Implementation method for integration of online models of urban drainage system

Abstract

The invention relates to the technical field of informatization water affair management, in particular to an urban drainage system online model integration implementation method, which comprises the following steps of: 1, establishing a drainage network dynamic analysis system and a model database based on a geographic information system; step 1.1, establishing a drainage pipe network dynamic analysis system comprising a drainage system dynamic database and a drainage system dynamic data acquisition and analysis system; 1.2, dynamically updating the model database; 2, an online pipe network model calculation engine is arranged, and model operation is carried out through the calculation engine; 3, establishing an online model operation background, verifying the data in the model database through the online model operation background, and outputting a model result; and 4, establishing a model online system, and realizing interaction between the user and the urban drainage system online model through the model online system.

Description

Technical Field

[0001] The present invention relates to the technical field of information-based water management, and in particular to a method for realizing online model integration of an urban drainage system. Background Art

[0002] At present, the hydraulic models of pipe networks around the world are mainly based on traditional offline hydraulic models, that is, they are established and calibrated based on the status of the pipe network in a certain period of time (generally 1 week). However, the operation of the drainage pipe network is actually continuous and dynamic, and the offline model cannot adapt to the actual changes in the operation status of the pipe network. It can only be compensated by regular or irregular re-modeling and calibration, which has poor flexibility. On the other hand, the maintenance and update of offline models requires high professional requirements and is difficult, while drainage companies are relatively weak in the field of hydraulic models. The difficulty in maintaining and updating models has become a common pain point and bottleneck for water supply companies, and is also a common problem faced by the in-depth application of drainage pipe network models.

[0003] With the rapid development of cutting-edge technologies such as IoT sensor monitoring and data communication transmission, the accumulation of big data on the operation status monitoring of drainage pipe network systems is becoming increasingly rich. However, due to the lack of professional monitoring data analysis and mining technology, it is difficult to play the role of massive monitoring big data in actual operation management. It is urgent to explore methods and technologies for using massive pipe network monitoring big data to promote the development of drainage system management towards refinement and intelligence. On the one hand, with the rapid advancement of urbanization, especially the development of urban-rural integration, the scale of urban drainage pipe network systems continues to increase, the pipe network structure becomes increasingly complex, and the difficulty of modeling increases; on the other hand, the gradual popularization of smart water construction in the drainage industry will put forward higher requirements for the real-time and accuracy of pipe network models. In summary, the traditional drainage pipe network verification technology has bottleneck problems such as poor real-time performance and low accuracy, and can no longer meet the drainage industry's technical requirements for refinement and intelligence of models in terms of waterlogging prevention and control, pipe network operation and maintenance, and pipeline optimization and renovation.

[0004] The existing model construction is divided into three major stages: preparation, construction, and application. The steps are divided into seven steps: problem identification and goal setting, model selection -> model principles, basic data acquisition, model grid division -> model generalization, boundary condition and initial condition setting, sensitivity analysis and parameter setting, model application and deployment. The data transmission of each step is transmitted offline. If the data of the previous step changes, the subsequent steps need to be modified accordingly, which makes the modeling process more cumbersome. The bottleneck problems such as low computing efficiency, poor real-time performance, strong dependence on manual labor, and low intelligence are becoming increasingly prominent. It can no longer meet the drainage industry's needs in waterlogging prevention and control, pipe network operation and maintenance, Pipeline optimization, updating and transformation The model proposes technical requirements such as refinement and intelligence. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides an implementation method for online model integration of an urban drainage system, aiming to solve the problem that the data transmission in each step of the construction of the existing model is transmitted offline, and the data of the previous step changes, and the subsequent steps need to be modified accordingly, resulting in a cumbersome modeling process, low computing efficiency, poor real-time performance, strong dependence on manual labor, and low intelligence.

