Novel power grid adjustable resource demand response method based on wide-area measurement data

Abstract

The invention discloses a novel power grid adjustable resource demand response method based on wide-area measurement data, and relates to the technical field of power regulation and control, and the method comprises the steps: building a gauge tensor through metric mapping based on a slice data set, and generating cluster topological data through a Riemann clustering algorithm according to a geodesic distance; constructing Hamiltonian by adopting quantum mapping based on cluster topological data, and generating a resource scheduling strategy through a quantum annealing algorithm; based on a resource scheduling strategy, multi-modal response verification is carried out on the multi-source measurement data through topology dynamic verification, and an instruction response verification report is generated; and according to the instruction response verification report, counting the proportion of the number of abnormal nodes in the cluster to generate a topology recombination decision, and optimizing a resource scheduling strategy through consistency verification. The intelligent degree gauge tensor is constructed by fusing the physical characteristics of the power grid and the communication topology, deep coupling cooperation of response clusters is achieved, and the bottleneck of large-scale real-time decision making is broken through.

Description

Technical Field

[0001] This invention relates to the field of power regulation technology, and in particular to a novel power grid adjustable resource demand response method based on wide-area measurement data. Background Technology

[0002] With the deepening of smart grid construction, wide-area measurement and response methods, based on synchronous phasor measurement units and high-speed communication networks, have achieved millisecond-level dynamic perception of grid status, providing a data foundation for demand response technologies. Traditional demand response mainly relies on load aggregators or price incentive signals to drive adjustable resources (such as distributed energy storage and flexible loads) for power regulation. In recent years, intelligent methods such as virtual power plant cluster control and resource scheduling strategies based on deep reinforcement learning have emerged. The application of software-defined networking and 5G slicing technology has also improved the real-time performance of resource response.

[0003] Although wide-area measurement response methods and intelligent scheduling algorithms have improved the accuracy of demand response, current technologies still face two major bottlenecks: First, the uncontrollability of physical topology means that traditional response mechanisms rely on static network models, which cannot adapt to the dynamic topology reconfiguration caused by the grid connection of new energy sources, leading to electrical coupling mismatch. Second, asynchronous verification is fragmented. The problems manifest as a lack of execution closed loop in topology reconfiguration decisions, measurement feedback relying on low-speed feedback that cannot capture sub-second power flow changes, and the lack of a collaborative verification mechanism among circuit breaker operation, communication reconfiguration, and synchronous phasor measurement unit measurements severely restricts response efficiency and safety. Summary of the Invention

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

[0005] Therefore, this invention provides a novel power grid adjustable resource demand response method based on wide-area measurement data to solve the problems of lack of dynamic adaptability of physical topology and asynchronous response verification.

[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 novel power grid adjustable resource demand response method based on wide-area measurement data, comprising: collecting multi-source measurement data and obtaining slice datasets through preprocessing; constructing a metric tensor based on the slice datasets through metric mapping, and generating cluster topology data based on geodesic distance using a Riemann clustering algorithm; constructing a Hamiltonian based on the cluster topology data using quantum mapping, and generating a resource scheduling strategy through a quantum annealing algorithm; verifying the multi-modal response of the multi-source measurement data through topology dynamic verification based on the resource scheduling strategy, and generating a command response verification report; generating a topology reorganization decision based on the command response verification report by statistically analyzing the proportion of abnormal nodes in the cluster, and optimizing the resource scheduling strategy through consistency verification.

[0008] As a preferred embodiment of the novel power grid adjustable resource demand response method based on wide-area measurement data described in this invention, the multi-source measurement data includes acoustic signature signals, voltage amplitude, current amplitude, electrical phase angle difference, frequency dynamic change rate, time-stamped synchronization signal, and data quality code.

[0009] The preprocessing includes abnormal data cleaning, time-scale synchronization correction, and feature enhancement filtering.

[0010] Based on the preprocessed multi-source measurement data, a slice dataset is obtained through voltage stability margin.

[0011] As a preferred embodiment of the novel power grid adjustable resource demand response method based on wide-area measurement data described in this invention, the specific steps for constructing a metric tensor based on the sliced ​​dataset through metric mapping are as follows:

[0012] Based on the 5G slice type in the slice dataset, communication parameters are obtained through multimodal parameter synchronous sensing, and feature transformation is performed through Cartesian tensor fusion decomposition to generate coupling strength factors.

[0013] The coupling coefficient is obtained through nonlinear field coupling based on the coupling strength factor and the electrical phase angle difference, and the metric tensor is constructed using the exponential positive definite symmetric method.

