Flood control scheduling method and system based on digital twinning

Abstract

The invention discloses a flood control scheduling method and system based on digital twinning, and the method comprises the steps: collecting time sequence satellite interference measurement data, obtaining a ground deformation rate and a coherence coefficient, calculating the leakage risk amount through a geological leakage model, and generating a risk data set; converting the risk data set into a boundary condition of fluid dynamic simulation, and performing flood routing simulation to generate a flood routing map; the flood routing map comprises a predicted submerging range and a predicted submerging water depth; based on the flood routing map, optimizing a gate topological relation through a map convolutional network, and generating a gate opening degree control instruction set; the gate opening control instruction set is injected into the digital twinborn body for simulation correction, the submerging water depth after correction is obtained, a submerging deviation value is calculated in combination with the predicted submerging water depth, and a verification report is generated; according to the method, the quantitative modeling of the underground leakage process is realized by collecting the time sequence satellite interference measurement data and constructing the structured risk data set.

Description

Technical Field

[0001] This invention relates to the field of flood control scheduling technology, and in particular to a flood control scheduling method and system based on digital twins. Background Technology

[0002] In recent years, with the rapid development of cutting-edge technologies such as remote sensing monitoring, artificial intelligence, and digital twins, flood control scheduling has gradually shifted from a traditional experience-driven model to a data-driven and model-driven model. Satellite synthetic aperture radar interferometry (InSAR) technology, due to its high spatiotemporal resolution for monitoring surface deformation, has been widely used in areas such as land subsidence and geological hazard identification. Meanwhile, the development of fluid dynamics simulation combined with high-performance computing platforms has made it possible to simulate flood evolution at a regional scale.

[0003] Despite significant progress, several key bottlenecks remain in practical applications. First, traditional flood control scheduling lacks effective modeling methods for underground seepage processes, making it difficult to convert surface deformation information into quantifiable seepage risk inputs into flood simulation models. This leads to inaccurate source term boundary condition settings, affecting prediction accuracy. Second, existing scheduling decisions are mostly based on static topology optimization of gate control strategies, failing to fully consider the coupling relationship between dynamic flow field characteristics and network topology, making it difficult to achieve precise, real-time basin-wide coordinated control. Summary of the Invention

[0004] The purpose of this invention is to provide a flood control scheduling method and system based on digital twins, which solves the problems of inaccurate modeling of underground seepage processes and insufficient dynamic adaptability of gate control strategies in the prior art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] In a first aspect, the present invention provides a flood control scheduling method based on digital twins, comprising: acquiring time-series satellite interferometry data, obtaining the surface deformation rate and coherence coefficient, and calculating the leakage risk through a geological leakage model to generate a risk dataset; converting the risk dataset into boundary conditions for fluid dynamics simulation and performing flood evolution simulation to generate a flood evolution map; the flood evolution map includes the predicted inundation range and the predicted inundation depth; based on the flood evolution map, optimizing the gate topology relationship through a graph convolutional network to generate a gate opening control instruction set; injecting the gate opening control instruction set into the digital twin for simulation correction, obtaining the corrected inundation depth, and calculating the inundation deviation value in conjunction with the predicted inundation depth to generate a verification report; and controlling the gates according to the verification report, simultaneously acquiring multi-source leakage data, and feeding it back to the geological leakage model for optimization.

[0007] As a preferred embodiment of the flood control scheduling method based on digital twins described in this invention, the specific steps for generating the risk dataset are as follows:

[0008] Based on time-series satellite interferometry data, an interferometric phase map is generated through phase unwrapping, and permanent scatterers are identified to obtain the surface deformation rate and coherence coefficient.

[0009] The surface deformation rate is input into the geological leakage model to calculate the leakage risk, and the leakage risk probability is calculated by combining the coherence coefficient.

[0010] The risk dataset is generated by combining the surface deformation rate, coherence coefficient, leakage risk quantity, and leakage risk probability according to geographic coordinates.

[0011] As a preferred embodiment of the flood control scheduling method based on digital twins described in this invention, the specific steps for generating the flood evolution map are as follows:

[0012] The risk dataset is input into a quantum graph neural network, which outputs a spatiotemporally continuous leakage source term tensor.

[0013] The leakage source term tensor is used as the boundary condition for fluid dynamics simulation. The fluid velocity field vector is calculated through the fluid dynamics equations to generate the original flow field data.

[0014] The original flow field data is fused with real-time monitoring data by using a quantum compressed sensing operator to generate corrected flow field data;

[0015] The flow field data is processed by a fractal geometric feature extraction algorithm to generate a flood evolution map.

[0016] As a preferred embodiment of the flood control scheduling method based on digital twins described in this invention, the specific steps for generating the gate opening control instruction set are as follows:

[0017] The fractal features in the flood evolution map are associated with the risk dataset and then linked to the gate nodes to generate a modulated set of node attributes.

