A method for tracing and correcting no solution of main grid and distribution network power flow model

Abstract

The application discloses a kind of main distribution network power flow model no solution traceability and correction method, it is related to power system dispatching technical field, including, acquisition real-time measurement data pre-processing, obtain quantum noise intensity coefficient;Quantum state density matrix is generated based on quantum noise intensity coefficient, and quantum fidelity is obtained by characteristic decomposition mapping;Based on quantum fidelity, the minimum eigenvalue is extracted, and the fault node coordinates are located by no solution traceability mapping, and the fault severity index is obtained by negative exponential mapping;Based on fault severity index, no solution traceability grade is divided, and fault scene is output;Based on fault scene, the set of power flow inversion control instructions is obtained, and the power flow no solution risk early warning value is corrected.The quantum fidelity mapping accurately locates the power grid fault source, eradicates the main distribution network model no solution problem, promotes the landing of quantum sensing industry and improves the fault response speed.

Description

Technical Field

[0001] This invention relates to the field of power system dispatching technology, and in particular to a method for tracing and correcting unsolvable power flow models in main and distribution networks. Background Technology

[0002] With the in-depth development of smart grids, power flow calculation in the main distribution network has formed a multi-technology integration system: traditional numerical analysis methods (such as the Newton-Raphson method) solve the power flow equations iteratively using the Jacobian matrix; topology correlation analysis technology (based on graph theory connectivity detection and tomographic scanning) achieves physical isolation of the grid fault domain; and the quantum computing optimization layer uses quantum annealing algorithms to handle high-dimensional non-convex constraint problems, improving the timeliness of large-scale distribution network calculations. In the field of physical parameter correction, sensitivity analysis (based on Jacobian matrix eigenvalue decomposition) and adaptive rule bases have gradually established a correlation model between equipment parameter offset and power flow convergence, forming the basic framework of "fault location-parameter correction".

[0003] However, current technologies suffer from two major structural defects: First, the fault tracing dimension is fragmented—topology analysis and physical correction are decoupled (e.g., the device association level is set only based on experience), making it impossible to quantify the dynamic relationship between the fault severity index and the topology search depth, leading to a high misjudgment rate of the impact range; Second, there is a quantum-classical co-inaccuracy—quantum fidelity is attenuated by power grid frequency noise interference, and existing pulse control methods (such as the OpenPulse framework) lack a feedback mechanism based on power flow inversion commands, making it impossible to achieve dynamic compensation of amplitude and phase parameters through gradient optimization. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method for tracing and correcting the unsolvable source of faults in the main distribution network power flow model to solve the problems of inaccurate quantization of fault impact range and decreased fidelity of quantum control.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for tracing and correcting unsolvable power flow models in main distribution networks, comprising: collecting and preprocessing real-time measurement data to obtain quantum noise intensity coefficients; generating a quantum density of states matrix based on the quantum noise intensity coefficients and obtaining quantum fidelity through eigenvalue decomposition mapping; extracting the minimum eigenvalue based on the quantum fidelity, locating the coordinates of fault nodes through unsolvable tracing mapping, and obtaining a fault severity index through negative exponential mapping; classifying unsolvable tracing levels based on the fault severity index and outputting fault scenarios; obtaining a power flow inversion control instruction set based on the fault scenarios and correcting unsolvable power flow risk warning values.

[0008] As a preferred embodiment of the unsolvable source tracing and correction method for the main distribution network power flow model described in this invention, the real-time measurement data includes active power, reactive power and voltage amplitude.

[0009] The preprocessing includes data cleaning, spatiotemporal consistency verification, and normalization.

[0010] As a preferred embodiment of the unsolvable source tracing and correction method for the main distribution network power flow model described in this invention, the acquisition of quantum noise intensity coefficient refers to calculating the quantum noise intensity coefficient based on the preprocessed real-time measurement data through a sliding window variance.

[0011] As a preferred embodiment of the unsolvable source tracing and correction method for the main distribution network power flow model described in this invention, the specific steps of generating a quantum state density matrix based on quantum noise intensity coefficients and obtaining quantum fidelity through eigenvalue decomposition mapping are as follows.

[0012] Noisy quantum gate sequences are generated based on quantum noise intensity coefficients, and quantum state density matrices are generated using quantum state tomography.

[0013] Construct a composite matrix based on the quantum state density matrix and the standard convergent state;

[0014] The eigenvalue spectrum is obtained by combining the quantum density of states matrix with the composite matrix, and the quantum fidelity is obtained by fidelity mapping.

