loop construction, in line with the physical principle that "main channels prioritize energy transmission" in power systems. Compared to traditional graph theory loop detection, this method not only focuses on structural integrity but also incorporates energy flow information, making the extraction results more physically meaningful. This method enhances the ability to identify implicit energy paths in power grids and is particularly suitable for locating abnormal power consumption and analyzing the causes of losses.

[0055] 5. Structural matching and edge set comparison are performed on the loop sets extracted from multiple threshold subgraphs to ultimately generate a unique base loop set for the entire network, avoiding information redundancy and computational duplication. Compared to the "loop duplicate counting" problem existing in existing technologies, this step introduces set equality judgment logic, effectively improving computational efficiency and accuracy. This approach maintains the integrity of physical loop coverage while reducing model complexity, facilitating aggregate modeling for downstream energy assessment and enhancing the stability and maintainability of system calculations.

[0056] 6. Using the current and resistance parameters of each branch within each base cycle as input, a weighted average energy consumption value is calculated, accurately reflecting its contribution to the energy transmission process. Compared to traditional single-branch calculations or node energy conservation methods, this approach better captures energy consumption characteristics at the circuit level, especially in complex networks with diverse circuits and intertwined paths. This method has broad application value in accurately calculating energy consumption and analyzing local overloads in power grids, providing strong support for energy-saving scheduling.

[0057] 7. A circuit persistence assessment mechanism based on frequency of occurrence and a normalized index was designed. This mechanism evaluates circuit activity under multiple thresholds using a joint "time-energy" dimension. Traditional line loss estimation often overlooks circuits that are frequently active but weak at certain "critical current levels." This method, however, addresses this information blind spot by using a persistence index, reflecting the stability and criticality of energy-consuming circuits. This offers significant advantages in optimizing dynamic grid operation and identifying vulnerable structures.

[0058] 8. By weighting the average energy loss of each base cycle and its durability ratio, a comprehensive performance indicator—the network cycle energy metric—is output for theoretical line loss estimation. Compared to the traditional macro-calculation method of subtracting the total node input from the node output, this method can reflect the structural distribution and dynamic fluctuations of energy consumption within the power grid, providing a more granular and explanatory perspective. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0061] Example, see Figure 1 , a method for calculating theoretical line loss of a power grid based on automatic data collection, comprising:

[0062] Obtain the grid topology and related measurement data, including the set of measurement points at each node in the grid and the set of connected branches, automatically collect the absolute value of the branch current and the voltage difference between the two ends of the branch, and calculate the equivalent resistance of the branch based on this;

[0063] A current threshold sequence is constructed based on the collected current data, and the current set is deduplicated and arranged in descending order to obtain multiple current thresholds;

[0064] For each current threshold, a corresponding threshold subgraph is constructed, which contains all branches with current greater than the threshold;

[0065] Construct a spanning forest for each threshold subgraph and extract the corresponding base cycle set based on the spanning forest. The base cycle is a simple loop formed by a branch and a unique path connecting the two end nodes of the branch in the spanning forest.

[0066] The base cycle sets extracted from each threshold subgraph are deduplicated and merged to obtain a unique base cycle set for the entire network.

[0067] Calculate the average loss energy based on each basis cycle;

[0068] The persistence of each base cycle is measured and normalized based on its frequency of occurrence at different thresholds to obtain the corresponding persistence ratio;

[0069] The energy indicators of all base cycles are weighted and accumulated to output the overall cycle energy measurement value of the power grid as the estimation result of the theoretical line loss.

[0070] Through eight steps, including grid topology and measurement data acquisition, current threshold sequence construction, threshold subgraph generation, base cycle extraction, de-duplication merging, energy calculation, persistence measurement, and final energy output, the method systematically achieves automated and structured measurement of theoretical grid line losses. This method addresses the problems of traditional manual analysis, such as modeling difficulties, reliance on expert experience, and poor real-time performance. First, through automatic data acquisition and graph theory modeling, the entire process of complex grid line loss estimation is automated. Second, through cycle energy and persistence indicators, a multi-dimensional and multi-level line loss evaluation mechanism is introduced, which improves the accuracy and physical rationality of the evaluation, making it easier to use for operation and scheduling optimization and abnormal energy consumption identification.

