Data assimilation method suitable for real-time flood simulation of SWMM model
The ensemble Kalman filtering method simultaneously assimilates the node water level and flow rate of the SWMM model, which solves the problem of mismatch between water level and flow rate, improves the simulation accuracy and model adaptability, and meets the emergency response needs.
Patent Information
- Application Number
- CN202510568528.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-30
- Publication Date
- 2025-08-01
AI Technical Summary
In the real-time flood simulation of existing SWMM models, only assimilating the node water level causes the node water level to mismatch the flow rate and flow of the connected pipeline, introducing calculation errors, and the error accumulates over time, affecting the simulation accuracy.
The ensemble Kalman filtering method is used, combining the measured node water level and the node water level and flow output from the SWMM model, and the water level and flow are assimilated simultaneously through the data assimilation method, and the water dynamic equation is used to correct the flow to ensure the consistency between the water level and the flow.
It significantly improves the accuracy of flood simulation, reduces error accumulation, enhances model robustness, adapts to complex urban drainage scenarios, and meets the timeliness of emergency response.
Smart Images
Figure CN120409349A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of flood control, and particularly relates to a data assimilation method applicable to real-time flood simulation of the SWMM model. Background Art
[0002] Real-time flood simulation based on numerical models is an important basic support for understanding the development trend of flood disasters and making scientific emergency decisions in a timely manner. However, errors are inevitably introduced during the real-time rolling operation of numerical models. As time goes by, the errors will continue to accumulate and become larger and larger, eventually leading to untrustworthy simulation results. To solve such problems, data assimilation of the state variables simulated by the model based on the observed state variables has become a common method in the academic and industrial circles. For the SWMM model, since only the modification of the node water level is included in its data interface file and the actual observed values are mostly based on water level observations, current data assimilation for the SWMM model mostly targets the node water level. This results in a mismatch between the node water level and the state variables such as the flow velocity and flow rate of the pipes connected to it at the beginning of the simulation calculation, introducing new calculation errors. Therefore, a data assimilation method applicable to real-time flood simulation of the SWMM model needs to be designed to solve the above problems. Summary of the Invention
[0003] The technical problem to be solved by the present invention is to provide a data assimilation method applicable to real-time flood simulation of the SWMM model, aiming to solve the problem that the SWMM model in the prior art conducts data assimilation for the node water level, resulting in a mismatch between the node water level and the state variables such as the flow velocity and flow rate of the pipes connected to it, thus introducing new calculation errors. It has the characteristics of being able to assimilate the corresponding river and pipe flows while assimilating the node water level, reducing cumulative errors, and improving the real-time simulation accuracy of the model.
[0004] To achieve the above technical effects, the technical solution adopted by the present invention is as follows: A data assimilation method applicable to real-time flood simulation of the SWMM model, comprising: S1, setting model variables, including the measured node water level and the node water level and flow rate output by the SWMM real-time model; S2, based on the ensemble Kalman filter method, combining the set model variables to perform data assimilation on the output variables of the SWMM real-time model; S3, updating the flow rate based on the assimilated analysis water level.
[0005] Preferably, in step S1, the node water level and flow rate output by the SWMM real-time model include: the node water level output by the SWMM real-time model at time t and the flow rate , where the flow rate includes river flow or pipe flow; h = {h1, h2,..., hNs}, q = {q1, q2, …, q Ns} is a one-dimensional vector representing the water levels and flows of all output nodes and pipe segments. N is the number of nodes / river reaches output by the model, and the output nodes correspond one-to-one with the output river channels / pipes.
[0006] Preferably, in step S1, the measured node water level is , H = {H1, H2, …, H No} is a one-dimensional vector representing the measured node water levels. No is the number of observation points, and No < Ns is satisfied.