[0015] As a preferred embodiment of the unsolvable source tracing and correction method for the main distribution network power flow model described in this invention, the steps of extracting the minimum feature value based on quantum fidelity, locating the coordinates of the fault node through unsolvable source tracing mapping, and obtaining the fault severity index through negative exponential mapping are as follows.

[0016] Minimum eigenvalues ​​are extracted based on quantum fidelity, and noise-sensitive spectral state vectors are obtained;

[0017] The coordinates of component components are located based on noise-sensitive spectral state vectors, and the coordinates of fault nodes are located through power grid topology mapping.

[0018] The severity index of a fault is obtained based on the coordinates of the faulty node and the minimum eigenvalue.

[0019] As a preferred embodiment of the unsolvable source tracing and correction method for the main distribution network power flow model described in this invention, the unsolvable source tracing level classification based on the fault severity index refers to classifying the fault level based on the fault severity index and the fault node coordinates through unsolvable source tracing level mapping, and generating a topology subgraph through dynamic expansion.

[0020] As a preferred embodiment of the unsolvable source tracing and correction method for the main distribution network power flow model described in this invention, the specific steps for outputting the fault scenario are as follows:

[0021] The faulty device ID is obtained by using a multi-source spatiotemporal graph convolutional network based on topological subgraphs;

[0022] Based on the fault severity index and fault level, the physical characteristics of the fault are obtained through the power grid physical characteristic rule base, and the power flow unsolvable risk warning value is output through the unsolvable risk criterion.

[0023] The fault scenario is output by encapsulating the fault device ID and fault physical characteristics through spectral field topology.

[0024] As a preferred embodiment of the unsolvable source tracing and correction method for the main distribution network power flow model described in this invention, the specific steps for obtaining the power flow inversion control instruction set based on the fault scenario and correcting the unsolvable power flow risk warning value are as follows:

[0025] A twin environment is constructed based on fault scenarios and real-time measurement data. A standardized state space is output based on the quantum state fusion vector, and a power flow inversion control instruction set is obtained.

[0026] Execute the power flow reversal control instruction set, collect quantum physical state feedback of the quantum state density matrix, and calculate the fidelity improvement rate;

[0027] The risk warning value for correcting the unsolvable trend based on the fidelity improvement rate.

[0028] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the unsolvable source tracing and correction method for the main distribution network power flow model as described in the first aspect of the present invention.

[0029] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the unsolvable source tracing and correction method for the main distribution network power flow model as described in the first aspect of the present invention.

[0030] The beneficial effects of this invention are as follows: it accurately locates the source of power grid faults through quantum fidelity mapping, and achieves exponential swallowing of unsolvable early warning values ​​based on twin-driven risk correction. Simultaneously, it uses the minimum eigenvalue to drive fault tracing and power flow convergence guarantee, ultimately solving the unsolvable problem of the main distribution network model, promoting the industrial application of quantum sensing and improving fault response speed. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 The flowchart shows the source tracing and correction method for unsolvable power flow models in main and distribution networks.

[0033] Figure 2 A flowchart for obtaining quantum fidelity.

[0034] Figure 3 A flowchart for calculating the fault severity index.

[0035] Figure 4 A flowchart for correcting risk warning values. Detailed Implementation

[0036] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0037] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0038] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0039] Reference Figures 1-4 This is one embodiment of the present invention, which provides a method for tracing and correcting unsolvable power flow models in main distribution networks, including the following steps:

[0040] S1. Collect and preprocess real-time measurement data to obtain the quantum noise intensity coefficient;

[0041] Real-time measurement data includes active power, reactive power, and voltage amplitude;

[0042] It should be noted that real-time active power measurement data is a quantitative parameter of the actual energy transmission rate in the power grid. Essentially, it describes the instantaneous power component of electrical energy converted into mechanical and thermal energy. Its core functions are real-time monitoring of the power grid's power generation and consumption balance, maintaining power system frequency stability, and driving dynamic simulation modeling of the noisy quantum gate sequence generation process. Real-time reactive power measurement data characterizes the energy exchange rate required to establish and maintain the electromagnetic field, reflecting the periodic energy exchange intensity between the electric and magnetic fields. Its core functions are quantifying the operating status of voltage support equipment, providing a voltage regulation control benchmark, and serving as the input source for calculating the variance of the quantum noise intensity coefficient. Real-time voltage amplitude measurement data... The instantaneous scalars that record the effective voltage values ​​of power grid nodes and the potential difference parameters that measure electric field strength serve as the core physical quantities for assessing the voltage stability margin of the power system, triggering overvoltage and undervoltage protection actions, and reconstructing the quantum density of states matrix using quantum state tomography. Real-time active power measurement data ensures dynamic balance between energy supply and demand, real-time reactive power measurement data maintains stable exchange of electromagnetic field energy, and real-time voltage amplitude measurement data monitors the safety boundary of electric field strength. Finally, the quantum noise intensity coefficient is calculated through a sliding window variance to provide data support for modeling dissipative noise injection, thus collaboratively driving the entire process of fault diagnosis, quantum state evolution, and power flow inversion control.

