CN-120613790-B - Dynamic digital twin-based power grid evolution behavior modeling method and system

CN120613790BCN 120613790 BCN120613790 BCN 120613790BCN-120613790-B

Abstract

The invention provides a dynamic digital twin-based power grid evolution behavior modeling method and system, and relates to the technical field of digital twin of smart power grids. According to the method, interaction characteristics of power station output, user load response and energy storage scheduling strategies are captured in real time through a digital twin interface, a behavior coupling factor containing complementarity and conflict weights is formed, and an implicit tensor fusion technology and nonlinear projection are utilized to construct a power grid response relation and a multi-scale demand elastic curved surface. And then, through the steps of virtual behavior anchor point disturbance and the like, the offset is separated to identify an unbalanced region, an evolution track is generated, an optimal cooperative path is extracted, and the behavior coupling factor weight is dynamically adjusted, so that the rapid convergence of the global cooperative stable domain is realized. The invention covers modules such as behavior coupling factor generation and the like, supports various operations, effectively solves various problems in power grid resource interaction, and improves dynamic adaptability and collaborative efficiency of power grid evolution modeling.

Inventors

CHANG GUOGANG
Hou yinchuan
LIU SHUHE
QU JIAN
XU ZHENG
ZHANG JINGHUI
Xu Meijin
XU LIANGFENG

Assignees

北京妙微科技有限公司

Dates

Publication Date: 20260505
Application Date: 20250613

Claims (8)