[0071] The acquisition of the grid topology and its related measurement data, including the set of measurement points of each node in the grid and the set of connected branches, automatically collecting the absolute value of the branch current and the voltage difference between the two ends of the branch, and calculating the equivalent resistance of the branch based on the absolute value of the branch current and the voltage difference between the two ends of the branch, specifically includes:

[0072] Establish network structure and basic measurement data to lay the physical and topological foundation for subsequent cycle extraction;

[0073] Assume that the node set of all measurement points in the network is V = {v1,v2,...,v N}; where v i is the i-th node; N is the total number of nodes; represents all measurement points, which is convenient for unified identification and indexing;

[0074] Suppose the branch set containing all network connection edges is E={e1,e2,...,eM}; where e k is the kth branch; M is the total number of branches; represents all connection relationships in the network and is used for graph theory operations;

[0075] Automatic collection branch e k The real-time current absolute value I k and branch e k Voltage difference ΔU k ; Obtain the electrical status of each branch to support energy calculation;

[0076] Calculate branch e k The equivalent resistance The measurement results are converted into resistance parameters and subsequently used for energy calculations.

[0077] By introducing topological structure modeling (node ​​sets and branch sets), automatically collecting real-time electrical parameters (current and voltage), and calculating resistance based on voltage-current, a systematic bridge is established between the physical parameters of the power grid and its graph-theoretic structure. This step addresses issues such as unclear structure, difficult parameter acquisition, and highly subjective manual input in power grid modeling. Converting actual measurement data into key parameters for graph model input provides a solid data foundation for subsequent circuit construction and energy calculations, while also improving the timeliness and adaptability of the data and enabling the system to dynamically adapt to changes in operating conditions.

[0078] The current threshold sequence is constructed based on the collected current data, and the current set is deduplicated and arranged in descending order to obtain multiple current thresholds, specifically including:

[0079] Automatically generate a series of critical thresholds from the measured current for multi-level sub-graph analysis;

[0080] Construct current set S = {I k |k=1,2,...,M}; collect the current of the entire network to prepare for threshold generation;

[0081] Remove duplicates from the current set S and sort them in descending order, and get the threshold sequence τ1>τ2>…>τ L ; Among them, {τ j}=unique(S), which is the threshold sequence; L≤M, which is the total number of thresholds; unique(·) is a set deduplication function that removes duplicate elements; threshold sequences of different current levels are obtained without the need for exogenous parameters.

[0082] By deduplicating and sorting the set of branch current values, a series of thresholds are automatically generated to form a current threshold sequence. This step solves the problem of traditional modeling where threshold setting relies on experience and lacks adaptability. It does not rely on subjective human parameter setting and has greater versatility and automation. High-current loops are often more significant for line losses. This method quantifies current levels and decomposes the power grid into multiple subgraphs with distinct strengths and weaknesses. This helps identify important and marginal loops, enhances the model's resolution, and supports multi-scale analysis and intelligent computing.

[0083] For each current threshold, constructing a corresponding threshold subgraph specifically includes:

[0084] By screening branches at different current levels, a multi-level network subgraph is constructed to capture the timing of loop occurrence;

[0085] For each threshold τ j Construct a subgraph:

[0086] G j =(V,E j ), E j ={e k ∈E|I k ≥τ j}; Among them, G j is the threshold τ j The subgraph below; E j To include all k ≥τ j branch; generated at the current level τ j and above available branch subgraphs.

[0087] By constructing a subgraph based on each current threshold, only branches with currents above the current threshold are retained, forming a set of threshold subgraphs, thus achieving multi-level partitioning of the network structure. This mechanism solves the problems of coarse-grained circuit extraction, high noise, and heavy computational burden in full-graph modeling. Each subgraph level represents a specific energy flow level, allowing observation of changes in the network's internal structure from multiple perspectives, improving the ability to identify critical states of circuit generation. This is particularly suitable for modeling low-voltage distribution networks or scenarios with multi-level loads, making analysis more targeted and scalable.