[0006] To achieve the above object, the present invention adopts the following technical solution:

[0007] A method for implementing online model integration of an urban drainage system comprises the following steps:

[0008] Step 1: Establish a drainage network dynamic analysis system and model database based on geographic information system;

[0009] Step 1.1: Establish a drainage network dynamic analysis system including a drainage system dynamic database and a drainage system dynamic data collection and analysis system, wherein the drainage system dynamic database includes a plant and station facility database, a pipe network facility database, a river facility database, a non-point source drainage user database and a monitoring database; the drainage system dynamic data collection and analysis system includes a data collection terminal and a web-based data analysis platform. The data collection terminal collects data online in real time and updates it to the drainage system dynamic database. The web-based data analysis platform views the data collected by the data collection terminal online in real time. Click to view the data on the web-based data analysis platform to display and analyze the data;

[0010] Step 1.2: Dynamically update the model database, and transmit the data to the model database online through the drainage system dynamic database, so that the model database is updated as the drainage system dynamic database is updated; wherein the model database includes the pipe network database, rainfall database, terrain / ground elevation database, current underlying surface database, current river network database and measured database;

[0011] Step 2: Deploy the online pipe network model calculation engine and perform model calculations through the calculation engine;

[0012] Step 3: Establish an online model running background, verify the data in the model database through the online model running background, and output the model results;

[0013] Step 4: Establish a model online system to enable users to interact with the online model of the urban drainage system through the model online system.

[0014] Furthermore, in the step 1.2, the drainage system dynamic database synchronously updates the data to the model database, and a corresponding mapping relationship is established between the drainage system dynamic database and the model database. At the same time, an automatic data conversion script is set on the web-based data analysis platform to convert the data style of the drainage system dynamic database into the model database style in real time and store it in the model database.

[0015] Furthermore, in step 1.2, the pipe network database includes inspection wells, drainage pipes / channels and hydraulic structures in the pipe network; the terrain / ground elevation database includes terrain points or raster files within the project scope; the rainfall database includes historical measured rainfall data, short-duration rainfall, measured rainfall / radar forecast rainfall; the current underlying surface database includes current use maps, current drainage divisions, sewage volume related data and current elevated positions; the current river network database includes river location data, river section data, river friction coefficient, river blue line and river embankment, lake capacity and depth and water surface area, lake bottom topography, river bottom topography and hydraulic structures on the river; the measured database includes historical waterlogging point data, sewage treatment plant inlet flow, sewage treatment plant effluent flow, sewage treatment plant sewage concentration, residential sewage data, enterprise sewage data, river and lake water quality data and non-point source pollutant data.

[0016] Furthermore, the inspection well includes: inspection well system type, XY coordinates, inspection well area, ground elevation and well bottom elevation; drainage pipe / channel includes: pipeline type, upstream and downstream node number, pipeline shape, pipeline diameter or channel width / height, roughness coefficient; hydraulic structures in the pipe network include: weir, water pump, intercepting well, regulating reservoir and flap gate, among which, the weir includes: upstream and downstream nodes, weir type, weir top elevation, weir top width, flow coefficient, and the water pump includes: water pump type, water pump system type, water collection tank size, upstream and downstream nodes , pump flow-lift curve, pump start water level, pump shut-off water level, acceleration delay time, deceleration delay time, pump station: dispatching rules, storage tank includes: XY coordinates, ground elevation, pool bottom elevation, water level / area relationship curve, interception well includes: XY coordinates, well bottom elevation, interception well overflow water level, well surface elevation, siphon includes: pipe top elevation, pipe diameter, arch elevation, outlet elevation, siphon starting water level, weir discharge coefficient, siphon discharge coefficient, flap gate includes: upstream and downstream nodes, bottom elevation, diameter; historical measured drop Rainfall data refers to annual rainfall data for a typical year or historical rainfall data for several months; short-duration rainfall refers to the local rainstorm intensity formula; measured rainfall / radar forecast rainfall refers to measured rainfall / radar forecast rainfall data of a rain gauge; current status maps refer to maps of various types of buildings; current drainage zoning maps refer to maps of divided catchment areas; sewage volume-related data include: catchment area, population density / number of people in each catchment area, population drainage equivalent, drainage time variation coefficient / daily variation coefficient; river channel location data refers to the centerline data of the river channel location; river channel The section data include: section location and section data, where the section data is in the form of coordinates + elevation or offset + elevation; the river friction coefficient is the Manning coefficient of the river; the river blue line and river bank are the river range and riverbank; the hydraulic structures on the river include: bridges, weirs and gates, where the bridge includes: upstream / downstream connecting section data, upstream / downstream bridge section data, bridge deck data; the weir includes: upstream and downstream nodes, weir type, weir top elevation, weir top width and flow coefficient; the gate includes: gate bottom elevation, width and opening height.