[0014] As a preferred embodiment of the novel power grid adjustable resource demand response method based on wide-area measurement data described in this invention, the specific steps for generating cluster topology data using the Riemann clustering algorithm based on geodesic distance are as follows:

[0015] The inertial intensity sequence is obtained by numerical mapping of the metric tensor using the Ricci scalar method, and the geodesic distance is generated using the Dijkstra algorithm.

[0016] Primary clusters are obtained using the spatial proximity method based on geodesic distance and inertial intensity sequences, and cluster topology data is generated through structured encapsulation.

[0017] As a preferred embodiment of the novel power grid adjustable resource demand response method based on wide-area measurement data described in this invention, the method involves constructing a Hamiltonian using quantum mapping based on cluster topology data. The specific steps are as follows:

[0018] The qubit layout is generated by index mapping based on cluster topology data, and the boundary constraint encoding is obtained by Pauli operators.

[0019] Based on the qubit layout and boundary constraint encoding, the Hamiltonian is constructed by linear superposition through adjacency matrix filtering.

[0020] As a preferred embodiment of the novel power grid adjustable resource demand response method based on wide-area measurement data described in this invention, the specific steps for generating the resource scheduling strategy using the quantum annealing algorithm are as follows:

[0021] Based on Hamiltonian, quantum state wave functions are generated by uniformly superimposing qubits through transverse field pulses, and resource scheduling strategies are generated by probability mapping in combination with cluster topology data.

[0022] As a preferred embodiment of the novel power grid adjustable resource demand response method based on wide-area measurement data described in this invention, the steps of verifying multi-modal responses to multi-source measurement data through topology dynamic verification based on resource scheduling strategies and generating a command response verification report are as follows.

[0023] Based on resource scheduling strategies and multi-source measurement data, abnormal nodes and communication latency are identified through spatiotemporal alignment cross-validation.

[0024] By detecting sudden changes in sound intensity, dual-modal analysis is performed to extract the voiceprint features from the device's voiceprint signal, and the active periods of the voiceprint are obtained through spatiotemporal binding.

[0025] Based on abnormal nodes, communication latency, and active periods of voiceprints, a command response verification report is generated through structured encapsulation.

[0026] As a preferred embodiment of the novel power grid adjustable resource demand response method based on wide-area measurement data described in this invention, the steps of generating a topology reorganization decision by statistically analyzing the proportion of abnormal nodes within the cluster based on the command response verification report, and optimizing the resource scheduling strategy through consistency verification, are as follows:

[0027] The number of abnormal nodes is counted based on the command response verification report, and an abnormal node distribution mapping table is generated through cluster affiliation mapping.

[0028] Based on the abnormal node distribution mapping table, the health index is obtained by the proportion of abnormal nodes, and a cluster status table is generated.

[0029] Based on the cluster status table, low health indices are filtered using discrete confidence levels, and topology reorganization decisions are generated.

[0030] The topology reconfiguration decision adopts a feedback mechanism to execute the opening and closing of physical circuit breakers, and optimizes the resource scheduling strategy by reallocating 5G slice resources.

[0031] 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 novel power grid adjustable resource demand response method based on wide-area measurement data as described in the first aspect of the present invention.

[0032] 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 novel power grid adjustable resource demand response method based on wide-area measurement data as described in the first aspect of the present invention.

[0033] The beneficial effects of this invention are as follows: by integrating the physical characteristics of the power grid and the communication topology to construct an intelligent metric tensor, deep coupling and collaboration of the response cluster are achieved. Furthermore, by using quantum constraint coding to transform complex resource scheduling into parallel optimization, the bottleneck of large-scale real-time decision-making is broken. Combined with the acoustic-electric dual-modal verification mechanism, abnormal nodes are accurately intercepted. Ultimately, the core obstacles of traditional methods, such as topology fragmentation, optimization lag, and lack of security verification, are optimized. Attached Figure Description

[0034] 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.

[0035] Figure 1 This is a flowchart of a novel power grid adjustable resource demand response method based on wide-area measurement data.

[0036] Figure 2 This is a flowchart of the demand response method.

[0037] Figure 3 A flowchart for generating the cluster topology.

[0038] Figure 4 A flowchart for topology reconfiguration. Detailed Implementation

[0039] 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.

[0040] 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.

[0041] 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.