[0043] Preprocessing includes data cleaning, spatiotemporal consistency verification, and normalization.

[0044] It should be noted that data cleaning processes outliers in real-time measurement data. Outliers are identified in real-time active power, reactive power, and voltage amplitude measurement data. Data points deviating from the mean by more than three standard deviations are marked as outliers. Missing values ​​are filled using linear interpolation, replacing them with the mean of consecutive valid points to ensure data continuity. Spatiotemporal consistency verification ensures the physical reliability of the measurement data. First, it verifies the timestamp alignment by checking that the time series of all real-time measurement data points meet the requirement of equal-interval sampling to eliminate time jitter errors. Second, it verifies power balance constraints based on the power grid topology, implementing inflow and outflow difference verification for real-time active power measurement data at nodes. Reactive power measurement data, combined with voltage amplitude measurement data, undergoes capacitive reactance matching to eliminate spatially conflicting data. Normalization achieves data scale uniformity by using the max-min normalization method to linearly transform real-time measurement data of different dimensions to the [0,1] interval for reverse recovery calculations.

[0045] Based on the preprocessed real-time measurement data, the quantum noise intensity coefficient is calculated using the sliding window variance.

[0046] It should be noted that, firstly, a sliding window of fixed length of 20 points is used. For real-time active power measurement data, 20 consecutive real-time sampling points are extracted at each sliding window position, and the variance of this 20-point sequence is calculated. Simultaneously, based on real-time reactive power measurement data, a 20-point reactive power real-time measurement data sequence corresponding to the same time sliding window is extracted, and its variance is calculated. Next, for real-time voltage amplitude measurement data, a 20-point voltage amplitude real-time measurement data sequence covered by the current sliding window is extracted, and its variance is calculated. Subsequently, the ratio of the variance of the real-time active power measurement data in the sliding window to the square of the maximum standard deviation of the real-time active power measurement data recorded in the preprocessing stage is calculated; the ratio of the variance of the real-time reactive power measurement data in the current sliding window to the square of the maximum standard deviation of the real-time reactive power measurement data recorded in the preprocessing stage is also calculated; the ratio of the variance of the real-time voltage amplitude measurement data in the current sliding window to the square of the maximum standard deviation of the real-time voltage amplitude measurement data recorded in the preprocessing stage is also calculated. Finally, the quantum noise intensity coefficient is generated.

[0047] The quantum noise intensity coefficient includes the quantum noise intensity coefficient of generated active power, the quantum noise intensity coefficient of reactive power, and the quantum noise intensity coefficient of voltage amplitude.

[0048] It should be noted that the expression for the active power quantum noise intensity coefficient is:

[0049] ;

[0050] in, For time indexing Real-time active power measurement data The quantum noise intensity coefficient, It is real-time measurement data of active power. For time indexing; Real-time active power measurement data within the sliding window The variance value; Real-time active power measurement data The historical maximum standard deviation;

[0051] The expression for the reactive power quantum noise intensity coefficient is:

[0052] ;

[0053] in, For time indexing Real-time reactive power measurement data The quantum noise intensity coefficient, It is real-time reactive power measurement data. For time indexing; Real-time reactive power measurement data within the sliding window The variance value; Real-time reactive power measurement data The historical maximum standard deviation;

[0054] The expression for the voltage amplitude quantum noise intensity coefficient is:

[0055] ;

[0056] in, For time indexing Real-time measurement data of voltage amplitude The quantum noise intensity coefficient, It is real-time voltage amplitude measurement data. For time indexing; Real-time measurement data of voltage amplitude within the sliding window The variance value; Real-time measurement data of voltage amplitude The historical maximum standard deviation;

[0057] S2. Generate the quantum state density matrix based on the quantum noise intensity coefficient, and obtain the quantum fidelity through eigenvalue decomposition mapping; see the detailed steps of S2. Figure 2 ,include:

[0058] Based on the quantum noise intensity coefficient, a noisy quantum gate sequence is generated by dissipative noise injection, and a quantum state density matrix is ​​generated by quantum state tomography.