1. The power grid evolution behavior modeling method based on dynamic digital twin is characterized by comprising the following steps of: S1, capturing output characteristics of a power station, a user load response mode and an energy storage operator scheduling strategy in real time through a digital twin interface, and abstracting the output characteristics, the user load response mode and the energy storage operator scheduling strategy into behavior coupling factors, wherein the behavior coupling factors comprise potential complementarity and conflict weights of resource interaction between entities; S2, constructing a power grid response relation based on a behavior coupling factor, mapping output fluctuation of a power station into a user side demand elastic curved surface through nonlinear projection, and forming a power grid decision space with an energy storage charging and discharging strategy, wherein the connection strength between nodes in the power grid response relation is reversely corrected by the historical conflict rate and the real-time complementarity of entity behaviors; S3, perturbing the power grid response relationship by introducing a virtual behavior anchor point to generate a power grid evolution model; S4, disturbing a power grid response relation through a virtual behavior anchor point, triggering a coupling chain local resonance, and generating a model evolution track reflecting resource interaction dynamics; S5, injecting the optimal cooperative paths into a resistance evolution channel of the digital twin, detecting cooperative deviation through virtual-real behavior chain resonance, and dynamically adjusting the fusion weight of the behavior coupling factors until the deviation is lower than a preset tolerance threshold; the behavior coupling factors realize asymmetric complementary configuration of power stations, users and energy storage operator resources in space-time dimension through coupling driving optimization of implicit tensor fusion and power grid response relation, and the power grid decision space enables the behavior of a power grid evolution model to be converged into a global collaborative stable domain through chain resonance constraint of decision verification; the virtual behavior anchor point is an orthogonal coupling chain based on a power grid decision space, a breakpoint of resource interaction in historical evolution is extracted, a complementary path weight attenuation extremum and a conflict path mutation threshold value are combined to cross and position a space-time heterogeneous anchor point, an anchor point with the largest deviation from real-time complementarity is screened, virtual output or load disturbance is injected into an associated node of the anchor point through digital twinning to cause local resonance of the coupling chain, a track offset after disturbance is separated through orthogonal basis vector independence, a novel breakpoint which is not covered by historical conflict is identified, the novel breakpoint is fed back to a complementary path pruning rule, the conflict threshold value is dynamically updated, and convergence of the coupling chain after correction is verified under the deviation measurement of a power grid evolution stability domain.
2. The method for modeling power grid evolution behavior based on dynamic digital twin according to claim 1, wherein the process of abstracting the power grid evolution behavior into behavior coupling factors in the step S1 is as follows: S11, inputting time sequence data of power station output, user load and energy storage strategies into a three-dimensional implicit tensor model, decomposing output fluctuation periodic characteristics, extracting a load elastic distribution mode, analyzing an energy storage regulation rule, and extracting implicit cross-entity interaction characteristics in three decomposition residual errors; s12, generating a dynamic mask matrix based on historical grid conflict events, pruning a power station-energy storage resource competition path in the interaction characteristics, and reserving a power station-user supply and demand complementary path; S13, calculating the difference between the adjustment delay of the power station and the response delay of the user in the complementary path, dynamically weakening the original matching degree through an exponential decay function, and generating an effective complementary coefficient; S14, carrying out double-channel fusion on the effective complementary coefficient and the real-time adjustment difference degree of power station-energy storage, carrying out complementary term weighted superposition, and carrying out conflict term threshold filtering to generate a coupling matrix; S15, detecting force mutation or load distortion in real time, triggering weight adjustment, enhancing complementary path weight and inhibiting conflict items, and correcting pruning rules and attenuation parameters through digital twin feedback.
3. The power grid evolution behavior modeling method based on dynamic digital twin according to claim 1, wherein the power grid response relation construction process in the step S2 is as follows: S21, mapping output fundamental frequency and high-frequency components of a power station to a price sensitive area and an overload risk area at a user side through non-orthogonal projection based on output period characteristics and a load elastic mode, generating a multi-scale elastic curved surface required, and capturing a distortion diffusion path of output fluctuation at the user side; S22, dispersing charge and discharge rate and capacity margin into space-time gradient units according to an energy storage regulation rule, and establishing a buffer association matrix of the space-time gradient units and a supply and demand unbalanced region in an elastic curved surface, wherein association strength is regulated and controlled by a complementary path after pruning in the step S1, and conflict paths are eliminated through threshold filtering; s23, dynamically fusing conflict suppression factors and real-time complementarity based on the effective complementary coefficients and the historical conflict rate, and weighting the initial connection strength; s24, expanding and fusing the required elastic curved surface and the energy storage buffer matrix according to the implicit tensor basis in the step S1, embedding the corrected dynamic strength as orthogonal dimension constraint, and forming a power grid decision space containing an asymmetric coupling chain, wherein the orthogonality guarantees the resolvability of the disturbance track of the anchor point of the subsequent virtual behavior and the deviation measure of the stable domain.