[0018] The node attribute set is input into the quantum-enhanced graph convolutional network, the output node state set is optimized to generate the gate connection weight matrix;

[0019] The gate connection weight matrix is ​​fused with the real-time flow field vorticity characteristics to generate a gate opening control command set.

[0020] As a preferred embodiment of the flood control scheduling method based on digital twins described in this invention, the specific steps for injecting the gate opening control command set into the digital twin for simulation correction and obtaining the corrected submerged water depth are as follows:

[0021] The gate opening control instruction set is encoded into the quantum bit rotation angle, and the quantum phase is modulated according to chaotic dynamics to generate the initial quantum state;

[0022] The initial quantum state is input into the digital twin, and the final quantum wave function is obtained by time evolution.

[0023] Compressed sensing measurements are performed on the final-state quantum wavefunction to generate the original flood depth.

[0024] The original submerged depth is differentiated by error function and the chaotic parameters are calculated to generate the corrected submerged depth.

[0025] As a preferred embodiment of the digital twin-based flood control scheduling method of the present invention, the specific steps for calculating the inundation deviation value by combining the predicted inundation depth and generating a verification report are as follows:

[0026] Based on the corrected flood depth and the predicted flood depth, the flood deviation value is calculated using quantum state fidelity.

[0027] The flooding deviation values ​​are correlated with spatial coordinates to generate a spatiotemporal deviation field, which is then mapped to the chaotic attractor phase space. Risk values ​​are generated by integrating along the trajectory.

[0028] The cumulative risk field is generated by performing fractional integral and spatial convolution on the risk value.

[0029] The cumulative risk field is encoded as a quantum hologram, and combined with chaotic timestamps and dynamic risk classification to generate a verification report.

[0030] As a preferred embodiment of the digital twin-based flood control scheduling method of the present invention, the steps of controlling the gates according to the verification report, simultaneously collecting multi-source seepage data, and feeding it back to the geological seepage model for optimization are as follows:

[0031] Based on the verification report, gate control is implemented, and multi-source leakage data is collected simultaneously.

[0032] The multi-source leakage data includes seepage flow rate, crack distribution data, and soil data;

[0033] The seepage flow rate, crack distribution data, and soil data are fused into a spatiotemporal seepage tensor, and the geological seepage model is dynamically optimized through chaotic integral operators and quantum verification.

[0034] Secondly, this invention provides a flood control scheduling system based on digital twins, comprising a data acquisition module, a flood simulation module, a gate control module, a simulation verification module, and a feedback calibration module. The data acquisition module is used to acquire time-series satellite interferometry data, obtain the surface deformation rate and coherence coefficient, and calculate the leakage risk through a geological leakage model to generate a risk dataset. The flood simulation module is used to convert the risk dataset into boundary conditions for fluid dynamics simulation and perform flood evolution simulation to generate a flood evolution map; the flood evolution map includes the predicted inundation range and predicted inundation depth. The gate control module is used to optimize the gate topology based on the flood evolution map using a graph convolutional network to generate a gate opening control instruction set. The simulation verification module is used to inject the gate opening control instruction set into the digital twin for simulation correction, obtain the corrected inundation depth, and calculate the inundation deviation value based on the predicted inundation depth to generate a verification report. The feedback calibration module is used to control the gates according to the verification report, simultaneously acquire multi-source leakage data, and feed it back to the geological leakage model for optimization.

[0035] Thirdly, 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 flood control scheduling method based on digital twins as described in the first aspect of the present invention.

[0036] Fourthly, 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 flood control scheduling method based on digital twins as described in the first aspect of the present invention.

[0037] The beneficial effects of this invention are as follows: by collecting time-series satellite interferometry data and constructing a structured risk dataset, quantitative modeling of the underground seepage process is realized, improving the accuracy of boundary conditions in flood evolution simulation and enhancing the reliability of prediction; furthermore, by optimizing the gate topology relationship based on graph convolutional networks and generating control instruction sets, the coordinated optimization of dynamic flow field and network topology is realized, improving the real-time performance and coordination of basin-level flood control scheduling. Attached Figure Description

[0038] 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.

[0039] Figure 1 This is a flowchart of a flood control scheduling method based on digital twins.

[0040] Figure 2 A flowchart for generating a risk dataset.

[0041] Figure 3 A flowchart for generating a flood evolution map.

[0042] Figure 4 A flowchart for generating the gate opening control instruction set. Detailed Implementation

[0043] 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.

[0044] 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.

[0045] 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.

[0046] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a flood control scheduling method based on digital twins, comprising the following steps:

[0047] S1: Collect time-series satellite interferometry data, obtain the surface deformation rate and coherence coefficient, and calculate the leakage risk through a geological leakage model to generate a risk dataset.

