Method, system and medium for generating dynamic calculation model of time difference method flow online monitoring

Abstract

The present application relates to a method, system and medium for generating a dynamic calculation model for online monitoring of flow using the time difference method. The method includes the following steps: online monitoring of flow rate comparison collection and data screening; establishing an ultrasonic time difference method derivation model and an objective function, and determining the constraints in the model calculation; based on the established objective function, using a particle swarm optimization algorithm to iteratively calculate the time difference method derivation model to deduce the optimal solution; performing deep learning based on actual hydrological monitoring production applications, adding new working conditions to the model for secondary optimization and solution; generating an ultrasonic time difference method calculation model based on the optimal solution. The present application can be used for real-time flow monitoring of hydrological sections affected by water conservancy projects such as rivers, reservoirs, and weirs, and can quickly and dynamically calibrate flow model parameters to meet the requirements of hydrological flow monitoring specifications, and can be widely used in hydrological monitoring, water resources management and other fields.

Description

Technical Field

[0001] The present application relates to the field of water conservancy and hydrological instrument measurement, and specifically to a method, system and medium for generating a dynamic calculation model for online flow monitoring using a time difference method. Background Art

[0002] Affected by human activities and climate change, extreme hydrological and meteorological disasters occur frequently, and the hydrological characteristics of the basin have changed significantly. At the same time, affected by water conservancy projects, the requirements for basin hydrological monitoring elements and monitoring accuracy are constantly increasing. At present, the ultrasonic time difference method is widely used in the monitoring of flow processes such as rivers, weirs, and reservoirs. How to ensure monitoring accuracy and expand the adaptability of instruments and equipment urgently needs to establish an effective flow calculation model to adapt to the changing river environment and better serve the increasingly stringent hydrological and water resources management. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a method, system and medium for generating a dynamic calculation model for online flow monitoring using an ultrasonic time difference method, which can be applied to online flow monitoring and calculation of rivers, reservoir inflows and outflows, weirs and canals, etc., to better and more quickly calibrate model parameters, and is suitable for non-stable water flows affected by multiple factors, and for measuring stations with complex flow state changes, thereby solving the technical problem of lag in model parameter calibration.

[0004] To achieve the above objectives, this application provides the following technical solutions:

[0005] In a first aspect, an embodiment of the present application provides a method for generating a dynamic calculation model for online flow monitoring using a time difference method, comprising the following specific steps:

[0006] Online monitoring flow comparison measurement collection and data screening, obtain basic water flow characteristic parameters of the hydrological station, and collect the flow results measured by ADCP and cableway current meter method and the online monitoring data of ultrasonic time difference method in the same period;

[0007] Establish the ultrasonic time difference method derivation model and objective function, and determine the constraints in the model calculation;

[0008] Based on the established objective function, the particle swarm optimization algorithm is used to iteratively calculate the time difference method derivation model to derive the optimal solution;

[0009] Based on deep learning in actual hydrological monitoring production applications, new working conditions are added to the model for secondary optimization solution;

[0010] Based on the optimal solution, an ultrasonic time difference method calculation model is generated.

[0011] The online monitoring flow rate comparison measurement collection and data screening obtain the basic water flow characteristic parameters of the hydrological station, and collect the measured flow results of the navigation ADCP and cableway current meter method and the online monitoring data of the ultrasonic time difference method in the same period, specifically,

[0012] Obtain the water-passing section and water level process of the hydrological station, obtain the hydrological test environment of the hydrological station and the installation information of the ultrasonic time difference method flow measurement equipment, calculate the local water-passing section area of ​​each layer of the river section, and the local flow velocity time series data of each layer obtained by the ultrasonic time difference method, perform data quality inspection and screening on the measured flow and the ultrasonic time difference method online monitoring data, process the analysis and inspection data, and deduce the local average flow velocity of each layer of the river section according to the local flow velocity of each layer obtained by the ultrasonic time difference method.

[0013] The local flow rate of each layer of the river channel cross-section is calculated based on the obtained local cross-sectional area of ​​each layer of the river channel cross-section and the local average flow velocity of each layer of the river channel cross-section.