[0088] The step of constructing a spanning forest for each threshold subgraph and extracting a corresponding basis cycle set based on the spanning forest specifically includes:

[0089] In each subgraph, we select branches to remove loops and extract the most basic loop structure;

[0090] For each subgraph G j Construct the generated forest T as follows: j And extract the base cycle set C j :

[0091] S110, E j For the middle branch, press I k Sort from large to small and record it as a sequence Among them, e i,h G j The branch with the hth largest current; I(e i,h ) is its corresponding current amplitude; the branch with large current is given priority to ensure that the main energy flow channel is emphasized when extracting the circuit;

[0092] S120, initialize the generated forest Prepare an empty set for accumulating acyclic branches;

[0093] S130, add branches one by one, for h=1 to |E j |Execute steps S131 to S132:

[0094] S131, branch e i,h Connecting nodes (u i,h ,v i,h ); where u i,h ,v j,h Branch e i,h The starting point and end point;

[0095] S132, if the edge e i,h Join T j After, u i,h With v i,h In T j is still not connected, then T j ←T j ∪{e i,h};

[0096] Construct the largest forest without loop connectivity;

[0097] S140, let the base cycle edge set be B j =E j \T j ; B j Represents subgraph G j The set of branches that are not added to the generated forest; the branches that are not absorbed by the forest are the edges that form independent loops;

[0098] S150, base loop structure:

[0099] For each e∈B j : Among them, e is the connection node u e With v e edge; c j (e) is the forest T generated by branch e jA simple circuit consisting of a unique path connecting its endpoints; For the forest j Connecting node u e With v e Combine each base cycle edge with the unique path connecting its endpoints in the forest into a simple loop;

[0100] S160, forming a base cycle set C j ={c j (e)|e∈B j};C j Represents the set of all base cycles under threshold level j; summarizes all base cycles and provides a basic unit for network-wide cycle analysis.

[0101] By constructing a spanning forest within each subgraph and extracting a set of base cycles, the team applied traditional graph traversal algorithms to real-world power grid structure analysis, resolving issues such as incomplete cycle extraction and unclear physical meaning of loops. A "preferential selection of high-current branches" strategy ensured that the extracted cycles were more representative of the main energy flow paths. Furthermore, the team used the spanning forest and the remaining edges to construct simple loops, ensuring the structural rationality and uniqueness of the loops. This not only identified a set of closed paths with minimum coverage but also established physically meaningful basic units for subsequent energy calculations, reducing redundancy and error accumulation.

[0102] Deduplication and merging of the base cycle sets extracted from each threshold subgraph to obtain a unique base cycle set for the entire network specifically includes:

[0103] Merge the loops generated at all levels, remove duplicates, and form a unique set of loops;

[0104] Union all base loops Converges all levels of base loops;

[0105] For any If the edge sets are the same, that is, {e|e∈c}={e|e∈c'}, then only one copy is retained, and the final base cycle set is obtained Remove duplicate loops with the same edge set to ensure uniqueness.

[0106] By comparing the edges of cycle sets extracted at multiple thresholds and removing duplicate loops with identical structures, a unique base cycle set is ultimately formed for the entire network. This solves the problem of overlapping loops in multi-level subgraphs, which can lead to overly large calculation results. This improves the compactness of data representation and analysis efficiency, avoiding energy consumption estimation distortions caused by duplicate loops. It also reduces the computational burden of subsequent steps, making the system more efficient and stable, and suitable for large-scale network analysis scenarios.

[0107] The calculation of the average energy loss based on each base cycle specifically includes:

[0108] Quantify the average power loss physically borne by each loop;

[0109] For each loop c∈C, perform the following steps:

[0110] S210, calculate the number of loop edges |c|=#{e|e∈c}; where #{·} is a set cardinality operator used for counting; determine the loop size;

[0111] S220, calculate average energy loss Among them, I e is the current amplitude of branch e; R e is the equivalent resistance of branch e; calculate the average power loss of each branch in the loop.