[0048] S1.1: Based on time-series satellite interferometry data, an interferometric phase map is generated through phase unwrapping, and permanent scatterers are identified to obtain the surface deformation rate and coherence coefficient.

[0049] It should be noted that time-series satellite interferometry data is obtained by repeated orbital observations of the same area by synthetic aperture radar satellites. Time-series satellite interferometry data includes multi-temporal synthetic aperture radar image information, satellite orbital parameters, imaging time information, and radar equipment parameters.

[0050] Furthermore, the synthetic aperture radar image information is unwrapped using a minimum cost flow algorithm to generate an interferometric phase map. Permanent scatterer targets are screened using a dual standard of amplitude deviation index and phase stability. Based on the phase time series of permanent scatterers, the surface deformation rate is extracted through linear regression analysis. At the same time, the phase consistency of adjacent pixels is obtained to obtain the coherence coefficient. The interferometric phase map reflects the phase change characteristics of surface deformation. The identification of permanent scatterers ensures the stability of the monitored targets. The surface deformation rate quantifies the annual change of surface displacement, and the coherence coefficient characterizes the stability of the radar echo signal.

[0051] S1.2: Input the surface deformation rate into the geological seepage model to calculate the seepage risk, the expression is:

[0052] Q = k·e μ·ΔV ;

[0053] Where Q represents the leakage risk, k represents the bedrock permeability coefficient (0.2≤k≤0.35), μ represents the deformation sensitivity factor, and ΔV represents the surface deformation rate.

[0054] The specific process includes inputting the surface deformation rate into the geological leakage model, and then calculating the leakage risk using the bedrock permeability coefficient and deformation sensitivity factor. The bedrock permeability coefficient reflects the inherent permeability characteristics of the geological structure, the deformation sensitivity factor characterizes the weighting coefficient of the surface deformation on the leakage, and the surface deformation rate directly reflects the intensity of surface displacement change at the monitoring point. The bedrock permeability coefficient, deformation sensitivity factor, and surface deformation rate work together to generate a quantified leakage risk.

[0055] Furthermore, the pre-training process of the geological seepage model requires the collection of historical time-series satellite interferometry data, including surface deformation rate, coherence coefficient, and actual seepage event records. Surface deformation features are extracted using phase unwrapping and permanent scatterer identification techniques. An initial seepage risk calculation relationship is constructed by combining the bedrock permeability coefficient and deformation sensitivity factor. The deformation sensitivity factor is calibrated using historical seepage event data. The backpropagation algorithm is used to optimize the geological seepage model parameters to minimize the error between the seepage risk output by the geological seepage model and the measured seepage amount. Cross-validation is used during the training process to prevent overfitting. Finally, a pre-trained geological seepage model that can accurately reflect the mapping relationship between surface deformation rate and seepage risk is obtained.

[0056] S1.3: Calculate the leakage risk probability based on the coherence coefficient, the expression is:

[0057]

[0058] Where P represents the probability of leakage risk, C represents the coherence coefficient (0≤C≤1), and λ represents the attenuation coefficient (0.3<λ<0.7).

[0059] The specific process involves the coherence coefficient reflecting the stability and reliability of surface deformation monitoring data. The leakage risk probability ranges from zero to one, with higher values ​​indicating more reliable monitoring results. The leakage risk probability is calculated using the attenuation coefficient of the coherence coefficient, which ranges from 0.3 to 0.7 and is used to adjust the influence of the coherence coefficient on the leakage risk probability. When the coherence coefficient is low, the surface deformation monitoring data has higher noise, and the leakage risk probability increases accordingly; conversely, when the coherence coefficient is high, the reliability of the monitoring data is enhanced, and the leakage risk probability decreases.

[0060] S1.4: Combine the surface deformation rate, coherence coefficient, leakage risk quantity, and leakage risk probability according to geographic coordinates to generate a risk dataset.

[0061] The specific process includes: the surface deformation rate characterizes the intensity of surface displacement change at the monitoring point; the coherence coefficient reflects the reliability of deformation monitoring data; the leakage risk is calculated through a geological leakage model; the leakage risk probability is determined by the coherence coefficient and the attenuation coefficient; and through geographic information spatial registration technology, precise matching is performed according to latitude and longitude coordinates. The surface deformation rate, coherence coefficient, leakage risk, and leakage risk probability of each monitoring point are integrated into a structured record, ultimately forming a risk dataset containing spatial location information and multi-dimensional risk indicators.

[0062] S2: Convert the risk dataset into boundary conditions for fluid dynamics simulation and perform flood evolution simulation to generate a flood evolution map; the flood evolution map includes the predicted inundation range and the predicted inundation depth.

[0063] S2.1: Input the risk dataset into the quantum graph neural network and output a spatiotemporally continuous leakage source term tensor.