[0059] It should be noted that the quantum noise intensity coefficient is directly converted into dissipative noise channel parameters through an exponential function mapping; for each quantum logic gate operation moment in the quantum gate sequence, the quantum noise intensity mapping parameters are adjusted using the quantum noise intensity coefficient; and this dissipative noise generator is integrated into the ideal quantum logic gate operation through a noise injection operator method to achieve the superposition of dissipative noise effects; all quantum logic gate operations are processed in chronological order, and a noisy quantum gate sequence is generated for each quantum logic gate. Quantum state tomography initializes the quantum register to the zero state; the noisy quantum gate sequence is loaded to drive the quantum state evolution; the quantum state tomography uses a complete Paulige measurement combination to perform projection measurement operations on the evolved quantum state; the process of preparing the quantum state, loading the noisy quantum gate sequence, and performing projection measurement is repeated one thousand times, and the measurement results are collected and recorded statistical data; the quantum state tomography incorporates a maximum likelihood estimation algorithm to process the measurement data, establishes a negative log-likelihood function and adds a positive semi-definite constraint, optimizes and minimizes the objective function, and iteratively updates it through gradient descent to converge and generate the quantum state density matrix.

[0060] A composite matrix is ​​constructed based on the quantum state density matrix and the standard convergent state through bilinear projection transformation;

[0061] It should be noted that, firstly, the quantum density of states (DSO) matrix data output by quantum state tomography is read; then, the mathematical representation of the standard convergent state is loaded; the DSO matrix is ​​projected onto the subspace containing the standard convergent state; this projection mapping process is mathematically expressed as a bilinear operator acting on the DSO matrix and the standard convergent state; finally, the projected components of the DSO matrix and the standard convergent state are fused through tensor product operations to generate a bilinear projection transformation matrix; this resulting matrix is ​​the composite matrix. The dimension of the composite matrix is ​​the product of the dimension of the DSO matrix and the dimension of the standard convergent state, and the final output is a composite matrix describing the deviation characteristics of the DSO matrix relative to the standard convergent state.

[0062] Furthermore, the standard convergent state is the mathematical expression of the static quantum state target state, serving as a benchmark for evaluating the deviation of the quantum state density matrix. The standard convergent state, represented as a fixed-parameter pure-state or mixed-state density matrix, provides a subspace projection benchmark during the bilinear projection transformation, enabling the quantitative extraction of the quantum state density matrix components projected onto the subspace containing the standard convergent state. In the tensor direct product stage, it serves as the foundation for the composite matrix's features, preserving the deviation information of the quantum state density matrix relative to the standard convergent state through algebraic operations. Finally, the composite matrix fully encapsulates the magnitude and correlation characteristics of the quantum state density matrix's deviation from the standard convergent state, providing crucial input data for subsequent quantum noise evaluation steps.

[0063] Based on the quantum density of states matrix combined with the composite matrix, the eigenvalue spectrum is obtained through eigenvalue decomposition, and the quantum fidelity is obtained through fidelity mapping.

[0064] It should be noted that Hermitian matrix eigenvalue decomposition is performed on the composite matrix to solve the algebraic characteristic equation, obtain all characteristic solutions, and generate an eigenvalue spectrum by arranging the real eigenvalues ​​in descending order and storing the corresponding eigenvector groups simultaneously. The largest eigenvalue in the eigenvalue spectrum is extracted, and the largest eigenvalue in the eigenvalue spectrum is used to generate a quantum fidelity that characterizes the approximation of the quantum state density matrix to the ideal target state through fidelity mapping.

[0065] S3. Extract the minimum eigenvalue based on quantum fidelity, locate the fault node coordinates through unsolvable source tracing mapping, and obtain the fault severity index through negative exponential mapping; see the detailed steps of S3. Figure 3 ,include:

[0066] Based on quantum fidelity, the minimum eigenvalue is extracted using a feature spectrum descent algorithm, and the noise-sensitive spectral state vector is obtained through index mapping.

[0067] It should be noted that, based on quantum fidelity, an end-indexing operation is performed on the descending eigenvalue spectrum obtained by eigenvalue decomposition. The minimum eigenvalue is extracted as the key output parameter, and the original storage index of the minimum eigenvalue in the eigenvalue spectrum is located. The corresponding eigenvector is then directly extracted according to the index based on the eigenvector sequence stored in the eigenvalue decomposition stage. This fully realizes the accurate extraction of the end-feature parameters and corresponding quantum state vectors of the descending eigenvalue spectrum, ultimately outputting a noise-sensitive spectral state vector carrying the correlation parameter of the minimum eigenvalue.