[0042] Reference Figures 1-4 This is one embodiment of the present invention, which provides a novel power grid adjustable resource demand response method based on wide-area measurement data, comprising the following steps:

[0043] S1. Collect multi-source measurement data and obtain slice datasets through preprocessing;

[0044] Multi-source measurement data includes acoustic signature signal, voltage amplitude, current amplitude, electrical phase angle difference, frequency dynamic rate of change, time-scale synchronization signal, and data quality code;

[0045] It should be noted that the acoustic signature represents the acoustic characteristics of power equipment during operation, used to detect mechanical vibrations and abnormal discharges; the voltage amplitude indicates the potential difference amplitude at power grid nodes, used to measure voltage stability; the current amplitude reflects the current carrying capacity of line conductors, used to monitor overload risk; the electrical phase angle difference quantifies the AC phase offset angle between different nodes, used to analyze the direction of power transmission; the frequency dynamic change rate describes the rate of change of the power grid cycle, used to assess the rotational inertia reserve; the time-stamped synchronization signal provides a unified time stamp accurate to the microsecond level, used to achieve cross-regional measurement data alignment; and the data quality code marks the validity level of the collected data, used to filter abnormal data. The acoustic signature identifies hidden equipment faults; the voltage amplitude, current amplitude, and electrical phase angle difference form a real-time monitoring system for electrical quantities; the frequency dynamic change rate senses stability margin; the time-stamped synchronization signal ensures the consistency of data timing across the entire domain; and the data quality code establishes a reliable data filtering mechanism.

[0046] Preprocessing includes outlier cleaning, timescale synchronization correction, and feature enhancement filtering;

[0047] It should be noted that outlier cleaning filters outliers based on data quality codes. For example, if the data quality code indicates that cleaning is triggered at three times, interpolation of the nearest valid value is used as a replacement. Time-scale synchronization correction aligns the time scales of all measurement data based on the time-scale synchronization signal, for example, using GPS clock pulses to calibrate the 100-microsecond deviation interval. Feature enhancement filtering improves the signal-to-noise ratio of the acoustic signature signal through Butterworth second-order bandpass filtering, for example, enhancing the frequency range of 50 Hz to 4 kHz by 15 dB. These three operations form a sequential processing chain: outlier cleaning ensures data validity, time-scale synchronization correction achieves precise clock alignment across data sources, and feature enhancement filtering focuses on extracting key acoustic and electrical feature information.

[0048] Based on the preprocessed multi-source measurement data, a slice dataset is obtained through voltage stability margin;

[0049] It should be noted that, firstly, the eigenvalues ​​of the Jacobian matrix are constructed using voltage amplitude, current amplitude, and electrical phase angle difference; for example, the smallest eigenvalue is selected as the stability margin benchmark. Finally, the measurement data is segmented into millisecond-level time windows based on the time-stamped synchronization signal, for example, forming a slice data packet containing the current acoustic signature signal and electrical quantities every 200 milliseconds. The voltage stability margin drives the generation of time-correlated slice datasets by continuously quantifying the grid's disturbance rejection capability.

[0050] The slice dataset includes electrical dynamic measurements, 5G slice types, and control link characteristics.

[0051] It should be noted that electrical dynamic measurements include millisecond-level sequences of voltage amplitude, current amplitude, electrical phase angle difference, and frequency dynamic change rate, used to capture real-time fluctuations in the electrical state of the power grid; 5G slice type identifiers identify URLLC (Ultra-Reliable Low-Latency Communication) and eMBB (Enhanced Mobile Broadband) network slice modes, used to adapt to the allocation of communication resources for different control services; control link characteristics record the hop count, latency, and packet loss rate parameters of the command transmission path, used to quantify the performance of the response channel. This set of parameters forms a three-in-one collaborative mechanism: electrical dynamic measurements establish a precise profile of the physical state of the power grid, 5G slice type dynamically configures communication resources to ensure transmission determinism, and control link characteristics evaluate command reachability, jointly supporting the data fusion foundation for millisecond-level response decisions.

[0052] S2. Based on the slice dataset, construct the metric tensor through metric mapping, and generate cluster topology data through Riemann clustering algorithm according to geodesic distance;

[0053] Based on the 5G slice type in the slice dataset, communication parameters are obtained through multimodal parameter synchronous sensing, and feature transformation is performed through Cartesian tensor fusion decomposition to generate coupling strength factors.

[0054] It should be noted that multimodal parameter synchronous sensing parses communication parameters from 5G slice types, including scheduling period, air interface latency, and link jitter rate. For example, the scheduling period for URLLC slices is in the hundreds of microseconds range, and the scheduling period for eMBB slices is in the milliseconds range. Through mathematical transformations, the communication parameters are mapped to a unified tensor space, generating scheduling period tensors, air interface latency tensors, and link jitter rate tensors. The scheduling period tensor uses discrete unit vector encoding, the air interface latency tensor is mapped to an exponential decay function, and the link jitter rate tensor is transformed into a probability density distribution. Then, a cross-product fusion operation is performed, for example, the scheduling period tensor and the air interface latency tensor form a sixth-order tensor. Finally, eigenvalue decomposition is performed on the fused tensor, and the largest eigenvector component is selected as the coupling strength factor.