[0014] The data quality inspection and screening of the measured flow and ultrasonic time difference method online monitoring data specifically includes eliminating data showing interruption of stratified flow velocity, obviously unreasonable stratified flow velocity, and data whose measured average flow velocity is not between the flow velocities of each layer.

[0015] The ultrasonic time difference method derivation model and objective function are established, and the constraints in the model calculation are determined as follows:

[0016] Establish the multi-objective function of the ultrasonic time difference method dynamic model, including system error μ and standard deviation σ,

[0017]

[0018]

[0019] Where n is the number of times the flow rate is calculated, Q i is the average flow model calculated value of the i-th section, Q ci is the measured value of the average flow rate of the i-th section, μ is the system error, σ is the standard deviation,

[0020] The ultrasonic time difference method uses the first-order formula to deduce the model as shown below:

[0021]

[0022] Where Q is the calculated value of the single flow model of the section, a i is the model calibration value of the ith layer of the stratified weight coefficient, A i is the local water flow area of ​​the layer, V i is the average flow velocity of the ith layer in the stratified ultrasonic time difference method during a single flow test, b is the correction coefficient,

[0023] Determine the constraints to ensure that the results of the analysis and optimization calculations conform to the actual environment. The constraints are that the calculated Q is a positive value, a i a should be between -10 and 10, b should be between -200 and 200, μ should be less than 1%, and σ should be less than 10%.

[0024] Based on the established objective function, the particle swarm optimization algorithm is used to iteratively calculate the time difference method derivation model to deduce the optimal solution:

[0025] Determine the particle swarm size, particle dimension, number of iterations and inertia weight. The particle swarm size is 30; the particle dimension is the number of layers installed by the time difference method + 2, that is, a i and b; the number of iterations k is 100; the inertia weight ω is introduced, and the inertia weight is dynamically adjusted to balance the globality and convergence speed of convergence, and the number of iterations is linearly decreased.

[0026]

[0027] In the formula, ω max is the initial maximum inertia weight, ω min is the minimum inertia weight when iterating to the maximum number of generations, k max is the maximum number of iterations, ω(k) is the inertia weight value after iteration k,

[0028] Randomly initialize the position and velocity of each particle, that is, initialize the particle swarm size to determine the number of a i The arrays of a and b are used to calculate the flow rate of the corresponding model for each comparison according to the model formula, and the objective function of each particle is counted. The smaller the objective function, the better it is. The best result is selected from all particles as the group optimal solution.

[0029] Update the speed and position of each particle. In the jth iteration calculation, the current position of the i-th particle is Current speed is The optimal solution of the current particle in the past iteration process is The best solution in the particle swarm in the past is (a1, a2, ... a c ,b), where c is the number of velocity layers, then the new velocity of the i-th particle is

[0030]

[0031] Where c1, c2 are learning factors, taking the constant 2, r1, r2 are random numbers uniformly distributed in the interval [0, 1],

[0032] Then the corresponding new position of the i-th particle is

[0033] a i(j+1)=a ij +v i(j+1)

[0034] Repeat the above steps to calculate the objective function of each particle, select the best result as the optimal solution of the particle, and select the optimal solution of all particles.

[0035] Based on deep learning in actual hydrological monitoring production applications, new working conditions are added to the model for secondary optimization solution. Specifically,

[0036] When adding measured flow results, first analyze the time difference flow data within the response period, check the rationality of the data, and conduct quality inspection and screening of the test flow data;

[0037] If the new measured flow results are used to calculate the system error and standard deviation within the specification requirements, recalibration is not required. If the system error and standard deviation change significantly, recalibration should be performed.

[0038] Continue to calculate the particle swarm optimization algorithm, and the initial particle position can be determined randomly, or the interval [a i -σ a ,a i +σ a ], and initialize the position and velocity of each particle within this interval.

[0039] In a second aspect, the present application provides a system for generating a dynamic calculation model for online flow monitoring using a time difference method, comprising:

[0040] Data collection and screening module, used to obtain basic flow characteristic parameters of hydrological stations, and collect flow measurement results of ADCP and cableway current meter method and online monitoring data of ultrasonic time difference method in the same period;

[0041] Model building module, used to establish ultrasonic time difference method derivation model and objective function, and determine the constraints in model calculation;

[0042] The derivation module is used to iteratively calculate the time difference method derivation model based on the established objective function and deduce the optimal solution by using the particle swarm optimization algorithm;

[0043] The secondary optimization solution module performs deep learning based on actual hydrological monitoring production applications and adds new working conditions to the model for secondary optimization solution.