[0064] The specific process includes: the risk dataset contains the surface deformation rate, coherence coefficient, leakage risk quantity, and leakage risk probability associated with geographic coordinates; the quantum graph neural network processes the risk dataset through quantum state encoding, uses a graph structure to represent the topological relationship between monitoring points, performs feature extraction and spatiotemporal interpolation operations in the quantum circuit, and transforms the risk dataset into a continuously distributed leakage source term tensor. The leakage source term tensor has time and spatial dimensions and can fully describe the dynamic change characteristics of the leakage risk field in the study area.

[0065] S2.2: Using the leakage source term tensor as the boundary condition for fluid dynamics simulation, the fluid velocity field vector is calculated through the fluid dynamics equations to generate the original flow field data. The expression is:

[0066]

[0067] Where n represents the time step number, u n+1u represents the fluid velocity field vector at the (n+1)th time step. n This represents the fluid velocity field vector at the nth time step, Δt1 represents the time interval between two adjacent time variables, and ρ represents the fluid density. Let v represent the pressure field gradient at the nth time step, and v represent the fluid kinematic viscosity coefficient. Let represent the Laplace operator for the fluid velocity field vector at the nth time step, g represent the gravitational acceleration vector, f represent the Coriolis force, η represent the seepage-momentum conversion coefficient, and L2 represent the double logarithm function. This represents the spatial gradient of the leakage source term tensor.

[0068] The specific process includes: the leakage source term tensor contains the spatiotemporally continuous leakage risk distribution characteristics; the fluid dynamics simulation uses the leakage source term tensor as the boundary condition input; the fluid dynamics equations consider the balance relationship between fluid viscous force, pressure gradient and leakage driving force; the Navier-Stokes equations are solved discretically using the finite volume method; the velocity component of each grid cell is calculated; and the original flow field data containing three-dimensional velocity vectors is generated. The original flow field data completely records the fluid motion state within the simulation area.

[0069] S2.3: The original flow field data is fused with real-time monitoring data by using a quantum compressed sensing operator to generate corrected flow field data.

[0070] The specific process includes: the original flow field data contains the spatiotemporal distribution of the fluid velocity field vector output by the neuromorphic computing chip; the quantum compressed sensing operator uses a quantum random measurement matrix to perform dimensionality reduction sampling on the original flow field data; at the same time, it receives measured data such as water level and flow velocity collected by real-time monitoring stations; the simulation data and measured data are fused in Hilbert space through the principle of quantum superposition; the quantum compressed sensing operator optimizes the data fusion process based on the criterion of minimizing reconstruction error, eliminates the deviation between the original flow field data and the measured data, and finally generates calibrated flow field data that retains both simulation accuracy and measured accuracy. The calibrated flow field data has higher spatiotemporal consistency and physical authenticity.

[0071] The simulation data comes from the raw flow field data generated by using the risk dataset as boundary conditions through fluid dynamics simulation calculations.

[0072] S2.4: Process the corrected flow field data using a fractal geometric feature extraction algorithm to generate a flood evolution map.

[0073] The specific process includes: correcting the flow field data to include the spatiotemporal distribution information of the fluid velocity field vector; using a fractal geometric feature extraction algorithm to obtain the box dimension and Hausdorff dimension of the vortex structure in the corrected flow field data; quantifying the irregularity of the flow field through multifractal spectrum analysis; spatially mapping the fractal feature parameters with the velocity gradient field; and generating a flood evolution map that simultaneously includes the predicted inundation range and the predicted inundation depth.

[0074] S3: Based on the flood evolution map, the gate topology is optimized through graph convolutional networks to generate a gate opening control instruction set.

[0075] S3.1: Associate the fractal features in the flood evolution map with the risk dataset to the gate nodes to generate a modulated set of node attributes.

[0076] The specific process includes: fractal features in the flood evolution map reflect the nonlinear dynamic characteristics of the flood diffusion process; the risk dataset contains the surface deformation rate, coherence coefficient, leakage risk amount, and leakage risk probability associated with geographic coordinates; the fractal features and risk dataset are associated with the control range of each gate node through a spatial matching algorithm; and the fractal features, predicted inundation depth, and leakage risk probability are integrated into a multidimensional feature vector using a feature weighted fusion method. Each gate node corresponds to a feature vector set that integrates flood evolution characteristics and geological risk characteristics. After normalization and spatiotemporal alignment, a modulated node attribute set is formed. The node attribute set fully represents the comprehensive risk status and flood evolution trend within the control area of ​​each gate.

[0077] S3.2: Input the node attribute set into the quantum-enhanced graph convolutional network, output the node state set, and optimize it to generate the gate connection weight matrix.