[0055] The coupling coefficient is obtained through nonlinear field coupling based on the coupling strength factor and the electrical phase angle difference, and the metric tensor is constructed using the exponential positive definite symmetric method.

[0056] It should be noted that, firstly, the coupling strength factor and the electrical phase angle difference are used to generate the coupling coefficient through a nonlinear function; then, a symmetric matrix with the number of nodes of order is created, and the off-diagonal elements are assigned values ​​according to the node connection relationship, while the node elements without direct electrical connection are set to zero; then, a positive definiteness check of the matrix is ​​performed, and when negative eigenvalues ​​appear, all negative eigenvalues ​​are forcibly raised to 0.01 to ensure positive definiteness; finally, the metric tensor matrix that satisfies the positive definite symmetry condition is output.

[0057] The inertial intensity sequence is obtained by numerical mapping of the metric tensor using the Ricci scalar method, and the geodesic distance is generated using the Dijkstra algorithm.

[0058] Numerical mapping of the metric tensor using the Ricci scalar method is employed to obtain the inertia intensity sequence and nodes. inertial strength Equal to its Richter scalar Based on metric tensor ( , The simplified engineering derivation formula for (node ​​index) is: , in, For nodes Local metric tensor The elements of the inverse matrix;

[0059] For nodes The notation of the decimation form (calculated from the partial derivative of the metric tensor): ), For nodes Local coordinate partial derivatives; For nodes The inertial strength constitutes the inertial strength sequence. ( (Total number of grid nodes).

[0060] Dijkstra's algorithm uses the elements of the metric tensor as weights in the adjacency matrix to generate the minimum cumulative weight path between nodes. With nodes geodetic distance ;

[0061] The formula is: ;

[0062] The engineering implementation clock, simplified using discrete path cumulative weights, is as follows: ,

[0063] in, For nodes arrive The set of all consecutive paths;

[0064] For nodes arrive The set of all discrete edges (based on power grid topology);

[0065] For metric tensor clock nodes and Continuous weights (non-interlocking elements);

[0066] For nodes arrive The geodesic distance (minimum cumulative weight).

[0067] It should be noted that the Ricci scalar method first obtains the Krzy symbol from the metric tensor, then constructs the Ricci tensor using the Krzy symbol, and finally shrinks it into a Ricci scalar. The Ricci scalar value output by each node is the inertial intensity sequence element. Subsequently, the elements of the metric tensor matrix are used as the adjacency matrix weights. Starting from the node with the largest inertial intensity sequence value, Dijkstra's algorithm is applied to traverse all node paths. The minimum cumulative weight of the node path is the geodesic distance.

[0068] Primary clusters are obtained using the spatial proximity method based on geodesic distance and inertial intensity sequences, and cluster topology data is generated through structured encapsulation.

[0069] It should be noted that, firstly, the node with the largest inertial intensity sequence value is used as the starting core node, and primary cluster members are formed by searching the geodesic distance and the distance to the core node. When there are multiple core nodes, this process is repeated in descending order of inertial intensity sequence value. If a node belongs to multiple primary clusters at the same time, it is assigned to the cluster of the core node with the smallest geodesic distance. Then, a structured encapsulation operation is performed to transform each primary cluster into an independent cluster unit. The independent cluster unit contains a list of member nodes and a core node identifier. Finally, all cluster units are merged to form complete cluster topology data.

[0070] Furthermore, primary clusters are intermediate node grouping structures generated during the spatial proximity method, driven by geodesic distance and inertial intensity sequences. First, the node with the largest inertial intensity sequence value is selected as the core node; primary cluster members are formed by searching the distances between the geodesic distance and the core node; when multiple core nodes exist, this operation is repeated in descending order of inertial intensity sequence values; for boundary nodes belonging to multiple core nodes, they are assigned to the group containing the core node with the smallest geodesic distance value. Essentially, primary clusters are preliminary topological grouping units in the node clustering process, containing core node identifiers and a list of member nodes, which are subsequently structured and encapsulated into sub-units of cluster topology data.

[0071] Cluster topology data includes generator cost parameters, load baseline power, and tie-line limits.

[0072] It should be noted that generator cost parameters characterize the unit power production cost coefficient of generating units within the cluster, used to quantify dispatch economics; load baseline power describes the typical power level of aggregated loads in the cluster, used to calculate adjustable resource capacity; tie-line limits define the maximum power exchange value at the cluster boundary, used to constrain the safety boundary of power interaction between regions. Generator cost parameters drive economic dispatch decisions, load baseline power anchors the demand response potential assessment benchmark, and tie-line limits construct power exchange safety barriers, collectively constituting the cluster topology data.

[0073] S3. Based on cluster topology data, a Hamiltonian is constructed using quantum mapping, and a resource scheduling strategy is generated using the quantum annealing algorithm.