[0068] Based on the noise-sensitive spectral state vector, the component coordinates are located by peak component indexing, and the coordinates of the fault node are located by power grid topology mapping.

[0069] It should be noted that the stage operation of locating component coordinates based on peak component indexing of the noise-sensitive spectral state vector involves traversing all element amplitude values ​​of the noise-sensitive spectral state vector, identifying the peak component corresponding to the largest amplitude component through the maximum value index, and calling the index-coordinate transformation function to input the peak component index and output the component coordinate value. The stage operation of component coordinate values ​​is achieved by locating the coordinates of the fault node through power grid topology mapping. First, the power grid topology mapping dictionary is obtained through the hash mapping table of power grid node coordinates indexed by the quantum ground state. Then, the ground state index of the quantum state amplitude component is used as the key to query the power grid topology mapping dictionary, and an exact key-value matching algorithm is executed to obtain the spatial coordinate parameters of the corresponding power grid node. Finally, the spatial coordinate parameters are output as the coordinates of the fault node.

[0070] The severity index of a fault is obtained by using a negative exponential function based on the coordinates of the faulty node and the minimum eigenvalue.

[0071] It should be noted that the fault node coordinate values ​​are obtained through spatial coordinate analysis based on the fault node coordinates, and the minimum eigenvalue parameter associated with the noise-sensitive spectral state vector is also obtained. A negative exponential function is constructed using the natural constant as the base, where the quantum noise conversion coefficient is dimensionless. The fault node coordinate values ​​are extracted, and Euclidean modulus calculations are performed to generate coordinate quantization weights, enabling the fault severity index to achieve a geometric weight amplification effect. Finally, the fault severity index, representing the quantitative assessment result of the fault degree of the target power grid node, is output. The expression for the fault severity index is:

[0072] ;

[0073] in, This represents the severity index of the fault. The minimum eigenvalue parameter associated with the noise-sensitive spectral state vector; For the fault node, the horizontal coordinate component in the Cartesian coordinate system is used. For the longitudinal coordinate components of the Cartesian coordinate system of the fault node; Based on the correlation analysis of historical power grid noise data and fault records, a preset quantum noise conversion coefficient (with a value range of 0 to 5) is fitted using least squares.

[0074] S4. Classify the level of unsolvable source tracing based on the fault severity index and output the fault scenario;

[0075] Based on the fault severity index and fault node coordinates, the fault level is divided by the unsolvable source tracing level mapping, and a topological subgraph is generated by dynamic expansion.

[0076] It should be noted that a pre-defined grading rule table is used to map the unsolvable source tracing level based on the fault severity index input. The rule defines the fault severity index value range [0,1] as a minor fault level, (1,3] as a moderate fault level, and (3,5] as a severe fault level. During the dynamic expansion and generation of the topology subgraph, the coordinate data of the fault node is loaded, and the connection information of the associated nodes is obtained by calling the power grid topology connection relationship database. A breadth-first search algorithm is executed to traverse three layers of adjacent nodes and lines with the fault node coordinates as the center. During the expansion process, the subgraph generation range is dynamically controlled by adjusting the weight coefficient of the search step size according to the fault level. Finally, the node set and line connection relationship constitute the topology subgraph.

[0077] Furthermore, the pre-defined hierarchical rule table for unsolvable source tracing mapping plays a fundamental role as a configuration dataset storage structure component that solidifies the mapping relationship between the fault severity index range and discrete fault levels. Through the fault severity index, it achieves discrete classification of fault severity, establishing a deterministic mapping relationship from quantitative indicators to semantic levels. During the dynamic expansion and generation of the topology subgraph, it dynamically adjusts the adjacency layer expansion depth parameter of the breadth-first search algorithm based on the mapped fault level values ​​to control the scale of topology subgraph generation, thereby ensuring precise matching between topology analysis resource allocation and fault impact, and realizing a configurable response mechanism for refined topology analysis based on fault levels.

[0078] Based on the topological subgraph, a multi-source spatiotemporal graph convolutional network is used to propagate spatiotemporal features and obtain the fault device ID.