[0044] The calculation model generation module generates the ultrasonic time difference method calculation model based on the optimal solution.

[0045] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a program code, and when the program code is executed by a processor, the steps of the method for generating a dynamic calculation model for online flow monitoring using the time difference method as described above are implemented.

[0046] Compared with the prior art, the beneficial effects of the present invention are: it can be applied to online monitoring and calculation of flow in rivers, reservoir inflows and outflows, weirs and canals, etc., to calibrate model parameters better and faster, and is suitable for non-stable water flows affected by various factors and measuring stations with complex flow changes, and solves the technical problem of hysteresis in model parameter calibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 It is an installation arrangement and flow calculation principle diagram of the ultrasonic time difference device according to an embodiment of the present invention.

[0049] Figure 2 It is a schematic diagram of the method flow of an embodiment of the present invention.

[0050] Figure 3 This is a system block diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0052] The terms "comprises," "comprising," or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0053] The terms "first", "second", etc. are only used to distinguish one entity or operation from another entity or operation, and should not be understood as indicating or implying relative importance, nor should they be understood as requiring or implying any such actual relationship or order between these entities or operations.

[0054] like Figure 1 , Figure 1 This is the installation layout of the ultrasonic time difference equipment and the flow calculation principle diagram, which reflects the principle and method of river test section switch and layered flow calculation. In the figure, V1 is the bottom layer flow velocity, Vn is the upper layer flow velocity, and V2~Vn-1 is the middle layer flow velocity. The monitoring equipment divides the river section flow into n layers, and the flow of each layer is calculated using the layer average flow velocity × layer area, where the uppermost layer average flow velocity uses the uppermost layer flow velocity, the lowermost layer average flow velocity uses the lowermost layer flow velocity, and the middle layer flow velocity uses the average value of the adjacent layer flow velocities.

[0055] Combined with Figure 2 Taking the basic test section of a hydrological station on the Han River as an example, the derivation method of the dynamic model of online flow monitoring using the ultrasonic time difference method under the influence of water conservancy projects is explained in detail.

[0056] Step 1: Online monitoring flow rate comparison collection and data screening. Obtain the basic flow characteristic parameters of the hydrological station, and collect the flow rate results measured by the ADCP and cableway current meter method and the online monitoring data of the ultrasonic time difference method in the same period.

[0057] Step 1a: Obtain the river section and water level process at the hydrological station.

[0058] Step 1b: Obtain the hydrological test environment of the hydrological station and the installation information of the ultrasonic time difference flow measurement equipment, and calculate the local water-passing cross-sectional area of ​​each layer of the river water-passing section.

[0059] Step 1c: obtaining the local flow velocity time series data of each layer by ultrasonic time difference method.

[0060] Step 1d: Check and screen the data quality of the measured flow and ultrasonic time difference method online monitoring data. When the following situations occur, the measured flow in the corresponding period will not be included in the modeling.

[0061] (1) The stratified flow rate is interrupted.

[0062] (2) The laminar velocity is obviously unreasonable, such as a large difference (more than 0.04 m / s when below 0.2, more than 20% when 0.2-0.6, and more than 10% when above 0.6), or the four-layer velocity is completely equal, or the bottom layer velocity is greater than the upper layer velocity, etc. (The above limit values ​​should be determined based on the analysis of the measuring station)

[0063] (3) The measured average flow velocity is not between the flow velocities of each layer.

[0064] Step 1e: Process the analysis and inspection data.

[0065] Based on the local flow velocity of each layer obtained by the ultrasonic time difference method, the local average flow velocity of each layer in the water-passing section of the river is deduced.

[0066] The local flow rate of each layer of the river channel cross-section is calculated based on the obtained local cross-sectional area of ​​each layer of the river channel cross-section and the local average flow velocity of each layer of the river channel cross-section.

[0067] Step 2: Establish the ultrasonic time difference method inference model and objective function, and analyze and determine the constraints in the model calculation.