[0074] The qubit layout is generated by index mapping based on cluster topology data, and the boundary constraint encoding is obtained by Pauli operators.

[0075] It should be noted that, firstly, the unique identifiers of all generator nodes, load nodes, and tie-line nodes in the cluster topology data are extracted and arranged in ascending order by node ID, and then qubit indices are continuously assigned. For example, generator node G01 is mapped to qubit q1, and load node L02 is mapped to qubit q2. Then, boundary constraint codes are generated for tie-line limit values. The power constraint of each tie-line is modeled as a Pauli operator combination term, where the tie-line limit value is directly converted into Pauli operator coefficients. At the same time, the generator cost parameters are mapped to linear Pauli Z-term coefficients to form boundary constraint codes.

[0076] Furthermore, boundary constraint encoding refers to the constraint rule encapsulation scheme that transforms the limit values ​​of tie lines and generator cost parameters in the cluster topology data into the mathematical expression form of Pauli operators. Specifically, the power constraint of each tie line is constructed as a combination of quadratic Pauli operators, and the generator cost parameter is constructed as a linear Pauli Z term, ultimately forming a set of mathematical expressions of constraint terms in the quantum Hamiltonian.

[0077] Furthermore, qubit layout refers to assigning a continuous integer index to each node by arranging the unique identifiers of generator nodes, load nodes, and tie-line nodes in the cluster topology data in ascending order according to the node ID. This establishes a one-to-one correspondence between physical nodes of the power grid and qubits, forming the physical basis encoding structure of scheduling variables in the quantum computing space.

[0078] Based on the qubit layout and boundary constraint encoding, the Hamiltonian is constructed by linear superposition through adjacency matrix filtering;

[0079] It should be noted that, firstly, based on the qubit layout index, the electrical connection relationships of nodes in the cluster topology data are extracted to generate an adjacency matrix. Elements of this adjacency matrix are 1 if there is a direct electrical connection between core nodes and 0 otherwise. Then, constraint terms corresponding to the non-zero elements in the adjacency matrix are selected from the boundary constraint encoding. Finally, a coefficient-weighted summation is performed on all selected constraint terms and all linear Pauli Z terms to generate the complete Hamiltonian. The expression for the complete Hamiltonian is: ,

[0080] in, This represents the set of qubit indices corresponding to all generator nodes in the qubit layout; express A single generator node identifier; Linear coefficients for mapping generator cost parameters (value range: [-2.0, 2.0]); For generator nodes Pauli Z operator for mapping; The set of node pairs generated by adjacency matrix filtering includes all combinations of nodes that have direct electrical connections. for Identifiers of nodes directly electrically connected in the middle. for Identifier of nodes with direct electrical connections; Nodes in cluster topology data With nodes Inter-line power transmission limit; The penalty coefficient for the quadratic term generated by the boundary constraint encoding (the value range is: positive real number greater than zero); node Pauli Z operator for mapping; node Pauli Z operator for mapping; For the complete Hamiltonian.

[0081] Based on Hamiltonian, quantum state wave function is generated by uniform superposition of qubits through transverse field pulses, and resource scheduling strategy is generated by probability mapping in combination with cluster topology data.

[0082] quantum state wave function Following the Schrödinger equation: ,

[0083] Initial state (after transverse field pulse):

[0084] in, To reduce Planck's constant (taken in engineering applications) (Normalization)

[0085] Let be the spin-up (0 state) and spin-down (1 state) basis vectors of the quantum bit q;

[0086] The ground state wave function at the end of the evolution (close to the ground state of the problem Hamiltonian H).

[0087] It should be noted that, firstly, a transverse field pulse is applied to put all qubits into a superposition state, which is mathematically represented by the quantum state wave function; then, through adiabatic evolution, the Hamiltonian is gradually changed from transverse field dominance to problem Hamiltonian dominance, and the ground state probability distribution is obtained by measuring the quantum state wave function after the evolution. After physical observation of each qubit, a binary numerical measurement result is obtained (where 0 represents spin up and 1 represents spin down), and the corresponding node state is determined according to the spin direction of the qubit in the binary numerical measurement result (this value directly corresponds to the on (1) or off (0) state of the power grid node). Combined with the rated power parameters of the generator nodes in the cluster topology data, the quantum state is converted into generator start-stop commands through probability mapping rules, and finally the resource scheduling strategy is output.

[0088] Quantum bit measurement probability and power dispatch formula:

[0089] Ground state measurement probability ( It is a sequence of spin states of qubits. ):

[0090]

[0091] Resource scheduling measurement (generator probability allocation): ,

[0092] in, A sequence of spin states The left arrow;

[0093] The rated power of generator g in the cluster topology data;

[0094] The spin state of the quantum bit corresponding to generator g ( Power on, (Shutdown)

[0095] The set of all possible spin state sequences;

[0096] Let g be the final scheduled power of generator g, which constitutes the complete resource scheduling strategy.