[0078] The specific process includes: a node attribute set containing flood evolution characteristics and geological risk parameters within the gate control area; a quantum-enhanced graph convolutional network encoding the node attribute set into quantum states; a controlled rotating gate operation in the quantum circuit to achieve feature transformation; graph convolution operation in Hilbert space to capture the topological relationships between gate nodes; quantum parallel recognition to accelerate the aggregation and updating process of node features; outputting a node state set containing control decision suggestions for each gate; dynamic rules based on the node state set to derive the collaborative control requirements between gates; a nonlinear mapping function to transform node state differences into connection strength; generating an optimized gate connection weight matrix reflecting the optimal collaborative relationship of the gate group; and the element values ​​of the optimized gate connection weight matrix characterizing the correlation of water flow regulation between different gates.

[0079] S3.3: The gate connection weight matrix is ​​fused with the real-time flow field vorticity characteristics to generate a gate opening control command set.

[0080] The specific process includes: a gate connection weight matrix containing optimized inter-gate collaborative control relationships; real-time flow field vorticity characteristics reflecting the current water body rotation state; coupling the gate connection weight matrix and real-time flow field vorticity characteristics in the feature space using a tensor fusion algorithm; balancing gate collaboration requirements and flow field dynamic characteristics using an adaptive weighting method; inputting the fused feature tensor into the control decision engine; deriving the ideal opening value of each gate based on fuzzy logic rules; verifying the compliance of instructions considering gate mechanical constraints; and finally generating a gate opening control instruction set containing timestamps and spatial location information. This gate opening control instruction set accurately matches the current flood evolution state and meets the overall scheduling requirements of the gate group.

[0081] The real-time flow field vorticity characteristics are obtained by performing curl calculations on the velocity field after obtaining the fluid velocity field vector from the fluid dynamics equations, resulting in the spatial rotation intensity distribution.

[0082] S4: Inject the gate opening control command set into the digital twin for simulation correction, obtain the corrected flood depth, and calculate the flood deviation value in combination with the predicted flood depth to generate a verification report.

[0083] S4.1: Encode the gate opening control instruction set into a qubit rotation angle, and modulate the quantum phase according to chaotic dynamics to generate the initial quantum state.

[0084] The specific process includes: a gate opening control command set containing the control parameters of each gate; mapping the values ​​of the gate opening control command set to the rotation angle of the qubit through a quantum encoder; each gate control command corresponding to a specific deflection of a qubit; using the Lorentz attractor equation to generate a nonlinear phase modulation signal in chaotic dynamics; coupling the nonlinear phase modulation signal with the qubit rotation angle in Hilbert space; and achieving precise control of the quantum state by chaotic perturbation through quantum controlled phase gate operation, ultimately generating an initial quantum state with chaotic characteristics.

[0085] S4.2: Input the initial quantum state into the digital twin and perform time evolution to obtain the final state quantum wave function.

[0086] The specific process includes: the initial quantum state carries the encoded information of the gate opening control instruction set and chaotic modulation characteristics; the quantum fluid Hamiltonian operator of the digital twin contains the kinetic energy term, potential energy term, and quantum pressure term of the fluid motion; the time evolution process of the quantum state is described by the Schrödinger equation; the quantum fluid Hamiltonian operator acts on the initial quantum state and performs unitary transformation within the discrete time step to simulate the quantum dynamics behavior of the fluid under gate control; the quantum coherence and entanglement characteristics are maintained during the evolution process; and the final state quantum wave function is obtained after a preset time interval.

[0087] The preset time interval is determined by the characteristic time scale of fluid motion and the quantum decoherence time. The time step is dynamically adjusted through the quantum Zeno effect to ensure evolution accuracy.

[0088] S4.3: Perform compressed sensing measurement on the final state quantum wave function and decode to generate the original submerged water depth.

[0089] The specific process includes: the final state quantum wave function contains quantum state information of fluid evolution; compressed sensing measurement uses a random projection matrix to perform dimensionality reduction sampling on the final state quantum wave function; the spatial distribution characteristics of the final state quantum wave function are reconstructed through quantum tomography; the decoding process uses a sparse reconstruction algorithm to map the quantum measurement results to the water depth distribution in physical space; and combined with the potential energy term transformation relationship in the quantum fluid Hamiltonian operator, the original flooding depth corresponding to the quantum evolution result is finally generated. This original flooding depth retains the non-classical characteristics of quantum simulation.

[0090] S4.4: Calculate the error function differentiation and chaotic parameters for the original submerged depth to generate the corrected submerged depth, expressed as:

[0091]

[0092] Where h1(x,t) represents the corrected flood depth at spatial location x and time t, h0(x,t) represents the original flood depth at spatial location x and time t, η(x,t) represents the chaotic-topography dynamic coupling coefficient at spatial location x and time t, t represents the current time, t0 represents the simulation start time, τ represents the integration time variable, K(x,τ) represents the spatiotemporal decay kernel function at spatial location x and time τ, and erf represents the Gaussian error function. This represents the original submerged water depth gradient vector at spatial location x and time τ.