[0112] By counting the current and resistance of each branch in a circuit and calculating its average energy loss, we can quantitatively assess the energy contribution of each circuit. This step addresses the coarse-grained nature of traditional line loss estimation, which uses nodes or paths as units for energy consumption assessment. First, circuit-level assessment better reflects the distribution of energy consumption within the structure. Second, it provides a quantifiable basis for ranking circuit importance. This approach facilitates the identification of high-loss circuits and enables targeted management for energy-saving optimization.

[0113] The persistence of each base cycle is measured, and normalized based on its frequency of occurrence at different thresholds to obtain a corresponding persistence ratio, specifically including:

[0114] Each circuit was assessed for its degree of consistent occurrence at multiple threshold levels;

[0115] The threshold level of the statistical loop c n(c) = #{j|c∈C j}, 1≤n(c)≤L; the threshold level number of the counting loop;

[0116] Normalized persistence ratio Convert the number of occurrences into a dimensionless persistence measure.

[0117] By counting and normalizing the number of times each circuit appears in multiple threshold subgraphs, a "persistence ratio" metric is generated, introducing temporal and structural stability evaluation in addition to energy consumption. This overcomes the problem of previous line loss analysis that only considers instantaneous power while ignoring circuit activity. This metric reflects whether a circuit is "stable and active" or "occasionally present" throughout the network, providing new data support for circuit importance ranking and operation and maintenance priority allocation, and is a key component in improving system cognitive capabilities.

[0118] The energy indicators of all base cycles are weighted and accumulated to output the cycle energy measurement value of the entire power grid as the estimation result of the theoretical line loss, which specifically includes:

[0119] The total network cycle energy index is obtained by combining the physical energy consumption and durability of each loop;

[0120] The weighted accumulation of all base cycles gives the overall cycle energy metric of the network:

[0121] P cycle =∑ c∈C E(c)×r(c); summarize the "average energy loss × durability" of each circuit and output the final indicator.

[0122] By weightedly summing the average energy loss and durability ratios of all circuits, a circulating energy metric for the entire network is output as a theoretical line loss estimate. This comprehensive metric addresses the difficulties in normalizing line loss results and the lack of a basis for structural decomposition. It maintains the physical authenticity of energy consumption data while reflecting the level of circuit activity within the network. This gives the line loss estimate both "strength" and "stability," making it more suitable for routine monitoring, fault detection, and energy-saving scheduling of power systems.

[0123] It should be noted that, in this document, relational terms such as first and second, etc., are used only 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 terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0124] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for calculating theoretical line loss of a power grid based on automatic data collection, characterized in that: include: Obtain the grid topology and related measurement data, including the set of measurement points at each node in the grid and the set of connected branches, automatically collect the absolute value of the branch current and the voltage difference between the two ends of the branch, and calculate the equivalent resistance of the branch based on this; A current threshold sequence is constructed based on the collected current data, and the current set is deduplicated and arranged in descending order to obtain multiple current thresholds; For each current threshold, a corresponding threshold subgraph is constructed, which contains all branches with current greater than the threshold; Construct a spanning forest for each threshold subgraph and extract the corresponding base cycle set based on the spanning forest. The base cycle is a simple loop formed by a branch and a unique path connecting the two end nodes of the branch in the spanning forest. The base cycle sets extracted from each threshold subgraph are deduplicated and merged to obtain a unique base cycle set for the entire network. Calculate the average loss energy based on each basis cycle; The persistence of each base cycle is measured and normalized based on its frequency of occurrence at different thresholds to obtain the corresponding persistence ratio; The energy indicators of all base cycles are weighted and accumulated to output the overall cycle energy measurement value of the power grid as the estimation result of the theoretical line loss.

2. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 1, characterized in that: The acquisition of the grid topology and its related measurement data, including the set of measurement points of each node in the grid and the set of connected branches, automatically collecting the absolute value of the branch current and the voltage difference between the two ends of the branch, and calculating the equivalent resistance of the branch based on the absolute value of the branch current and the voltage difference between the two ends of the branch, specifically includes: Assume that the node set of all measurement points in the network is V = {v1,v2,...,v N }; where v i is the i-th node; N is the total number of nodes; Suppose the branch set containing all network connection edges is E={e1,e2,...,e M }; where e k is the kth branch; M is the total number of branches; Automatic collection branch e k The real-time current absolute value I k and branch e k Voltage difference ΔU k ; Calculate branch e k The equivalent resistance 3. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 2, characterized in that: The current threshold sequence is constructed based on the collected current data, and the current set is deduplicated and arranged in descending order to obtain multiple current thresholds, specifically including: Construct current set S = {I k |k=1,2,...,M}; Remove duplicates from the current set S and sort them in descending order, and get the threshold sequence τ1>τ2>…>τ L ; Among them, {τ j }=unique(S), which is the threshold sequence; L≤M, which is the total number of thresholds; unique(·) is the set deduplication function, which removes duplicate elements.

4. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 3, characterized in that: For each current threshold, constructing a corresponding threshold subgraph specifically includes: For each threshold τ j Construct a subgraph: G j =(V,E j ), E j ={e k ∈E|I k ≥τ j }; Among them, G j is the threshold τ j The subgraph below; E j To include all k ≥τ j branch road.

5. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 4, characterized in that: The step of constructing a spanning forest for each threshold subgraph and extracting a corresponding basis cycle set based on the spanning forest specifically includes: For each subgraph G j Construct the generated forest T as follows: j And extract the base cycle set C j : S110, E j For the middle branch, press I k Sort from large to small and record it as a sequence Among them, e i,h G j The branch with the hth largest current; I(e i,h ) is its corresponding current amplitude; S120, initialization generation forest S130, add branches one by one, for h=1 to |E j |Execute steps S131 to S132: S131, branch e i,h Connecting nodes (u i,h ,v i,h ); where u i,h ,v i,h Branch e i,h The starting point and end point; S132, if the edge e i,h Join T j After, u i,h With v j,h In T j is still not connected, then T j ←T j ∪{e i,h }; S140, let the base cycle edge set be B j =E j \T j ; S150, base loop structure: For each e∈B j : Among them, e is the connection node u e With v e edge; c j (e) is the forest T generated by branch e j A simple circuit consisting of a unique path connecting its endpoints; For the forest j Connecting node u e With v e The only path; S160, forming a base cycle set C j ={c j (e)|e∈B j }.

6. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 5, characterized in that: Deduplication and merging of the base cycle sets extracted from each threshold subgraph to obtain a unique base cycle set for the entire network specifically includes: Union all base loops For any If the edge sets are the same, that is, {e|e∈c}={e|e∈c'}, then only one copy is retained, and the final base cycle set is obtained 7. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 6, characterized in that: The calculation of the average energy loss based on each base cycle specifically includes: For each loop c∈C, perform the following steps: S210, calculate the number of loop edges |c|=#{e|e∈c}; where #{·} is a set cardinality operator used for counting; S220, calculate average energy loss Among them, I e is the current amplitude of branch e; R e is the equivalent resistance of branch e.

8. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 7, characterized in that: The persistence of each base cycle is measured, and normalized based on its frequency of occurrence at different thresholds to obtain a corresponding persistence ratio, specifically including: The threshold level of the statistical loop c n(c) = #{j|c∈C j }, 1≤n(c)≤L; Normalized persistence ratio 9. A method for calculating theoretical line loss of a power grid based on automatic data collection according to claim 8, characterized in that: The energy indicators of all base cycles are weighted and accumulated to output the cycle energy measurement value of the entire power grid as the estimation result of the theoretical line loss, which specifically includes: The weighted accumulation of all base cycles gives the overall cycle energy metric of the network: P cycle =∑ c∈C E(v)×r(c)。

Images