[0097] S4. Based on resource scheduling strategy, multimodal response verification of multi-source measurement data is performed through topology dynamic verification, and command response verification report is generated.

[0098] Based on resource scheduling strategies and multi-source measurement data, abnormal nodes and communication latency are identified through spatiotemporal alignment cross-validation.

[0099] It should be noted that, firstly, the issuance timestamps corresponding to the generator start-up and shutdown commands in the resource scheduling strategy are extracted, and the voltage and current amplitudes in the multi-source measurement data are captured synchronously. A spatiotemporal alignment operation is performed to align the voltage and current amplitudes with the timestamp synchronization signal to generate command issuance timestamps and PMU measurement timestamps. The communication delay from the arrival of the generator start-up and shutdown command to the electrical quantity response is calculated, and the deviation of the resource scheduling strategy prediction value is obtained. Then, cross-validation is performed: by comparing the voltage amplitude change with the deviation of the resource scheduling strategy prediction value, electrical anomaly nodes are marked, and the period of sudden change in acoustic signature intensity is detected. Based on the time difference between the occurrence of the electrical anomaly node and the command issuance timestamp during its sudden change in acoustic signature intensity, the true anomaly node is determined. Finally, a list of anomaly node locations and corresponding communication delays are output.

[0100] Furthermore, communication latency is a quantified value of the time delay between command transmission and execution during the power grid demand response process. It is defined as the total time difference from the issuance of a point-to-point control command from the dispatching end to the terminal's execution and generation of an electrical response. Specifically, it originates from the spatiotemporal alignment cross-verification operation during the command response verification report generation process: extracting the time stamp of the resource scheduling strategy command issuance and the actual change time stamps of the voltage and current amplitudes of the local PMU in the multi-source measurement data, and using the time stamp synchronization signal as the absolute time reference to calculate the time difference. For abnormal nodes, the communication latency confirmed after verification in conjunction with the active period of the acoustic fingerprint is recorded in the communication latency statistics table field of the command response verification report. Its physical meaning directly reflects the reliability of the control link transmission and the timeliness of the terminal response.

[0101] By detecting sudden changes in sound intensity, dual-modal analysis is performed to extract the voiceprint features from the device's voiceprint signal, and the active periods of the voiceprint are obtained through spatiotemporal binding.

[0102] It should be noted that, firstly, short-time energy calculation is performed on the device's acoustic signature signal, and the acoustic intensity abrupt change interval is identified; within the acoustic intensity abrupt change interval, the Mel frequency cepstral coefficient method is used to extract acoustic signature features, for example, a thirteen-dimensional Mel frequency cepstral coefficient feature vector is generated every ten milliseconds; at the same time, the start time and end time markers of the acoustic intensity abrupt change are recorded based on the time-stamp synchronization signal; the spatiotemporal binding operation associates the acoustic signature feature vector with the corresponding time-stamp interval to generate a structured record of the acoustic signature active period, which includes the start time marker, the end time marker, and the acoustic signature feature vector sequence.

[0103] Based on abnormal nodes, communication latency, and active periods of voiceprints, a command response verification report is generated through structured encapsulation.

[0104] It should be noted that, firstly, the root node is established and named; the list of abnormal node locations is sorted by node ID and then encapsulated into a child node group, and each abnormal node is recorded with an entry to obtain the node ID, abnormality type code, and abnormality confidence percentage; the communication latency is encapsulated, and each record includes the starting terminal ID, the target terminal ID, and the latency value (in milliseconds); the voiceprint active period record is converted into a child node group, and each record is written with a start time stamp, an end time stamp, and a voiceprint feature vector hash value, and finally, an instruction response verification report is generated.

[0105] S5. Based on the instruction response verification report, generate a topology reorganization decision by statistically analyzing the percentage of abnormal nodes in the cluster, and optimize the resource scheduling strategy through consistency verification.

[0106] The number of abnormal nodes is counted based on the command response verification report, and an abnormal node distribution mapping table is generated through cluster affiliation mapping.

[0107] It should be noted that, firstly, the command response verification report is parsed to extract all abnormal node entries within the child node group and count the total number of abnormal nodes; then, the cluster unit member list in the cluster topology data is loaded, and the cluster affiliation mapping operation is performed: each abnormal node entry is traversed, the cluster unit identifier to which its node ID belongs is matched, and the number of abnormal nodes in that cluster unit is recorded; then, by calculating the proportion of abnormal nodes in each cluster unit, an abnormal node distribution mapping table is finally generated.