[0093] Where σ(x,τ) represents the water depth gradient normalization factor at spatial location x and time τ, and its expression is:

[0094]

[0095] It should be noted that N represents the number of sampling points in the neighborhood space, and i represents the index of the neighborhood space point. Indicates the spatial location x i The Euclidean norm of the original water depth gradient at time τ, where a represents the spatial characteristic frequency, and x i The coordinates of the i-th neighborhood point represent the spatial location.

[0096] The specific process includes: the original flooding depth reflects the preliminary results obtained from quantum fluid simulation; the digital twin's correction engine calculates the error function between the original flooding depth and the measured benchmark data; the spatial gradient distribution of the error function is solved using the variational method; at the same time, spatiotemporally varying chaotic parameters are generated based on the chaotic dynamics equations; the error gradient and chaotic parameters are dynamically coupled in the integral domain; historical errors are weighted using a spatiotemporal decay kernel function; the correction amplitude is constrained by a Gaussian error function; and finally, a corrected flooding depth that integrates quantum simulation characteristics and chaotic correction characteristics is generated. The corrected flooding depth satisfies both fluid dynamics laws and actual observation constraints.

[0097] The measured baseline data comes from actual measurement data such as water level and flow velocity collected by real-time monitoring stations, as well as on-site measured data such as surface seepage flow and crack distribution obtained by UAV multispectral sensors and synthetic aperture radar.

[0098] S4.5: Based on the corrected flood depth and the predicted flood depth, the flood deviation value is calculated using quantum state fidelity, expressed as follows:

[0099]

[0100] Where ΔE(x,t) represents the inundation deviation at spatial location x and time t, and Ω represents the spatial integration region for flood evolution simulation calculation. Indicates the spatial location x ′ And the quantum state conjugate wave function of the submerged water depth field h1 after correction at time t, Indicates the spatial location x ′ The quantum state wave function that predicts the submerged water depth at time t.

[0101] Where G(x,x) ′ ) represents spatial location x and x ′ The quantum correlation strength between them is expressed as:

[0102]

[0103] It should be noted that exp represents the natural exponential function, γ represents the spatial correlation length, and m represents the characteristic wavenumber of the quantum correlation operator. This represents the initial terrain gradient vector at spatial location x. Indicates the spatial location x ′ The initial terrain gradient vector at that location.

[0104] The specific process includes: correcting the inundation depth by including the simulation results after chaotic correction; predicting the inundation depth by using the extrapolation data from the flood evolution map; calculating the quantum state fidelity by encoding the corrected and predicted inundation depths as quantum state wave functions; comparing the similarity of the quantum state wave functions through quantum state inner product operations; calculating the quantum correlation strength in the Hilbert space integral domain; the correlation strength reflecting the topographic gradient relationship between spatial locations; and finally outputting the inundation deviation value characterizing the degree of difference in water depth.

[0105] S4.6: Correlate the flooding deviation value according to spatial coordinates to generate a spatiotemporal deviation field, and map it to the chaotic attractor phase space. Integrate along the trajectory to generate a risk value.

[0106] The specific process includes: the inundation deviation value contains the degree of water depth difference corresponding to the spatial location; a spatiotemporal deviation field is established by matching geographic coordinates; the phase space of the chaotic attractor uses a three-dimensional coordinate system to represent the deviation evolution characteristics; each data point in the spatiotemporal deviation field is mapped to a specific location in the phase space according to the Lorentz equation; time integration is performed along the chaotic trajectory; the time integration process considers the spatial autocorrelation and temporal persistence of the inundation deviation value; and finally outputs a risk value that comprehensively reflects the deviation intensity and evolution trend.

[0107] S4.7: Perform fractional integral and spatial convolution on the risk value to generate a cumulative risk field.

[0108] The specific process includes: the risk value reflects the integral result of the spatiotemporal deviation field in the chaotic phase space; the fractional integral uses the Riemann-Liouville definition to handle the time memory effect of the risk value; the non-integer order differential operation is achieved through the gamma function; the spatial convolution uses the Gaussian kernel function to weight and smooth the neighborhood risk value; the fractional integral retains the long-range correlation characteristics of historical risks; and the spatial convolution enhances the spatial continuity of the risk field. The two operations work together in the time and spatial domains to finally generate a cumulative risk field that has both time accumulation characteristics and spatial distribution patterns. The cumulative risk field quantifies the spatiotemporal aggregation degree of risk factors during the flood evolution process.

[0109] S4.8: Encode the cumulative risk field into a quantum hologram, combine it with chaotic timestamps and dynamic risk classification, and generate a verification report.