[0007] Preferably, in step S2, the ensemble Kalman filter method is used to assimilate the measured water levels and the simulated water levels through data assimilation, which specifically includes the following steps: S201. Generate a set of model state variables. The number of set members is M, and each member is represented as: ; where, represents the predicted state vector of the i-th set member at time t, including the simulated water level and the flow ; and are respectively the simulated water level and flow of the i-th set member at time t; S202. Map the set members to the observation space through the observation operator H to obtain the predicted observed values: ; where, H represents the observation operator matrix, with a dimension of No × Ns , and is used to extract the water level values corresponding to the measured nodes from all model nodes; is the predicted observed water level vector of the i-th member at time t; S203. Calculate the ensemble mean and covariance: Predicted water level mean: ; Predicted covariance matrix: ; where, represents the average predicted water level vector of all set members at time t; represents the predicted error covariance matrix, with a dimension of Ns × Ns , which characterizes the uncertainty of the model prediction; S204. Calculate the Kalman gain matrix: ; wherein, R is the observation error covariance matrix; the dimension is No×No, representing the uncertainty of the measured water level data; is the Kalman gain matrix, with the dimension of Ns×No, used for weighted adjustment of the contributions of model prediction and measured data; S205, update the ensemble members to obtain the analyzed water level: ; wherein, is a random perturbation term subject to the N(0,R) distribution; is the measured water level vector at time t; is the analyzed water level vector of the i-th member at time t S206, take the mean of the analyzed water levels of the ensemble members to obtain : .
[0008] wherein, represents the final analyzed water level.
[0009] Preferably, in step S3, based on the assimilated analyzed water level update the river / pipe flow rate ; specifically including: According to the difference between the assimilated water level and the water level before assimilation , combined with the inertial term of the hydrodynamic equation, calculate the flow rate correction amount through the following formula:
[0010] wherein, U is the average flow velocity of the river / pipe simulated by the model, Aa is the cross-sectional area calculated according to , and A is the cross-sectional area calculated according to .
[0011] Preferably, in step S3, superimpose the flow rate correction amount ΔQ and the flow rate simulated by the model to obtain the assimilated river / pipe flow rate: ; and perform assimilation calculations on the simulated data at subsequent times in sequence.
[0012] Preferably, after the flow rate correction, the following steps are further included: Verify the rationality by solving the simplified Saint-Venant equation. If the residual is greater than the preset value, readjust the covariance weight of the ensemble Kalman filter or increase the number of ensemble members M; the calculation formula is: ; Among them, δ is a preset threshold value.
[0013] Preferably, a data assimilation device applicable to real-time flood simulation of the SWMM model, the device is used for the data assimilation method applicable to real-time flood simulation of the SWMM model; the device includes: Data input module: used to receive measured node water level data , as well as the node water level and flow rate output by the SWMM real-time model; preprocess the input data, including outlier removal, missing data interpolation, and multi-source data fusion; Ensemble generation and assimilation module: It includes an ensemble generation sub-module and a Kalman filter calculation sub-module. The ensemble generation sub-module is used to generate an ensemble of model state variables ; the Kalman filter calculation sub-module is used for the data assimilation process and outputs the analyzed water level ; Flow correction module: According to the analyzed water level , through the formula calculate the flow correction amount, and superimpose it with the simulated flow rate to generate the assimilated flow rate ; configured to iteratively verify the compatibility with the hydrodynamic equation, and if the residual exceeds the limit, trigger an error feedback to the ensemble generation and assimilation module; Dynamic parameter optimization module: Take the Manning coefficient n of the SWMM model as an extended state variable, and synchronously update the parameter value through the ensemble Kalman filter; Result output module: Based on the assimilated and , generate a spatial distribution map of the inundation depth, and push it to the urban emergency management platform through the API interface, and output early warning signals and emergency decision-making suggestions in real time.
[0014] Preferably, a computer device includes a memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to execute the data assimilation method applicable to real-time flood simulation of the SWMM model.
[0015] Preferably, a computer-readable storage medium stores computer instructions thereon, and the computer instructions are used to cause a computer to execute the data assimilation method applicable to real-time flood simulation of the SWMM model.
[0016] The beneficial effects of the present invention are as follows: 1. Significantly improve the accuracy of flood simulation and reduce the risk of error accumulation: This patent solves the problem of initial state mismatch caused by only assimilating water levels in traditional methods by simultaneously assimilating the node water levels and channel / pipeline flow data of the SWMM model. In traditional methods, only adjusting the water level will cause the hydrodynamic parameters such as the associated flow rate and velocity to deviate from the real physical process, and the simulation results will gradually become distorted as errors accumulate over time. This patent uses the Ensemble Kalman Filter to fuse the measured water levels with the model output and synchronously corrects the flow rate based on the inertial term in the hydrodynamic equation to ensure the physical consistency of water levels and flow rates in the spatio-temporal dimension.