[0068] Step 2a: Establish the multi-objective function of the ultrasonic time difference method dynamic model, including the system error μ and the standard deviation σ.

[0069]

[0070]

[0071] Where n is the number of times the flow rate is calculated, Q i is the average flow model calculated value of the i-th section, Q ci is the measured value of the average flow rate of the ith section, μ is the systematic error, and σ is the standard deviation.

[0072] Step 2b: In the derivation process, the model formula generally adopts a polynomial. In the example, a first-order formula can be used as shown below:

[0073]

[0074] Where Q is the calculated value of the single flow model of the section, a i is the model calibration value of the ith layer of the stratified weight coefficient, A i is the local water flow area of ​​the layer, V i is the average flow velocity of the ith layer in the stratified ultrasonic time difference method during a single flow test, and b is the correction coefficient.

[0075] Step 2c: Determine the constraints to ensure that the results of the analysis and optimization calculations are consistent with the actual environment.

[0076] (1) The calculated Q is generally a positive value, depending on the hydrological station.

[0077] (2)a i Generally it is between -10 and 10.

[0078] (3) b is generally between -200 and 200, depending on the hydrological station.

[0079] (4)μ should be less than 1% and σ should be less than 10%.

[0080] Step 3: Based on the established objective function, the particle swarm optimization algorithm is used to iteratively calculate the time difference method model to deduce the optimal solution.

[0081] Step 3a, determine the particle swarm size, particle dimension, number of iterations and inertia weight. In this example, although the flow conditions of the hydrological station are relatively complex, the derived parameters are relatively simple, and the particle swarm size can be 30; the particle dimension is the number of installation layers of the time difference method + 2, that is, a i and b; the number of iterations k is 100; the inertia weight ω is introduced, and the inertia weight is dynamically adjusted to balance the globality and convergence speed of convergence, and the number of iterations is linearly decreased.

[0082]

[0083] In the formula, ω max is the initial maximum inertia weight, ω min is the minimum inertia weight when iterating to the maximum number of generations, k max is the maximum number of iterations, and ω(k) is the inertia weight value after iteration k.

[0084] Step 3b: Randomly initialize the position and velocity of each particle, that is, initialize and determine the size of the particle swarm for a number of a i The array of a and b. Calculate the flow rate of each comparison model according to the model formula, and count the objective function of each particle. The smaller the objective function, the better. Select the best result from all particles as the group optimal solution.

[0085] Step 3c: Update the speed and position of each particle. In the jth iteration, the current position of the i-th particle is Current speed is The optimal solution of the current particle in the past iteration process is The best solution in the particle swarm in the past is (a1, a2, ... a c ,b), where c is the number of velocity layers. Then the new velocity of the i-th particle is

[0086]

[0087] Where c1 and c2 are learning factors, which are generally constants of 2, and r1 and r2 are random numbers uniformly distributed in the interval [0, 1].

[0088] Then the corresponding new position of the i-th particle is

[0089] a i(j+1) =a ij +v i(j+1)

[0090] Step 3d: Repeat steps 3a and 3b to calculate the objective function of each particle, select the best result as the optimal solution of the particle, and select the optimal solution of all particles.

[0091] Step 4: Use deep learning in practical applications to add new working conditions into the model system for secondary optimization solution.

[0092] Step 4a: When adding measured flow results, first analyze the time difference flow data within the response period to check the rationality of the data. Perform test flow data quality inspection and screening according to the requirements of step 1d.

[0093] Step 4b: If the new measured flow results are used to calculate the system error and standard deviation within the specification requirements, recalibration is not required. If the system error and standard deviation change significantly, recalibration should be performed.

[0094] Step 4c: Continue the particle swarm optimization algorithm to calculate according to step 3. The initial particle position can be determined randomly or the interval [a i -σ a ,a i +σ a ], and initialize the position and velocity of each particle within this interval.

[0095] Step 5: Based on the optimal solution, obtain the time difference method calculation model and apply it to flood reporting and compilation applications.

[0096] Step 5a: embed the time difference method calculation model formula into the flood reporting system software, obtain the ultrasonic time difference method velocity data of each layer in real time, and after a reasonable data check, calculate the time difference method flow and perform average sliding processing, as shown in the following formula, and use it as the corresponding flow for flood reporting.