[0017] In step 2, the online pipe network model computing engine is deployed as follows: configure the operating system environment of the cloud server computer, the computing engine operating environment and basic configuration files, install the computing engine program to the cloud server computer, and after the computing engine is deployed, the model operation is performed by the computing engine installed on the cloud server computer.

[0018] In step 2, the calculation method of the calculation engine is: input model input data, the calculation engine obtains the input data, performs calculations, and outputs the result output data after the calculations are completed. Among them, the calculation engine contains various algorithms for model calculations, and calculations can be performed as long as the input data input requirements are met.

[0019] In step 3, the online model running background operates as follows:

[0020] The data from the drainage network dynamic analysis system is transmitted to the model database through an interface;

[0021] The model database data is transmitted to the calculation engine by connecting to the model service interface. The calculation engine performs model calculations. Before the model calculations, data verification is first performed. If the data verification is incorrect, a verification error file is returned. The user can make online modifications based on the error content and re-overwrite the modified content and save it to the model library. At the same time, the modified content is recorded in the log. After the modification, the data continues to be verified until the verification is successful. The model runs the simulation.

[0022] After running the model simulation, check the model again to see if there is an error. If there is an error, the error content will pop up and the user can modify it online based on the error content. After the modification is completed, submit it online for further verification in the background until there is no error and the model result is returned to the server model result library.

[0023] Furthermore, the data verification method before the model operation is to perform compliance check and topological relationship check on the input data, and the main check content is to check whether the type and parameters of the verification data and the topological relationship are correct.

[0024] Furthermore, the verification of whether the model reports an error mainly depends on analyzing the changes in the loss function during the model training process. This involves observing the changes in the loss during the model training and verification process to determine whether the model is converged, overfitted, underfitted, or divergent. This can be determined by the following situations:

[0025] 1) Model convergence: When the training loss during training decreases and the validation loss on the validation set also decreases, and both tend to be stable, it means that the model has converged;

[0026] 2) Model overfitting: When the training loss decreases during training but the validation loss on the validation set increases instead of decreasing, or the train loss decreases but the val loss remains unchanged, it indicates that the model is overfitting;

[0027] 3) Model underfitting: When the training loss train loss during training and the validation loss valloss on the validation set remain at a relatively high level, or neither of them has an obvious downward trend, it indicates that the model is underfitting;

[0028] 4) Model divergence: When both the training loss (train loss) and the validation loss (valloss) on the validation set increase or fluctuate greatly during training, it indicates that the model is diverging.

[0029] Furthermore, the model online system includes a model input interface, a model parameter modification interface and a model result display interface; the model input interface is for inputting model range, model simulation data and boundary condition data; the model parameter modification interface is used to modify the data after the model data verification error and the model simulation operation error, and the modification interface is manually modified and submitted to complete the modification; the result display interface renders and displays the model operation results according to the operation result fields and different categories of field values; the data transmission processes involved in the input interface, model parameter modification interface and result display interface are all online transmission, real-time synchronous dynamic update, and the modification of the model database does not affect the drainage system dynamic database, but the modification of the drainage system dynamic database will synchronously affect the modification of the model database.