[0017] 2. Enhance the robustness of the model and adapt to complex urban drainage scenarios: Traditional data assimilation methods are highly dependent on the quality of observational data and are prone to failure when the observation points are sparse or there are outliers. This patent significantly improves the integrity and reliability of the input data by introducing a multi-source data fusion mechanism and a data preprocessing module to perform spatio-temporal alignment, outlier removal, and interpolation on the measured data.
[0018] 3. Achieve efficient real-time calculation and meet the timeliness requirements of emergency response. The calculation efficiency is significantly improved by optimizing the process design of the Ensemble Kalman Filter. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 is a flowchart of the present invention; Figure 2 is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION
[0020] Embodiment 1: As Figure 1 shown, a data assimilation method for real-time flood simulation applicable to the SWMM model includes: S1. Set model variables, including the measured node water levels and the node water levels and flow rates output by the SWMM real-time model; S2. Based on the Ensemble Kalman Filter method, perform data assimilation on the variables output by the SWMM real-time model in combination with the set model variables; S3. Update the flow rate based on the assimilated analysis water level.
[0021] Preferably, in step S1, the node water levels and flow rates output by the SWMM real-time model include: The node water level output by the SWMM real-time model at time t and the flow rate , where the flow rate includes channel flow rate or pipeline flow rate; h = {h1, h2,..., h Ns}, q = {q1, q2,..., q Ns} is a one-dimensional vector representing the water levels and flows of all output nodes and pipe segments. N is the number of nodes / river reaches output by the model, and the output nodes correspond one-to-one with the output river channels / pipes.
[0022] Preferably, in step S1, the measured node water level is , H = {H1, H2, …, H No} is a one-dimensional vector representing the measured node water levels. No is the number of observation points, and No < Ns is satisfied.
[0023] Preferably, in step S2, the ensemble Kalman filter method is used to integrate the measured water level and the simulated water level for data assimilation, which specifically includes the following steps: S201, generate a set of model state variables. The number of set members is M, and each member is expressed as: ; where, represents the predicted state vector of the i-th set member at time t, including the simulated water level and the flow ; and are respectively the simulated water level and flow of the i-th set member at time t; S202, map the set members to the observation space through the observation operator H to obtain the predicted observation values: ; where, H represents the observation operator matrix, with dimensions of No × Ns , and is used to extract the water level values corresponding to the measured nodes from all model nodes; is the predicted observation water level vector of the i-th member at time t; S203, calculate the ensemble mean and covariance: Predicted water level mean: ; Predicted covariance matrix: ; where, represents the average predicted water level vector of all set members at time t; represents the predicted error covariance matrix, with dimensions of Ns × Ns , characterizing the uncertainty of the model prediction; S204, calculate the Kalman gain matrix: ; where \(R\) is the observation error covariance matrix, with a dimension of \(N_o\times N_o\), representing the uncertainty of the measured water level data; is the Kalman gain matrix, with a dimension of \(N_s\times N_o\), used for weighted adjustment of the contributions of model predictions and measured data; S205. Update the ensemble members to obtain the analyzed water level: ; where is a random perturbation term obeying the \(N(0, R)\) distribution; is the measured water level vector at time \(t\); is the analyzed water level vector of the \(i\)-th member at time \(t\) S206. Take the mean of the analyzed water levels of the ensemble members to obtain : .
[0024] where represents the final analyzed water level.
[0025] Preferably, in step S3, based on the assimilated analyzed water level update the river / pipe flow rate ; specifically including: According to the difference between the assimilated water level and the water level before assimilation , combined with the inertial term of the hydrodynamic equation, calculate the flow rate correction amount through the following formula:
[0026] where \(U\) is the average velocity of the river / pipe simulated by the model, \(A_a\) is the cross-sectional area calculated according to , and \(A\) is the cross-sectional area calculated according to .
[0027] Preferably, in step S3, superimpose the flow rate correction amount \(\Delta Q\) and the flow rate simulated by the model to obtain the flow rate of the river / pipe section after assimilation: ; and perform assimilation calculations on the simulation data at subsequent times in sequence.