[0097]

[0098] Where Q t The smoothed flow rate for period t, w t is the time period weight coefficient, n is the number of smoothing groups; when w=1 / n, the weights of each time period are equal, that is, the arithmetic mean of n time periods, which is called the simple moving average algorithm; different weights can be set for different time periods, which is the weighted moving average. In this actual application, the five-point cubic smoothing algorithm is used.

[0099] Step 5b: Apply the model algorithm to the integrated flow calculation. According to the flow test integration specification, the results are error-checked. If the test result is poor, step 4 should be repeated to add more factors and recalibrate the multivariate model. After passing the test, the model is used to push the full data, calculate the time series flow, and calculate the water volume for the year, month, ten days, and day, and analyze the water volume with the upstream and downstream measuring stations.

[0100] The model algorithm was applied to the calculation of the flow of a hydrological station in Hanjiang River. First, the estimated flow was compared with the synchronous measured flow, as shown in Appendix 1. After statistical analysis, the maximum error was 4.95%, the minimum error was 0.05%, the system error was 0.22%, and the standard deviation was 2.15%, which met the requirements of the specification.

[0101] Appendix 1: Statistical table of flow estimation accuracy using the time difference method at a station on the Han River

[0102]

[0103]

[0104]

[0105] like Figure 3 As shown, the embodiment of the present application provides a model generation system for a dynamic calculation method of an ultrasonic time difference method flow online monitoring, comprising

[0106] Data collection and screening module 1 is used to obtain the basic water flow characteristic parameters of the hydrological station, and collect the flow results measured by the ADCP and cableway current meter method and the online monitoring data of the ultrasonic time difference method in the same period;

[0107] Model building module 2, used to establish the ultrasonic time difference method derivation model and objective function, and determine the constraint conditions in the model calculation;

[0108] The derivation module 3 is used to iteratively calculate the time difference method derivation model based on the established objective function and deduce the optimal solution by using the particle swarm optimization algorithm;

[0109] Secondary optimization solution module 4, based on deep learning in actual hydrological monitoring production applications, adds new working conditions into the model for secondary optimization solution;

[0110] The calculation model generation module 5 generates an ultrasonic time difference method calculation model based on the optimal solution.

[0111] An embodiment of the present application provides a computer-readable storage medium storing program code. When the program code is executed by a processor, the steps of the model generation method of the ultrasonic time difference method flow online monitoring dynamic calculation method are implemented as described above.