[0030] The method for implementing online model integration of an urban drainage system described in the present invention has the following beneficial effects:

[0031] The drainage network dynamic analysis system, model database, online model operation background and model online system are combined to realize the seamless connection of GIS data of the drainage network dynamic analysis system to model data. After the basic data of the drainage system dynamic database of the drainage network dynamic analysis system is dynamically updated through the drainage system dynamic data acquisition and analysis system, the corresponding data of the model database can also be updated synchronously, eliminating the complicated process of traditional model data processing, and realizing automatic data update in the model modeling process with good real-time performance. According to actual modeling needs, the calculation engine can be customized through the model online system, and different online calculation engines can be replaced. The calculation engine deployment and model database are deployed in a distributed manner. One calculation engine can be used in multiple online model operation background projects. The model results can be displayed through the model online system and customized display analysis can be performed, with high calculation efficiency, low manual dependence and high intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of a method for implementing online model integration of an urban drainage system of the present invention;

[0033] Figure 2 It is a connection schematic diagram of the drainage network dynamic analysis system of the present invention;

[0034] Figure 3 It is a schematic diagram of the operation flow of the online model operation background of the present invention. DETAILED DESCRIPTION

[0035] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0036] like Figure 1 and Figure 2 As shown, the implementation method of the online model integration of the urban drainage system of the present invention comprises the following steps:

[0037] S1: Establish a dynamic analysis system and model database for drainage network based on geographic information system;

[0038] S1.1: Establish a drainage network dynamic analysis system including a drainage system dynamic database and a drainage system dynamic data collection and analysis system, wherein the drainage system dynamic database includes a plant and station facility database, a network facility database, a river facility database, a non-point source drainage user database and a monitoring database; the drainage system dynamic data collection and analysis system includes a data collection terminal and a web-based data analysis platform. The data collection terminal collects data online in real time and updates it to the drainage system dynamic database. The web-based data analysis platform views the data collected by the data collection terminal online in real time. Click to view the data on the web-based data analysis platform to display and analyze the data;

[0039] S1.2: Dynamically update the model database, and transmit the data to the model database online through the drainage system dynamic database, so that the model database is updated as the drainage system dynamic database is updated; wherein the model database includes the pipe network database, rainfall database, terrain / ground elevation database, current underlying surface database, current river network database and measured database;

[0040] S2: Online pipe network model calculation engine layout, and model calculation through the calculation engine;

[0041] S3: Establish an online model running background, verify the data in the model database through the online model running background, and output the model results;

[0042] S4: Establish a model online system to enable users to interact with the online model of the urban drainage system through the model online system.

[0043] The drainage network dynamic analysis system, model database, online model operation background and model online system are combined to realize the seamless connection of GIS data of the drainage network dynamic analysis system to model data. After the basic data of the drainage system dynamic database of the drainage network dynamic analysis system is dynamically updated through the drainage system dynamic data acquisition and analysis system, the corresponding data of the model database can also be updated synchronously, eliminating the complicated process of traditional model data processing, and realizing automatic data update in the model modeling process with good real-time performance. According to actual modeling needs, the calculation engine can be customized through the model online system, and different online calculation engines can be replaced. The calculation engine deployment and model database are deployed in a distributed manner. One calculation engine can be used in multiple online model operation background projects. The model results can be displayed through the model online system and customized display analysis can be performed, with high calculation efficiency, low manual dependence and high intelligence.

[0044] Furthermore, in the step 1.2, the drainage system dynamic database synchronously updates the data to the model database, and a corresponding mapping relationship is established between the drainage system dynamic database and the model database. At the same time, an automatic data conversion script is set on the web-based data analysis platform to convert the data style of the drainage system dynamic database into the model database style in real time and store it in the model database.

[0045] Furthermore, in step 1.2, the pipe network database includes inspection wells, drainage pipes / channels and hydraulic structures in the pipe network; the terrain / ground elevation database includes terrain points or raster files within the project scope; the rainfall database includes historical measured rainfall data, short-duration rainfall, measured rainfall / radar forecast rainfall; the current underlying surface database includes current use maps, current drainage divisions, sewage volume related data and current elevated positions; the current river network database includes river location data, river section data, river friction coefficient, river blue line and river embankment, lake capacity and depth and water surface area, lake bottom topography, river bottom topography and hydraulic structures on the river; the measured database includes historical waterlogging point data, sewage treatment plant inlet flow, sewage treatment plant effluent flow, sewage treatment plant sewage concentration, residential sewage data, enterprise sewage data, river and lake water quality data and non-point source pollutant data.