[0110] The specific process includes: the cumulative risk field contains the spatiotemporal distribution characteristics of flood risk; the quantum hologram encoding process uses the principle of quantum interference to convert the risk field into phase and amplitude information; the chaotic timestamp generates an unforgeable time identifier through the Lorentz equation; the dynamic risk classification automatically divides the risk level based on the risk threshold; the quantum hologram records the complete three-dimensional information of the risk field; the chaotic timestamp ensures the authenticity of the report's time sequence; the dynamic risk classification provides intuitive risk assessment results; and the three are integrated to generate a verification report with quantum anti-counterfeiting characteristics and chaotic time sequence characteristics. The verification report fully retains the quantum holographic projection of the risk field and the classification assessment conclusions.

[0111] Risk thresholds are classified into tiered critical values ​​based on the numerical distribution characteristics of the cumulative risk field and statistical analysis of historical disaster data.

[0112] S5: Control the gate according to the verification report, collect multi-source leakage data simultaneously, and feed it back to the geological leakage model for optimization.

[0113] S5.1: Based on the verification report, perform gate control and simultaneously collect multi-source leakage data.

[0114] The specific process includes: the verification report contains the cumulative risk field information and dynamic risk classification results recorded by the quantum hologram; the gate control command is automatically triggered and executed according to the risk assessment level in the verification report; the UAV plans its trajectory according to the coordinates of the high-risk area marked in the verification report; it is equipped with multispectral sensors and synthetic aperture radar to collect data on surface seepage flow, soil moisture content and crack distribution; the gate control command execution is started synchronously with the UAV data acquisition; and the seepage data collected by the UAV and the gate control effect are matched in time and space.

[0115] S5.2: The seepage flow rate, crack distribution data and soil data are fused into a spatiotemporal seepage tensor, and the geological seepage model is dynamically optimized through chaotic integral operators and quantum verification.

[0116] The specific process includes: the seepage flow rate collected by the UAV reflects the actual seepage intensity; the crack distribution data characterizes the degree of surface fracturing; and the soil data indicates the development status of the seepage path. The three types of data are spatiotemporally registered to form a multidimensional seepage tensor. The chaotic integral operator obtains the nonlinear characteristics of the seepage field based on fractal theory. Quantum verification uses quantum state tomography to check the reliability of the data. The deformation sensitivity factor and attenuation coefficient of the geological seepage model are dynamically optimized based on the seepage tensor. The optimized geological seepage model can more accurately reflect the coupling relationship between the actual seepage process and surface deformation, realizing the online optimization of the geological seepage model.

[0117] This embodiment also provides a flood control scheduling system based on digital twins, including: a data acquisition module, a flood simulation module, a gate control module, a simulation verification module, and a feedback calibration module. The data acquisition module is used to acquire time-series satellite interferometry data, obtain the surface deformation rate and coherence coefficient, and calculate the leakage risk through a geological leakage model to generate a risk dataset. The flood simulation module is used to convert the risk dataset into boundary conditions for fluid dynamics simulation and perform flood evolution simulation to generate a flood evolution map. The flood evolution map includes the predicted inundation range and the predicted inundation depth. The gate control module is used to optimize the gate topology relationship through a graph convolutional network based on the flood evolution map to generate a gate opening control instruction set. The simulation verification module is used to inject the gate opening control instruction set into the digital twin for simulation correction, obtain the corrected inundation depth, and calculate the inundation deviation value in combination with the predicted inundation depth to generate a verification report. The feedback calibration module is used to control the gates according to the verification report, simultaneously acquire multi-source leakage data, and feed it back to the geological leakage model for optimization.

[0118] This embodiment also provides a computer device applicable to the flood control scheduling method based on digital twins, 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 flood control scheduling method based on digital twins as proposed in the above embodiment.

[0119] 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.

[0120] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the flood control scheduling method based on digital twins 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.

[0121] In summary, this invention achieves quantitative modeling of underground seepage processes by collecting time-series satellite interferometry data and constructing a structured risk dataset, thereby improving the accuracy of boundary conditions in flood evolution simulation and enhancing the reliability of predictions. Furthermore, by optimizing gate topology relationships based on graph convolutional networks and generating control instruction sets, it realizes the coordinated optimization of dynamic flow fields and network topology, improving the real-time performance and coordination of basin-level flood control scheduling.