4. The power grid evolution behavior modeling method based on dynamic digital twin according to claim 1, wherein the power grid evolution model is constructed as follows: S31, decomposing resonance response of the coupling chain in an orthogonal decision space based on virtual disturbance, dynamically adjusting connection weight of the power station-user-energy storage node according to the track offset, and reconstructing a disturbance propagation path; s32, utilizing the orthogonal basis vectors to separate the evolution tracks of all the nodes after disturbance, calculating the superposition quantity of the power station output deviation, the user load distortion rate and the energy storage adjustment hysteresis, and extracting the hidden unbalance area which is not covered by the conflict pruning; S33, mapping the hidden unbalance area to a coupling matrix, dynamically updating the complementary path effective coefficient and conflict item filtering rule, and reversely correcting the elastic curved surface mapping parameter through the digital twin interface; S34, verifying convergence of the corrected coupling chain under the stable domain deviation measurement, and triggering secondary disturbance of the novel anchor point if the threshold value is not reached.
5. The method for modeling power grid evolution behavior based on dynamic digital twin according to claim 4, wherein the identifying of the hidden unbalance area in the step S32 comprises the following steps: s321, performing time domain convolution on the power station output deviation and conflict weights in the behavior coupling factors to generate conflict association vectors, and performing inverse gradient matching on the user load distortion rate and the complementary coefficients to generate complementary association vectors; S322, based on the track offset direction after the disturbance of the virtual behavior anchor points, carrying out dynamic tensor superposition on the conflict association vector and the complementary association vector according to the basis vector of the orthogonal power grid decision space to generate an association matrix containing space-time game features, and carrying out abnormal path filtering through the historical conflict paths in the dynamic mask matrix; S323, if the increasing rate of the conflict weight in the association matrix exceeds the dynamic attenuation threshold value of the complementary coefficient, judging the power station-energy storage strategy conflict area; S324, carrying out space-time alignment on the judgment result and the complementary path after pruning, removing the repeated area of the event with the historical conflict, generating a dynamic map only containing novel unbalance, and reversely correcting the mapping weight of the unbalance supply and demand area in the elastic curved surface and the updating step length of the coupling matrix.
6. The power grid evolution behavior modeling method based on dynamic digital twinning is characterized in that the model evolution track is triggered by virtual behavior anchor point disturbance, a dynamic state transition sequence of a power grid evolution model subjected to virtual output or load disturbance in a power grid decision space is characterized, the generation process is that track offset after disturbance is separated through orthogonal basis vectors, a local resonance response of a coupling chain is combined, a space-time evolution path comprising power station output deviation, user load distortion and energy storage regulation hysteresis is formed, the extraction of an optimal collaborative path is based on a deviation minimization principle, and the evolution track is converged to a target stable domain through reverse correction of complementary path weights and conflict thresholds.
7. The power grid evolution behavior modeling method based on dynamic digital twin is characterized by comprising the steps of injecting an optimal collaborative path into a subtropic evolution channel of a digital twin, separating space-time coupling components in track offset caused by disturbance of a virtual behavior anchor point based on independence of orthogonal basis vectors, constructing a response vector of chain resonance, carrying out tensor inner product operation on the response vector and dynamic strength of a coupling matrix by combining boundary constraint conditions of a target stability domain, generating a collaborative bias index, dynamically adjusting fusion weight of a behavior coupling factor if the collaborative bias index exceeds a preset tolerance threshold, reversely correcting a dynamic mask matrix and complementary coefficient attenuation parameters through implicit tensor basis expansion, enhancing complementary path weight, inhibiting conflict items, and triggering secondary response of local resonance of the coupling chain.
8. A dynamic digital twin-based power grid evolution behavior modeling system capable of implementing the method of any one of claims 1-7, the system comprising: The behavior coupling factor generation module captures the output characteristics of the power station, the user load response mode and the energy storage operator dispatching strategy in real time through the digital twin interface and abstracts the power station output characteristics, the user load response mode and the energy storage operator dispatching strategy into behavior coupling factors, wherein the behavior coupling factors comprise potential complementarity and conflict weights of resource interaction among entities; The power grid evolution behavior model module is used for constructing a power grid response relation based on a behavior coupling factor, mapping output fluctuation of a power station into a user side demand elastic curved surface through nonlinear projection, and forming a power grid decision space with an energy storage charging and discharging strategy, wherein the connection strength between nodes in the power grid response relation is reversely corrected by the historical conflict rate and the real-time complementarity of entity behaviors; The collaborative path optimization module is used for generating a power grid evolution model by introducing a virtual behavior anchor point to disturb a power grid response relation, and extracting an optimal collaborative path based on the deviation degree of a model evolution track and a target stability domain, wherein the deviation degree is defined as a composite function of resource mismatch degree, collaborative delay cost and evolution inertia resistance; the feedback control module is used for injecting the optimal cooperative path into a resistance evolution channel of the digital twin body, detecting cooperative deviation through virtual-real behavior chain resonance, and dynamically adjusting the fusion weight of the behavior coupling factors until the deviation is lower than a preset tolerance threshold; the virtual behavior anchor point is an orthogonal coupling chain based on a power grid decision space, a breakpoint of resource interaction in historical evolution is extracted, a complementary path weight attenuation extremum and a conflict path mutation threshold value are combined to cross and position a space-time heterogeneous anchor point, an anchor point with the largest deviation from real-time complementarity is screened, virtual output or load disturbance is injected into an associated node of the anchor point through digital twinning to cause local resonance of the coupling chain, a track offset after disturbance is separated through orthogonal basis vector independence, a novel breakpoint which is not covered by historical conflict is identified, the novel breakpoint is fed back to a complementary path pruning rule, the conflict threshold value is dynamically updated, and convergence of the coupling chain after correction is verified under the deviation measurement of a power grid evolution stability domain.