[0079] It should be noted that, firstly, an undirected graph adjacency matrix is ​​constructed based on the connection relationships between nodes in the topological subgraph to represent the physical connection topology between nodes in the topological subgraph; a real-time measurement dataset of historical time series of topological subgraph nodes, including real-time active power measurement data, real-time reactive power measurement data, and real-time voltage amplitude measurement data, is loaded to construct a time-dimensional feature vector; the undirected graph adjacency matrix is ​​used as the spatial topology weight, and the time-series feature vector is superimposed and jointly optimized through spatiotemporal graph convolution to extract spatiotemporal features, obtaining a pre-trained multi-source spatiotemporal graph convolutional network; the adjacency matrix and time-series feature vector are input into the pre-trained multi-source spatiotemporal graph convolutional network; the multi-source spatiotemporal graph convolutional network aggregates the features of adjacent topological subgraph nodes through spatial graph convolutional layers to achieve spatial dimension information propagation, and extracts time-series pattern features through temporal convolutional layers to achieve time-dimensional information propagation; the network ends are connected to fully connected layers to output node-level fault probability prediction values; binarization discrimination is performed based on the fault probability prediction values ​​to output a set of topological subgraph node labels; the set of topological subgraph node IDs corresponding to the labeled topological subgraph nodes is extracted, and fault device IDs are generated through a global mapping table.

[0080] Based on the fault severity index and fault level, the physical characteristics of the fault are obtained through the power grid physical characteristic rule base, and the power flow unsolvable risk warning value is output through the unsolvable risk criterion.

[0081] It should be noted that, firstly, the power grid physical characteristic rule base receives the fault severity index value and fault level as input index keys, performs a pre-stored rule matching and retrieval operation, and returns the corresponding physical characteristic parameter set, including the theoretical value of short-circuit impedance (obtained according to the power system short-circuit capacity calculation procedure, with a value range of 0.05~20.0 pu), the fault current threshold (set according to the circuit breaker breaking capacity standard, with a value range of 10~50 kA), and the thermal stability limit time (calculated through the conductor thermal stability characteristic curve, with a value range of 1.0~3.0 seconds). Subsequently, the no-solution risk criterion processes the fault device ID and physical characteristic parameters, and establishes a logic tree for determining the no-solution condition of the power flow equation, including branch power exceeding the limit test, node voltage deviation test, and fault current exceeding the limit test. Based on the fault device ID list, the topology subgraph is traversed item by item to check the condition satisfaction and generate a Boolean judgment result vector. The final comprehensive judgment result vector and the fault severity index output a normalized power flow unsolvable risk warning value (the value range is a normalized real number interval of [0,1], and the output value is precisely limited to the distribution in the continuous interval from zero to one, and the larger the value, the higher the risk of unsolvable).

[0082] Furthermore, the power grid physical characteristic rule base is a predefined database structure functional component that stores quantitative parameters of fault characteristics of power system components and corresponding boundary condition judgment rules. Its core function is to: input the fault severity index value into the mapping index to preset the standard parameter range of physical characteristics, and at the same time encapsulate the power flow unsolvable judgment logic conditions, including branch power exceeding the limit condition, node voltage deviation condition, and fault current exceeding the standard condition, to provide a standardized physical characteristic input source and judgment rule benchmark for unsolvable risk judgment, ensuring that the fault analysis process has verifiable engineering physical basis.

[0083] Furthermore, fault physical characteristics are a set of quantitative parameters characterizing the inherent electrical response characteristics and safety boundaries of faulty equipment in a power system. When a short-circuit fault occurs in a power line, the equivalent total impedance value of the path through which the fault point current flows is quantified by the theoretical value of the short-circuit impedance, which constrains the change in the fault current amplitude. The equipment insulation withstand capability boundary is determined by the fault current threshold, and the thermal stability limit time is determined, which constrains the duration of the equipment's short-circuit thermal load. Through the comprehensive electromagnetic-thermal characteristics of the faulty equipment, a precise quantitative description of the equipment's physical response under fault conditions is achieved, supporting the determination of engineering physical rules for unsolvable risk criteria, and ensuring that the fault severity assessment results have strict power system engineering physical interpretability.

[0084] The fault scenario is output by encapsulating the fault device ID and fault physical characteristics through spectral field topology.

[0085] It should be noted that the process of performing spectral field topology encapsulation to output the fault scene based on the faulty device ID and faulty physical characteristics begins by inputting the device identifiers contained in the faulty device ID list and calling the physical characteristic parameter set to establish a key-value pair dictionary structure for device physical characteristics; then, based on the node connection relationship of the topology subgraph, the adjacency matrix is ​​extracted and the faulty device ID is injected into the topology graph as the core node, while the physical characteristic parameters are bound to the corresponding node attribute fields; the spectral field topology encapsulation operation constructs a directed graph structure in which the vertex set contains the faulty device ID and the three-layer adjacent device IDs, the edge set contains the line connection relationship of the topology subgraph, and the attribute matrix stores the node bound physical characteristic parameter values ​​to generate standard graph structure description language format data; finally, the fault scene is encapsulated and output.