[0046] Furthermore, the inspection well includes: inspection well system type, XY coordinates, inspection well area, ground elevation and well bottom elevation; drainage pipe / channel includes: pipeline type, upstream and downstream node number, pipeline shape, pipeline diameter or channel width / height, roughness coefficient; hydraulic structures in the pipe network include: weir, water pump, intercepting well, regulating reservoir and flap gate, among which, the weir includes: upstream and downstream nodes, weir type, weir top elevation, weir top width, flow coefficient, and the water pump includes: water pump type, water pump system type, water collection tank size, upstream and downstream nodes , pump flow-lift curve, pump start water level, pump shut-off water level, acceleration delay time, deceleration delay time, pump station: dispatching rules, storage tank includes: XY coordinates, ground elevation, pool bottom elevation, water level / area relationship curve, interception well includes: XY coordinates, well bottom elevation, interception well overflow water level, well surface elevation, siphon includes: pipe top elevation, pipe diameter, arch elevation, outlet elevation, siphon starting water level, weir discharge coefficient, siphon discharge coefficient, flap gate includes: upstream and downstream nodes, bottom elevation, diameter; historical measured drop Rainfall data refers to annual rainfall data for a typical year or historical rainfall data for several months; short-duration rainfall refers to the local rainstorm intensity formula; measured rainfall / radar forecast rainfall refers to measured rainfall / radar forecast rainfall data of a rain gauge; current status maps refer to maps of various types of buildings; current drainage zoning maps refer to maps of divided catchment areas; sewage volume-related data include: catchment area, population density / number of people in each catchment area, population drainage equivalent, drainage time variation coefficient / daily variation coefficient; river channel location data refers to the centerline data of the river channel location; river channel The section data include: section location and section data, where the section data is in the form of coordinates + elevation or offset + elevation; the river friction coefficient is the Manning coefficient of the river; the river blue line and river bank are the river range and riverbank; the hydraulic structures on the river include: bridges, weirs and gates, where the bridge includes: upstream / downstream connecting section data, upstream / downstream bridge section data, bridge deck data; the weir includes: upstream and downstream nodes, weir type, weir top elevation, weir top width and flow coefficient; the gate includes: gate bottom elevation, width and opening height.

[0047] Specifically, the data content included in the model database is shown in Table 1.

[0048] Table 1

[0049]

[0050]

[0051]

[0052] In step 2, the online pipe network model calculation engine is specifically deployed as follows: configure the operating system environment of the cloud server computer, the calculation engine operating environment and basic configuration files, install the calculation engine program to the cloud server computer, and after the calculation engine is deployed, the model calculation is performed by the calculation engine installed on the cloud server computer. This calculation engine can be used as a calculation engine for multiple drainage pipe network information systems at the same time. It is not necessary to deploy the calculation engine on the database server of the pipe network drainage system. After the calculation engine is deployed, the model calculation is performed by the background calculation engine service. There is no need to open the operation interface of the calculation engine, and the written code script program is used to call it in the background service.

[0053] In step 2, the calculation method of the calculation engine is: input model input data, the calculation engine obtains the input data, performs calculations, and outputs the result output data after the calculations are completed. Among them, the calculation engine contains various algorithms for model calculations, and calculations can be performed as long as the input data input requirements are met.

[0054] like Figure 3 As shown, in step 3, the online model running background operates in the following specific ways:

[0055] The data from the drainage network dynamic analysis system is transmitted to the model database through an interface;

[0056] The model database data is transmitted to the calculation engine by connecting to the model service interface. The calculation engine performs model calculations. Before the model calculations, data verification is first performed. If the data verification is incorrect, a verification error file is returned. The user can make online modifications based on the error content and re-overwrite the modified content and save it to the model library. At the same time, the modified content is recorded in the log. After the modification, the data continues to be verified until the verification is successful. The model runs the simulation.

[0057] After running the model simulation, check the model again to see if there is an error. If there is an error, the error content will pop up and the user can modify it online based on the error content. After the modification is completed, submit it online for further verification in the background until there is no error and the model result is returned to the server model result library.

[0058] Furthermore, the data verification method before the model operation is to perform compliance check and topological relationship check on the input data, and the main check content is to check whether the type and parameters of the verification data and the topological relationship are correct.