[0112] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0113] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0114] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0115] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0116] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0117] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0118] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0119] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for generating a dynamic calculation model for online flow monitoring using a time difference method, characterized in that: The specific steps include: Online monitoring flow comparison measurement collection and data screening, obtain basic water flow characteristic parameters of the hydrological station, and collect the flow results measured by ADCP and cableway current meter method and the online monitoring data of ultrasonic time difference method in the same period; The above steps are specifically as follows: Obtain the river water section and water level process of the hydrological station, obtain the hydrological test environment of the hydrological station and the installation information of the ultrasonic time difference method flow measurement equipment, and calculate the local water section area of ​​each layer of the river water section, the local flow velocity time series data of each layer obtained by the ultrasonic time difference method, perform data quality inspection and screening on the measured flow and the ultrasonic time difference method online monitoring data, process the analysis and inspection data, and deduce the local average flow velocity of each layer of the river water section based on the local flow velocity of each layer obtained by the ultrasonic time difference method. According to the obtained local water section area of ​​each layer of the river water section and the local average flow velocity of each layer of the river water section, calculate and obtain the local flow of each layer of the river water section; Establish the ultrasonic time difference method derivation model and objective function, and determine the constraints in the model calculation; Based on the established objective function, the particle swarm optimization algorithm is used to iteratively calculate the time difference method derivation model to deduce the optimal solution, specifically, Determine the particle swarm size, particle dimension, number of iterations and inertia weight. The particle swarm size is 30; the particle dimension is the number of layers installed by the time difference method + 2, that is and b; the number of iterations k is 100; the inertia weight is introduced , and dynamically adjust the inertia weight to balance the globality and convergence speed of convergence, using a linear decrease in the number of iterations, ， In the formula, is the initial maximum inertia weight, is the minimum inertia weight when iterating to the maximum number of iterations, is the maximum number of iterations, is the inertia weight value after iteration k, Randomly initialize the position and velocity of each particle, that is, initialize the particle swarm size to determine the number of The array of a and b is used to deduce the model formula according to the ultrasonic time difference method to calculate the flow rate of each test, and the objective function of each particle is counted. The smaller the objective function, the better. The best result is selected from all particles as the group optimal solution. Update the speed and position of each particle. In the jth iteration calculation, the current position of the i-th particle is , the current speed is , the optimal solution of the current particle in the past iteration process is , the optimal solution in the particle swarm in the past is , where c is the number of velocity layers, then the new velocity of the ith particle is ， in is the learning factor, which is a constant of 2. yes A random number uniformly distributed in the interval, Then the corresponding new position of the i-th particle is ， Repeat the above steps to calculate the objective function of each particle, select the best result as the optimal solution of the particle, and select the optimal solution of all particles; Based on deep learning in actual hydrological monitoring production applications, new working conditions are added to the model for secondary optimization solution; Based on the optimal solution, an ultrasonic time-difference method calculation model is generated; The ultrasonic time difference method derivation model and objective function are established, and the constraints in the model calculation are determined as follows: Establish the ultrasonic time difference method derivation model and objective function, including system error and standard deviation , ， Where n is the number of times the flow rate is calculated. is the calculated value of the average flow rate of the qth section, is the measured value of the average flow rate of the qth section, Systematic error, is the standard deviation, The ultrasonic time difference method uses the first-order formula to deduce the model as shown below: ， In the formula, Q is the calculated value of single flow rate of the section, is the rated value of the p-th layer weight coefficient, is the local water flow area of ​​the layers, is the average flow velocity of the pth layer in the stratified ultrasonic time difference method during a single flow test, b is the correction coefficient, Determine the constraint conditions to ensure that the results of the analysis and optimization calculations conform to the actual environment. The constraint conditions are that the calculated Q is a positive value. a is between -10 and 10, b is between -200 and 200, Should be less than 1%. Should be less than 10%; Based on deep learning in actual hydrological monitoring production applications, new working conditions are added to the model for secondary optimization solution. Specifically, When adding measured flow results, first analyze the time difference flow data within the response period, check the rationality of the data, and conduct quality inspection and screening of the test flow data; If the new measured flow results are used to calculate the system error and standard deviation within the specification requirements, they should not be recalibrated. If the system error and standard deviation change significantly, they should be recalibrated. Continue to calculate the particle swarm optimization algorithm. The initial particle position can be determined randomly or the interval can be appropriately expanded in the original optimal solution value. , initialize the position and velocity of each particle within this interval.

2. The method for generating a dynamic calculation model for online flow monitoring using a time difference method according to claim 1, characterized in that: The data quality inspection and screening of the measured flow and ultrasonic time difference method online monitoring data specifically includes eliminating data showing interruption of stratified flow velocity, obviously unreasonable stratified flow velocity, and data whose measured average flow velocity is not between the flow velocities of each layer.

3. A system for generating a dynamic calculation model for online flow monitoring using a time difference method, used to implement the method described in any one of claims 1 to 2, characterized in that: include Data collection and screening module, used to obtain basic flow characteristic parameters of hydrological stations, and collect flow measurement results of ADCP and cableway current meter method and online monitoring data of ultrasonic time difference method in the same period; Model building module, used to establish ultrasonic time difference method derivation model and objective function, and determine the constraints in model calculation; The derivation module is used to iteratively calculate the time difference method derivation model based on the established objective function and deduce the optimal solution by using the particle swarm optimization algorithm; The secondary optimization solution module performs deep learning based on actual hydrological monitoring production applications and adds new working conditions to the model for secondary optimization solution. The calculation model generation module generates the ultrasonic time difference method calculation model based on the optimal solution.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, and when the program codes are executed by a processor, the steps of the method for generating a dynamic calculation model for online flow monitoring using the time difference method as described in any one of claims 1 to 2 are implemented.

Images