[0086] S5. Obtain the power flow inversion control instruction set based on the fault scenario, and correct the power flow unsolvable risk warning value. See below for the specific steps of S5. Figure 4 ,include:

[0087] A twin environment is constructed based on fault scenarios and real-time measurement data. A standardized state space is output based on the quantum state fusion vector, and a power flow inversion control instruction set is obtained.

[0088] It should be noted that, based on the fault scenario and combined with real-time measurement data, a twin environment is constructed. First, the fault scenario is loaded to extract the physical characteristic parameters of the faulty equipment, which are then injected into the digital twin platform to establish an electrical parameterized model. This model is then synchronously loaded with the latest real-time measurement data package, containing real-time active power, reactive power, and voltage amplitude measurement data, driving the dynamic simulation of the twin environment. A standardized state space transformation is performed using quantum state fusion vector input: Quantum fidelity and noise-sensitive spectral state vectors are used to construct the quantum state fusion vector input space. Principal component analysis is used to extract the first four eigenvectors with the largest variance, forming a four-dimensional standardized state space coordinate base. Under the standardized state space coordinates, a controlled differential equation is established, and the physical characteristic parameters of the faulty equipment are set as boundary constraints to solve the objective function. This allows for the reconstruction of the current state space trajectory and the output of the power flow inversion control instruction set before the fault.

[0089] The power flow inversion control command set includes transformer tap adjustment, line parameter compensation, and reactive power switching commands;

[0090] It should be noted that the transformer tap change command is an operational command that adjusts the transformer winding turns ratio to precisely control the node voltage. It has the ability to quickly respond to voltage deviations and plays a role in stabilizing the voltage of key nodes in fault scenarios. The line parameter compensation command is a dispatch command that configures series or parallel impedance compensation devices on the transmission line and modifies the equivalent resistance or reactance parameters of the line. It corrects the abnormal power flow distribution caused by the fault in real time by optimizing the equivalent impedance parameters, and solves the problem of active-reactive power flow imbalance. The reactive power switching command is a switching operation command that controls the connection / disconnection status of parallel capacitor banks or reactor banks to provide or absorb local reactive power. It plays a core supporting role in achieving local reactive power balance in the fault area and maintaining the grid voltage level within a reasonable range.

[0091] Execute the power flow reversal control instruction set, collect quantum physical state feedback of the quantum state density matrix, and calculate the fidelity improvement rate;

[0092] It should be noted that the physical actuator receives the power flow inversion control command set and completes the field equipment operation; after the operation completion delay window, it re-acquires real-time power grid measurement data; based on the new real-time measurement data, it regenerates a noisy quantum gate sequence through dissipative noise injection and reconstructs a new quantum state density matrix using quantum state tomography; based on the quantum state density matrix, it reads the new quantum fidelity, and simultaneously performs eigenvalue decomposition based on the new quantum state density matrix to obtain the composite matrix and perform fidelity mapping calculation to obtain the current fidelity improvement rate. The expression for the fidelity improvement rate is:

[0093] ;

[0094] in, It is the fidelity improvement rate; For new quantum fidelity; Quantum fidelity before executing the power flow reversal control instruction set; The maximum standard deviation of the real-time voltage amplitude measurement data recorded during the preprocessing stage; The minimum value in the eigenvalue spectrum reflects the maximum perturbation intensity of quantum noise. The change in Euclidean distance of the noise-sensitive spectral state vector;

[0095] Based on the fidelity improvement rate, the quantum power frequency coupling algorithm is used to correct the risk warning value of unsolvable power flow.

[0096] It should be noted that the quantum power frequency coupling algorithm is based on the fidelity improvement rate and the current power flow unsolvable risk warning value. First, the fundamental frequency of the real-time active power measurement data is extracted and the frequency traction factor is calculated by combining the mean of the off-diagonal modulus values ​​of the quantum density of states matrix. The warning value is corrected based on the frequency traction factor and the output of the fidelity adjustment function. If the detected fundamental frequency deviation of the power grid exceeds 0.2 Hz, the frequency traction factor is locked at 0.99. When the fidelity improvement rate is greater than 30%, the gain compensation mechanism is activated to adjust the correction value a second time. Finally, the power flow unsolvable risk warning value in the range [0,1] is output.

[0097] This embodiment also provides a computer device applicable to the case of unsolvable source tracing and correction methods for main and distribution network power flow models, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the unsolvable source tracing and correction methods for main and distribution network power flow models proposed in the above embodiment.