[0028] Preferably, after the flow rate correction, the following steps are further included: Verify the rationality of by solving the simplified Saint-Venant equation. If the residual is greater than the preset value, readjust the covariance weight of the ensemble Kalman filter or increase the number of ensemble members \(M\); the calculation formula is: ; where \(\delta\) is the preset threshold.
[0029] Preferably, a data assimilation device applicable to real-time flood simulation of the SWMM model, the device is used for the data assimilation method applicable to real-time flood simulation of the SWMM model; the device includes: Data input module: used to receive measured node water level data , as well as the node water level and flow output by the SWMM real-time model; preprocess the input data, including outlier removal, missing data imputation, and multi-source data fusion; Ensemble generation and assimilation module: Includes an ensemble generation sub-module and a Kalman filter calculation sub-module. The ensemble generation sub-module is used to generate an ensemble of model state variables ; the Kalman filter calculation sub-module is used for the data assimilation process and outputs the analyzed water level ; Flow correction module: According to the analyzed water level , through the formula Calculate the flow correction amount, and superimpose it with the simulated flow To generate the assimilated flow ; configured to iteratively verify The compatibility with the hydrodynamic equation, and if the residual exceeds the limit, trigger an error feedback to the ensemble generation and assimilation module; Dynamic parameter optimization module: Take the Manning coefficient n of the SWMM model as an extended state variable, and synchronously update the parameter values through ensemble Kalman filtering; Result output module: Based on the assimilated and , generate a spatial distribution map of the inundation depth, and push it to the urban emergency management platform through the API interface, and real-time output warning signals and emergency decision-making suggestions.
[0030] Embodiment 2: A computer device includes a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the data assimilation method applicable to real-time flood simulation of the SWMM model.
[0031] Such as Figure 2As shown, it is a schematic structural diagram of a computer device provided by an optional embodiment of the present invention. The computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device such as a display device coupled to the interface. In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations, for example, as a server array, a set of blade servers, or a multi-processor system. Figure 2 Taking one processor 10 as an example.
[0032] The processor 10 can be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 can further include a hardware chip. The above hardware chip can be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above programmable logic device can be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.
[0033] [[ID=~8]]The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.
[0034] The memory 20 can include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the computer device, etc. In addition, the memory 20 can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some optional embodiments, the memory 20 can optionally include a memory remotely set relative to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.
[0035] The memory 20 can include a volatile memory, such as a random access memory; the memory can also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 can also include a combination of the above types of memories.
[0036] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or communication networks.
[0037] Embodiment 3: Preferably, a computer-readable storage medium stores computer instructions thereon, and the computer instructions are used to cause a computer to execute the data assimilation method for real-time flood simulation applicable to the SWMM model.
[0038] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code that is originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiment is implemented.
Claims
1. A data assimilation method for real-time flood simulation applicable to the SWMM model, characterized in that, Including: S1, set model variables, including the measured node water levels and the node water levels and flows output by the SWMM real-time model; S2, based on the Ensemble Kalman Filter method, assimilate the output variables of the SWMM real-time model by combining the set model variables; S3, update the flow based on the assimilated analysis water levels.
2. The data assimilation method applicable to real-time flood simulation of SWMM model according to claim 1, characterized in that In the step S1, the node water levels and flows output by the SWMM real-time model include: The node water levels and flows output by the SWMM real-time model at time t and flows , where the flows include channel flows or pipe flows; h = {h1, h2, …, h Ns}, q = {q1, q2, …, q Ns} are one-dimensional vectors representing the water levels and flows of all output nodes and pipe segments, N is the number of nodes / river reaches output by the model, and the output nodes correspond one-to-one with the output channels / pipes.
3. A data assimilation method for real-time flood simulation applicable to the SWMM model according to claim 2, characterized in that, In the step S1, the measured node water level is , H = {H1, H2, …, H No} is a one-dimensional vector representing the measured node water level. No is the number of observation points, and No < Ns is satisfied.