[0059] Furthermore, the verification of whether the model reports an error mainly depends on analyzing the changes in the loss function during the model training process. This involves observing the changes in the loss during the model training and verification process to determine whether the model is converged, overfitted, underfitted, or divergent. This can be determined by the following situations:

[0060] 1) Model convergence: When the training loss (train loss) decreases during training and the validation loss (val loss) on the validation set also decreases, and both tend to be stable, it means that the model has converged. In this case, the training and validation errors of the model are both decreasing, and the decreasing trend is stable, indicating that the model can learn and generalize well to new data;

[0061] 2) Model overfitting: When the training loss (train loss) decreases during training but the validation loss (val loss) on the validation set increases instead of decreasing, or when the train loss decreases but the val loss remains unchanged, the model is overfitting. Overfitting means that the model learns the training data too well and performs poorly on unseen data.

[0062] 3) Model underfitting: When the training loss train loss during training and the validation loss valloss on the validation set remain at a relatively high level, or neither of them has an obvious downward trend, it indicates that the model is underfitting. Underfitting means that the model does not have enough ability to learn the patterns in the training data;

[0063] 4) Model divergence: When both the training loss (train loss) and the validation loss (valloss) on the validation set increase or fluctuate greatly during training, this may indicate that there are problems with the model training. Model divergence may be caused by unreasonable network structure design, improper learning rate setting, etc.

[0064] Further, in step 4, the model online system includes a model input interface, a model parameter modification interface and a model result display interface;

[0065] The model input interface is used to input the model range, model simulation data and boundary condition data; the model simulation data input can input six types of data including pipe network data in the model database, rainfall data, terrain / ground elevation data, current river network data and measured data;

[0066] Model parameter modification interface, used to modify the data after model data verification error and model simulation operation error. The modification interface is manually modified and submitted to complete the modification;

[0067] The result display interface renders and displays the model's running results according to the running result fields and different categories of field values;

[0068] The data transmission processes involved in the input interface, model parameter modification interface and result display interface are all online transmission, real-time synchronous dynamic update, and the modification of the model database does not affect the drainage system dynamic database, but the modification of the drainage system dynamic database will synchronously affect the modification of the model database.

[0069] The above description is only a preferred embodiment of the present invention and does not limit the technical scope of the present invention. Therefore, any slight modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A method for implementing online model integration of urban drainage system, characterized in that: The following steps are involved: Step 1: Establish a drainage network dynamic analysis system and model database based on geographic information system; Step 1.1: Establish a drainage network dynamic analysis system including a drainage system dynamic database and a drainage system dynamic data collection and analysis system, wherein the drainage system dynamic database includes a plant and station facility database, a pipe network facility database, a river facility database, a non-point source drainage user database and a monitoring database; the drainage system dynamic data collection and analysis system includes a data collection terminal and a web-based data analysis platform. The data collection terminal collects data online in real time and updates it to the drainage system dynamic database. The web-based data analysis platform views the data collected by the data collection terminal online in real time. Click to view the data on the web-based data analysis platform to display and analyze the data; Step 1.2: Dynamically update the basic data of the model database, so that the model database is updated along with the update of the drainage system dynamic database; wherein the model database includes the pipe network database, rainfall database, terrain / ground elevation database, current underlying surface database, current river network database and measured database; Step 2: Deploy the online pipe network model calculation engine and perform model calculations through the calculation engine; Step 3: Establish an online model running background, verify the data in the model database through the online model running background, and output the model results; Step 4: Establish a model online system to enable users to interact with the online model of the urban drainage system through the model online system.

2. The method for realizing online model integration of urban drainage system according to claim 1, characterized in that: In step 1.2, the drainage system dynamic database synchronously updates the data to the model database, and a mapping relationship is established between the drainage system dynamic database and the model database. At the same time, an automatic data conversion script is set on the web data analysis platform to convert the data style of the drainage system dynamic database into the model database style in real time and store it in the model database.