[0098] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0099] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the solution-free tracing and correction method for the main distribution network power flow model as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0100] In summary, this invention achieves precise location of power grid fault sources through quantum fidelity mapping, exponential absorption of unsolvable early warning values ​​based on twin-driven risk correction, and simultaneously utilizes minimum eigenvalues ​​to drive fault tracing and power flow convergence assurance, ultimately solving the unsolvable problem of main and distribution network models, promoting the industrial application of quantum sensing and improving fault response speed.

[0101] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for tracing and correcting unsolvable power flow models in main distribution networks, characterized in that: include, Real-time measurement data is collected and preprocessed to obtain the quantum noise intensity coefficient; The quantum state density matrix is ​​generated based on the quantum noise intensity coefficient, and the quantum fidelity is obtained through eigenvalue decomposition mapping. The minimum feature value is extracted based on quantum fidelity, the coordinates of the fault node are located by unsolvable source tracing mapping, and the fault severity index is obtained by negative exponential mapping. The system classifies unsolvable source tracing levels based on the fault severity index and outputs the fault scenarios. The power flow inversion control instruction set is obtained based on the fault scenario, and the power flow unsolvable risk warning value is corrected.

2. The method for tracing and correcting unsolvable problems in the main distribution network power flow model as described in claim 1, characterized in that: The real-time measurement data includes active power, reactive power, and voltage amplitude. The preprocessing includes data cleaning, spatiotemporal consistency verification, and normalization.

3. The method for tracing and correcting unsolvable problems in the main distribution network power flow model as described in claim 2, characterized in that: The acquisition of quantum noise intensity coefficient refers to the calculation of quantum noise intensity coefficient based on preprocessed real-time measurement data through a sliding window variance.

4. The method for tracing and correcting unsolvable problems in the main distribution network power flow model as described in claim 3, characterized in that: The specific steps for generating the quantum state density matrix based on the quantum noise intensity coefficient and obtaining the quantum fidelity through eigenvalue decomposition mapping are as follows. Noisy quantum gate sequences are generated based on quantum noise intensity coefficients, and quantum state density matrices are generated using quantum state tomography. Construct a composite matrix based on the quantum state density matrix and the standard convergent state; The eigenvalue spectrum is obtained by combining the quantum density of states matrix with the composite matrix, and the quantum fidelity is obtained by fidelity mapping.

5. The method for tracing and correcting unsolvable problems in the main distribution network power flow model as described in claim 4, characterized in that: The specific steps are as follows: extracting the minimum eigenvalue based on quantum fidelity, locating the coordinates of the fault node through a no-solution source tracing mapping, and obtaining the fault severity index through a negative exponential mapping. Minimum eigenvalues ​​are extracted based on quantum fidelity, and noise-sensitive spectral state vectors are obtained; The coordinates of component components are located based on noise-sensitive spectral state vectors, and the coordinates of fault nodes are located through power grid topology mapping. The severity index of a fault is obtained based on the coordinates of the faulty node and the minimum eigenvalue.

6. The method for tracing and correcting unsolvable problems in the main distribution network power flow model as described in claim 5, characterized in that: The classification of unsolvable tracing levels based on the fault severity index refers to classifying fault levels based on the fault severity index and fault node coordinates through unsolvable tracing level mapping, and generating a topological subgraph through dynamic expansion.

7. The method for tracing and correcting unsolvable problems in the main distribution network power flow model as described in claim 6, characterized in that: The specific steps for the output fault scenario are as follows. The faulty device ID is obtained by using a multi-source spatiotemporal graph convolutional network based on topological subgraphs; Based on the fault severity index and fault level, the physical characteristics of the fault are obtained through the power grid physical characteristic rule base, and the power flow unsolvable risk warning value is output through the unsolvable risk criterion. The fault scenario is output by encapsulating the fault device ID and fault physical characteristics through spectral field topology.

8. The method for tracing and correcting unsolvable problems in the main distribution network power flow model as described in claim 7, characterized in that: The specific steps for obtaining the power flow inversion control instruction set based on the fault scenario and correcting the power flow unsolvable risk warning value are as follows. A twin environment is constructed based on fault scenarios and real-time measurement data. A standardized state space is output based on the quantum state fusion vector, and a power flow inversion control instruction set is obtained. Execute the power flow reversal control instruction set, collect quantum physical state feedback of the quantum state density matrix, and calculate the fidelity improvement rate; The risk warning value for correcting the unsolvable trend based on the fidelity improvement rate.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the unsolvable source tracing and correction method for the main distribution network power flow model as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the unsolvable source tracing and correction method for the main distribution network power flow model as described in any one of claims 1 to 8.

Images