4. A data assimilation method for real-time flood simulation applicable to the SWMM model according to claim 3, characterized in that In the step S2, the ensemble Kalman filtering method is adopted to comprehensively process the measured water level and the simulated water level for data assimilation, which specifically includes the following steps: S201, generate a set of model state variables, with the number of set members being M, and each member is represented as: ； Among them, represents the predicted state vector of the i-th ensemble member at time t, including the simulated water level and the flow rate ; and are the simulated water level and flow rate of the i-th ensemble member at time t, respectively; S202, map the set members to the observation space through the observation operator H to obtain predicted observation values: ； where \(H\) represents the observation operator matrix, with dimensions of No × Ns , which is used to extract the water level values corresponding to the measured nodes from the full model nodes; is the predicted observation water level vector of the \(i\)-th member at time \(t\); S203, calculate the set mean and covariance: Predicted water level mean: ； Predicted covariance matrix: ； Among them, represents the average predicted water level vector of all set members at time t; represents the prediction error covariance matrix, with a dimension of Ns × Ns , characterizing the uncertainty of model prediction; S204, calculate the Kalman gain matrix: ； where R is the observation error covariance matrix; the dimension is No×No, representing the uncertainty of the measured water level data; is the Kalman gain matrix, with the dimension of Ns×No, used for weighted adjustment of the contributions of model prediction and measured data; S205, update the set members to obtain the analysis water levels: ； wherein, is a random disturbance term subject to the N(0, R) distribution; is the measured water level vector at time t; is the analyzed water level vector of the i-th member at time t S206, the analysis water levels of the set members are averaged to obtain : ； Among them, represents the final analyzed water level.
5. A data assimilation method for real-time flood simulation applicable to the SWMM model according to claim 4, characterized in that, In the step S3, based on the assimilated analysis water level Update the river channel / pipeline flow ; Specifically including: Based on the assimilated water level and the water level before assimilation calculate the flow correction amount through the following formula by combining the inertial term of the hydrodynamic equation with the difference between them: where U is the average flow velocity of the river channel / pipeline simulated by the model, and Aa is the cross-sectional area of the water passage calculated according to and A is the cross-sectional area of the water passage calculated according to the cross-sectional area of the water passage.
6. A data assimilation method for real-time flood simulation applicable to the SWMM model according to claim 5, characterized in that, In the step S3, the flow correction amount ΔQ is superimposed on the model simulated flow to obtain the flow of the river reach / pipe segment after assimilation: ； And sequentially perform assimilation calculations on the simulation data at subsequent times.
7. A data assimilation method for real-time flood simulation applicable to the SWMM model according to claim 6, characterized in that After the flow is corrected, the following steps are further included: Verification by solving the simplified Saint-Venant equations for rationality. If the residual is greater than the preset value, readjust the covariance weight of the ensemble Kalman filter or increase the number of ensemble members M; the calculation formula is as follows: ； Wherein, δ is a preset threshold.
8. A data assimilation device applicable to real-time flood simulation of the SWMM model, characterized in that, The device is used to execute the data assimilation method for real-time flood simulation applicable to the SWMM model described in claims 1-7; the device includes: Data input module: used to receive the measured node water level data , as well as the node water levels and flows output by the SWMM real-time model; preprocess the input data, including outlier removal, missing data imputation, and multi-source data fusion; Set generation and assimilation module: It includes an ensemble generation sub-module and a Kalman filter calculation sub-module. The ensemble generation sub-module is used to generate an ensemble of model state variables ; The Kalman filter calculation sub-module is used for the data assimilation process and outputs the analyzed water level ; Flow correction module: According to the analyzed water level , calculate the flow correction amount through the formula , and superimpose it with the simulated flow to generate the assimilated flow ; configured for iterative verification of the compatibility with the hydrodynamic equation, and if the residual exceeds the limit, trigger error feedback to the ensemble generation and assimilation module; Dynamic parameter optimization module: Take the Manning coefficient n of the SWMM model as an extended state variable and synchronously update the parameter values through the Ensemble Kalman Filter; Result output module: Based on the assimilated and , generate a spatial distribution map of the inundation depth, and push it to the urban emergency management platform through the API interface, and output early warning signals and emergency decision-making suggestions in real time.
9. A computer device, characterized in that, Including a memory and a processor, the memory and the processor are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the data assimilation method for real-time flood simulation applicable to the SWMM model described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, Computer instructions are stored on a computer-readable storage medium, and the computer instructions are used to cause a computer to execute the data assimilation method for real-time flood simulation applicable to the SWMM model described in any one of claims 1 to 7.