[0108] Based on the abnormal node distribution mapping table, the health index is obtained by the proportion of abnormal nodes, and a cluster status table is generated.

[0109] It should be noted that, firstly, the abnormal node proportion field of the abnormal node distribution mapping table is read, and the health index is calculated using the health index formula; then, a cluster status table data structure is established, including cluster unit identifier, health index, number of abnormal nodes, and health status level (leveling rules: health index ≥ 90 is defined as healthy level, 70 ≤ health index < 90 is defined as warning level, and health index < 70 is defined as fault level); all rows of the abnormal node distribution mapping table are traversed, and the cluster unit identifier, calculated health index, original number of abnormal nodes, health status level determined according to the health index, and current Coordinated Universal Time timestamp of each row are sequentially filled into the corresponding column of the cluster status table; finally, a cluster status table containing complete status parameters of all cluster units is generated. Obtain the health index. The expression is: ,

[0110] in, This represents the total number of cluster unit nodes recorded in the cluster status table. This represents the number of abnormal nodes in the cluster unit as counted in the cluster status table. It is an exponential decay penalty term for communication latency on the cluster health status. This is the arithmetic mean of the communication delays of all abnormal nodes in the command response verification report. This is the time delay attenuation coefficient constant (the value range is: a real number greater than zero); This indicates a health index.

[0111] Furthermore, the abnormal node ratio is a quantitative indicator obtained from the abnormal node distribution mapping table. Specifically, it is defined as the percentage of the number of abnormal nodes in the abnormal node distribution mapping table to the total number of nodes in the cluster unit of the abnormal node distribution mapping table. Its physical meaning represents the proportion of failed or abnormal nodes in a specific cluster unit, and it is directly used to drive the calculation of the health index and the topology reorganization decision trigger judgment.

[0112] Based on the cluster status table, low health indices are filtered using discrete confidence levels, and topology reorganization decisions are generated.

[0113] It should be noted that, firstly, the discrete confidence level screening rule is set to select records in the cluster status table that are marked as fault level in the health status classification field (i.e., cluster units with a health index less than 70); then, each row of the cluster status table is traversed, and when the health index is less than 70 and the health status classification field is equal to the fault level, the cluster unit identifier of that row is extracted; and a topology reorganization decision is generated based on the set of filtered cluster unit identifiers.

[0114] Furthermore, the discrete confidence score is a classification label for the health status grading field in the cluster status table. It generates three discrete states based on the range of health index values: a health index value greater than or equal to 90 is labeled as healthy; a health index value between 70 and 90 is labeled as warning; and a health index value less than 70 is labeled as faulty. This parameter originates from the direct mapping rule based on the health index during the cluster status table generation process and is used to determine the triggering conditions for topology reorganization decisions.

[0115] The topology reconfiguration decision adopts a feedback mechanism to execute the opening and closing of physical circuit breakers, and optimizes the resource scheduling strategy by reallocating 5G slice resources.

[0116] First, the target cluster unit identifier list in the topology reconfiguration decision is parsed, and the cluster unit identifiers are converted into corresponding circuit breaker number sets through the physical device mapping table. Then, the physical circuit breaker opening and closing operations are performed, disconnecting the communication circuit breaker of the fault-level cluster unit according to the reconfiguration instruction code rules, while simultaneously closing the backup link circuit breaker. Simultaneously, the 5G slice resource reallocation process is triggered, dynamically adjusting the slice type according to the health index of the cluster status table—cluster units with a health index below 40 are switched to URLLC slices, and cluster units with a health index between 40 and 70 adopt eMBB slices. The communication latency factor of the resource scheduling strategy is also updated (the latency factor of URLLC slices is set to 10ms, and the latency factor of eMBB slices is set to 20ms).

[0117] This embodiment also provides a computer device applicable to a novel power grid adjustable resource demand response method based on wide-area measurement data, comprising: 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 novel power grid adjustable resource demand response method based on wide-area measurement data proposed in the above embodiment.

[0118] 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.

[0119] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the novel power grid adjustable resource demand response method based on wide-area measurement data 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.

[0120] In summary, this invention achieves deep coupling and collaboration of response clusters by integrating the physical characteristics of the power grid with the communication topology to construct an intelligent metric tensor, transforming complex resource scheduling into parallel optimization through quantum constraint coding, breaking through the bottleneck of large-scale real-time decision-making, and accurately intercepting abnormal nodes by combining an acoustic-electric dual-modal verification mechanism. Ultimately, it optimizes the core obstacles in traditional methods, such as topology fragmentation, optimization lag, and lack of security verification.