Description

Dynamic digital twin-based power grid evolution behavior modeling method and system Technical Field The invention relates to the technical field of digital twin of smart power grids, in particular to a power grid evolution behavior modeling method and system based on dynamic digital twin. Background Along with the evolution of the energy system to an intelligent direction, the application of the digital twin technology in the modeling of the power grid is gradually deepened, and the digital twin technology is mainly used for mirror image mapping and state deduction of a physical system and a virtual space. In the prior art, a local simulation model is built based on single-dimension data, or a multi-entity interaction relation is approximated by a linear superposition mode. However, with the increase of the permeability of distributed resources in the power grid and the complexity of the interaction scene, the multi-entity space-time behavior game presents high dynamic property and asymmetry, and the limitation of the prior art is increasingly highlighted. These problems make it difficult to achieve efficient synergy and dynamic optimization in the space-time dimension for the grid to respond to high proportions of renewable energy fluctuations and customer-side demands. Disclosure of Invention In order to achieve the above purpose, the invention provides a dynamic digital twin-based power grid evolution behavior modeling method and system; in a first aspect, an embodiment of the present invention provides a dynamic digital twin-based power grid evolution behavior modeling method; The method comprises the steps of capturing output characteristics of a power station, a user load response mode and an energy storage operator dispatching strategy in real time through a digital twin interface, abstracting the output characteristics, the user load response mode and the energy storage operator dispatching strategy into behavior coupling factors, extracting potential complementarity and conflict weight of resource interaction among entities through an implicit tensor fusion technology, further inputting time sequence data of the output power of the power station, the user load and the energy storage strategy into a three-dimensional implicit tensor model, decomposing output fluctuation periodic characteristics, load elastic distribution modes and energy storage regulation rules, extracting cross-entity interaction characteristics, generating a dynamic mask matrix based on historical grid conflict events, pruning a resource competition path of the power station-energy storage, reserving supply and demand complementation paths of the power station-user, calculating the difference of the adjustment delay of the power station and the user response delay, dynamically weakening original matching degree through an exponential decay function, generating an effective complementation coefficient, carrying out double-channel fusion on the effective complementation coefficient and the real-time adjustment difference of the power station-energy storage, weighting superposition of the complementation item, filtering the conflict item threshold, forming the coupling matrix, detecting the power mutation or load distortion in real time, triggering weight regulation to strengthen the complementation path weight and restrain conflict item, and correcting pruning rules and attenuation parameters through digital twin feedback. The method comprises the steps of constructing a power grid response relation based on a behavior coupling factor, mapping output fluctuation of a power station into a user side demand elastic curved surface through nonlinear projection, forming a power grid decision space with an energy storage charging and discharging strategy, mapping fundamental frequency and high frequency components of output of the power station to a user side price sensitive area and an overload risk area through non-orthogonal projection by utilizing output periodic characteristics and a load elastic mode, generating a multi-scale demand elastic curved surface, capturing a distortion diffusion path of the output fluctuation at the user side, dispersing a charging and discharging rate and a capacity margin into a space-time gradient unit according to an energy storage regulation rule, establishing a buffer association matrix of the space-time gradient unit and a supply and demand unbalanced area in the elastic curved surface, wherein association strength is regulated by the complementary path after pruning, dynamically fusing a conflict inhibition factor and real-time complementarity to construct initial connection strength based on an effective complementary coefficient and a real-time complementary degree, expanding and fusing the demand elastic curved surface and the energy storage buffer matrix according to an implicit tensor basis, embedding the corrected dynamic strength as orthogonal dimension constraint to form a power grid stabil