[0122] 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 flood control scheduling method based on digital twinning, characterized in that: The application relates to a method for optimizing flood control of a dam by using a quantum graph neural network. The method comprises the following steps: collecting time-series satellite interferometric measurement data, obtaining surface deformation rate and coherence coefficient, and calculating leakage risk quantity by a geological leakage model to generate a risk data set; converting the risk data set into boundary conditions of fluid dynamics simulation, and performing flood evolution simulation to generate a flood evolution atlas; the flood evolution atlas comprises a predicted inundation range and a predicted inundation water depth; optimizing a gate topology relationship by a graph convolution network based on the flood evolution atlas to generate a gate opening control instruction set; injecting the gate opening control instruction set into a digital twin for simulation correction to obtain corrected inundation water depth, and calculating an inundation deviation value by combining the predicted inundation water depth to generate a verification report; 2. The flood control scheduling method based on digital twinning of claim 1, wherein: controlling the gate according to the verification report, synchronously collecting multi-source leakage data, and feeding back to the geological leakage model for optimization. The specific steps of generating the risk data set are as follows: based on time-series satellite interferometric measurement data, generating an interferometric phase image by phase unwrapping, and identifying a permanent scatterer to obtain surface deformation rate and coherence coefficient; inputting the surface deformation rate into the geological leakage model to calculate the leakage risk quantity, and calculating the leakage risk probability in combination with the coherence coefficient; 3. The flood control scheduling method based on digital twinning of claim 1, wherein: combining the surface deformation rate, the coherence coefficient, the leakage risk quantity and the leakage risk probability according to geographical coordinates to generate the risk data set. The specific steps of generating the flood evolution atlas are as follows: inputting the risk data set into a quantum graph neural network to output a time-space continuous leakage source term tensor; taking the leakage source term tensor as boundary conditions of fluid dynamics simulation, calculating a fluid velocity field vector by a fluid dynamics equation to generate original flow field data; fusing the original flow field data with real-time monitoring data by a quantum compressive sensing operator to generate corrected flow field data; 4. The flood control scheduling method based on digital twinning of claim 1, wherein: processing the corrected flow field data by a fractal geometric feature extraction algorithm to generate the flood evolution atlas. The specific steps of generating the gate opening control instruction set are as follows: associating the fractal features in the flood evolution atlas with the risk data set to the gate node to generate a modulated node attribute set; inputting the node attribute set into a quantum-enhanced graph convolution network to output a node state set, and optimizing to generate a gate connection weight matrix; 5. The flood control scheduling method based on digital twinning of claim 1, wherein: fusing the gate connection weight matrix with real-time flow field vorticity features to generate the gate opening control instruction set. The specific steps of injecting the gate opening control instruction set into the digital twin for simulation correction to obtain the corrected inundation water depth are as follows: encoding the gate opening control instruction set into a quantum bit rotation angle, and generating an initial quantum state according to chaotic dynamics modulation quantum phase; inputting the initial quantum state into the digital twin to perform time evolution to obtain a final quantum wave function; performing compressive sensing measurement on the final quantum wave function to generate original inundation water depth; 6. The flood control scheduling method based on digital twinning of claim 1, wherein: performing error function differentiation and chaotic parameter calculation on the original inundation water depth to generate corrected inundation water depth. The specific steps of calculating the inundation deviation value by combining the predicted inundation water depth to generate the verification report are as follows: calculating the inundation deviation value by quantum state fidelity based on the corrected inundation water depth and the predicted inundation water depth; associating the inundation deviation value according to spatial coordinates to generate a time-space deviation field, and mapping to a chaotic attractor phase space to generate a risk value along a trajectory integral. Score integral and spatial convolution are performed on the risk value to generate a cumulative risk field; The cumulative risk field is encoded into a quantum hologram, combined with a chaotic timestamp and a dynamic risk classification to generate a verification report.

7. The flood control scheduling method based on digital twinning of claim 1, wherein: The gate control is performed according to the verification report, the multi-source leakage data is synchronously collected, and is fed back to the geological leakage model for optimization, and the specific steps are as follows, The gate control is performed according to the verification report, and the multi-source leakage data is collected; The multi-source leakage data includes seepage flow, fracture distribution data and soil data; The seepage flow, fracture distribution data and soil data are fused into a spatiotemporal leakage tensor, and the geological leakage model is dynamically optimized through chaotic integral operators and quantum verification.

8. A flood control scheduling system based on digital twinning, based on the flood control scheduling method based on digital twinning in any of claims 1-7, characterized in that: The method comprises a data collection module, a flood simulation module, a gate control module, a simulation verification module and a feedback calibration module, The data collection module is used to collect time-series satellite interferometric measurement data, obtain surface deformation rate and coherence coefficient, and calculate leakage risk quantity through a geological leakage model to generate a risk data set; The flood simulation module is used to convert the risk data set into boundary conditions of fluid dynamics simulation, and perform flood evolution simulation to generate a flood evolution atlas; the flood evolution atlas comprises a predicted inundation range and a predicted inundation water depth; The gate control module is used to optimize gate topology relationship through a graph convolution network based on the flood evolution atlas to generate a gate opening degree control instruction set; The simulation verification module is used to inject the gate opening degree control instruction set into a digital twin for simulation correction, obtain a corrected inundation water depth, and calculate an inundation deviation value in combination with the predicted inundation water depth to generate a verification report; The feedback calibration module is used to perform gate control according to the verification report, synchronously collect multi-source leakage data, and feed back to the geological leakage model for optimization. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to realize the steps of the flood control scheduling method based on digital twinning according to any one of claims 1-7.

10. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the flood control scheduling method based on digital twinning according to any one of claims 1-7.

Images