[0121] 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 novel power grid adjustable resource demand response method based on wide-area measurement data, characterized in that: include, Collect multi-source measurement data of the power grid and obtain slice datasets through preprocessing; Based on the slice dataset, a metric tensor is constructed through metric mapping, and cluster topology data is generated using the Riemann clustering algorithm based on geodesic distance; Based on cluster topology data, a Hamiltonian is constructed using quantum mapping, and a resource scheduling strategy is generated using the quantum annealing algorithm. Based on resource scheduling strategies, multimodal response verification of multi-source measurement data is performed through topology dynamic verification, and a command response verification report is generated. Based on the instruction response verification report, the topology reorganization decision is generated by statistically analyzing the percentage of abnormal nodes in the cluster, and the resource scheduling strategy is optimized through consistency verification.

2. The novel power grid adjustable resource demand response method based on wide-area measurement data as described in claim 1, characterized in that: The multi-source measurement data includes acoustic signature signal, voltage amplitude, current amplitude, electrical phase angle difference, frequency dynamic change rate, time-scale synchronization signal, and data quality code; The preprocessing includes abnormal data cleaning, time-scale synchronization correction, and feature enhancement filtering.

3. The novel power grid adjustable resource demand response method based on wide-area measurement data as described in claim 2, characterized in that: The specific steps for constructing a metric tensor based on the sliced ​​dataset using metric mapping are as follows. Based on the 5G slice type in the slice dataset, communication parameters are obtained through multimodal parameter synchronous sensing, and feature transformation is performed through Cartesian tensor fusion decomposition to generate coupling strength factors. The coupling coefficient is obtained through nonlinear field coupling based on the coupling strength factor and the electrical phase angle difference, and the metric tensor is constructed using the exponential positive definite symmetric method.

4. The novel power grid adjustable resource demand response method based on wide-area measurement data as described in claim 3, characterized in that: The specific steps for generating cluster topology data based on geodesic distance using the Riemann clustering algorithm are as follows: The inertial intensity sequence is obtained by numerical mapping of the metric tensor using the Ricci scalar method, and the geodesic distance is generated using the Dijkstra algorithm. Primary clusters are obtained using the spatial proximity method based on geodesic distance and inertial intensity sequences, and cluster topology data is generated through structured encapsulation.

5. The novel power grid adjustable resource demand response method based on wide-area measurement data as described in claim 4, characterized in that: The specific steps for constructing the Hamiltonian based on cluster topology data using quantum mapping are as follows. The qubit layout is generated by index mapping based on cluster topology data, and the boundary constraint encoding is obtained by Pauli operators. Based on the qubit layout and boundary constraint encoding, the Hamiltonian is constructed by linear superposition through adjacency matrix filtering.

6. The novel power grid adjustable resource demand response method based on wide-area measurement data as described in claim 5, characterized in that: The resource scheduling strategy generated by the quantum annealing algorithm includes generating quantum state wave functions by uniformly superimposing qubits using transverse field pulses based on Hamiltonians, and generating resource scheduling strategies by probability mapping in combination with cluster topology data.

7. The novel power grid adjustable resource demand response method based on wide-area measurement data as described in claim 6, characterized in that: The method based on resource scheduling strategy, using topology dynamic verification to perform multimodal response verification on multi-source measurement data, and generating a command response verification report, involves the following specific steps. Based on resource scheduling strategies and multi-source measurement data, abnormal nodes and communication latency are identified through spatiotemporal alignment cross-validation. By detecting abrupt changes in sound intensity, dual-modal analysis is performed to extract voiceprint features from the voiceprint signal, and active periods of the voiceprint are obtained through spatiotemporal binding. Based on abnormal nodes, communication latency, and active periods of voiceprints, a command response verification report is generated through structured encapsulation.

8. The novel power grid adjustable resource demand response method based on wide-area measurement data as described in claim 7, characterized in that: The process involves verifying the command response report, statistically analyzing the percentage of abnormal nodes within the cluster, generating a topology reorganization decision, and optimizing resource scheduling strategies through consistency checks. The specific steps are as follows: The number of abnormal nodes is counted based on the command response verification report, and an abnormal node distribution mapping table is generated through cluster affiliation mapping. Based on the abnormal node distribution mapping table, the health index is obtained by the proportion of abnormal nodes, and a cluster status table is generated. Based on the cluster status table, low health indices are filtered using discrete confidence levels, and topology reorganization decisions are generated. The topology reconfiguration decision adopts a feedback mechanism to execute the opening and closing of physical circuit breakers, and optimizes the resource scheduling strategy by reallocating 5G slice resources.

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 a novel power grid adjustable resource demand response method based on wide-area measurement data 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 a novel power grid adjustable resource demand response method based on wide-area measurement data as described in any one of claims 1 to 8.

Images