3. The method for realizing online model integration of urban drainage system according to claim 1, characterized in that: In step 1.2, the pipe network database includes inspection wells, drainage pipes / channels and hydraulic structures in the pipe network; the terrain / ground elevation database includes terrain points or raster files within the project scope; the rainfall database includes historical measured rainfall data, short-duration rainfall, measured rainfall / radar forecast rainfall; the current underlying surface database includes current use maps, current drainage divisions, sewage volume related data and current elevated positions; the current river network database includes river location data, river section data, river friction coefficient, river blue line and river embankment, lake capacity and depth and water surface area, lake bottom topography, river bottom topography and hydraulic structures on the river; the measured database includes historical waterlogging point data, sewage treatment plant inlet flow, sewage treatment plant effluent flow, sewage treatment plant sewage concentration, residential sewage data, enterprise sewage data, river and lake water quality data and non-point source pollutant data.

4. The method for realizing online model integration of urban drainage system according to claim 3 is characterized in that: Inspection wells include: Manhole system type, XY coordinates, manhole area, ground elevation and manhole bottom elevation; Drainage pipes / channels include: pipeline type, upstream and downstream node numbers, pipe shape, pipe diameter or channel width / height, and roughness coefficient; hydraulic structures in the pipe network include: weirs, water pumps, interception wells, regulating reservoirs, and flap gates. Among them, weirs include: upstream and downstream nodes, weir types, weir top elevation, weir top width, and flow coefficient; water pumps include: water pump type, water pump system type, water collection tank size, upstream and downstream nodes, water pump flow-lift curve, pump start water level, pump shut-off water level, acceleration delay time, deceleration Delay time, water pump station: dispatching rules, regulating pool includes: XY coordinates, ground elevation, pool bottom elevation, water level / area relationship curve, interception well includes: XY coordinates, well bottom elevation, interception well overflow water level, well surface elevation, siphon includes: pipe top elevation, pipe diameter, arch elevation, outlet elevation, siphon starting water level, weir discharge coefficient, siphon discharge coefficient, flap gate includes: upstream and downstream nodes, bottom elevation, diameter; historical measured rainfall data is the annual rainfall data of a typical year or the historical rainfall data of several months; short-term rainfall is the local rainstorm intensity formula; measured rainfall / radar forecast rainfall is the measured rainfall / radar forecast rainfall data of the rain gauge; the current map is the map of each type of building; the current drainage zoning map is the divided catchment area map; sewage volume related data include: catchment area, population density / number of people in each catchment area, population drainage equivalent, drainage time variation coefficient / daily variation coefficient; river channel location data is the centerline data of the river channel location; river channel section data includes: section location The data include location and section data, where the section data is in the form of coordinates + elevation or offset + elevation; the friction coefficient of the river channel is the Manning coefficient of the river channel; the blue line and river bank of the river channel are the river channel range and river bank line; the hydraulic structures on the river channel include: bridges, weirs and gates, where the bridge includes: upstream / downstream connecting section data, upstream / downstream bridge section data, bridge deck data; the weir includes: upstream and downstream nodes, weir type, weir top elevation, weir top width and flow coefficient; the gate includes: gate bottom elevation, width and opening height.

5. The method for realizing online model integration of urban drainage system according to claim 1, characterized in that: In step 2, the online pipe network model computing engine is deployed as follows: configure the operating system environment of the cloud server computer, the computing engine operating environment and basic configuration files, install the computing engine program to the cloud server computer, and after the computing engine is deployed, the model operation is performed by the computing engine of the cloud server computer.

6. The method for realizing online model integration of urban drainage system according to claim 5, characterized in that: The calculation method of the calculation engine is: input model input data, after the calculation engine obtains the input data, it performs calculations, and after the calculations are completed, it outputs the result output data. Among them, the calculation engine contains various algorithms for model calculations, and calculations can be performed as long as the input data input requirements are met.

7. The method for realizing online model integration of urban drainage system according to claim 1, characterized in that: In step 3, the online model running background operates as follows: The data from the drainage network dynamic analysis system is transmitted to the model database through an interface; The model database data is transmitted to the calculation engine by connecting to the model service interface. The calculation engine performs model calculations. Before the model calculations, data verification is first performed. If the data verification is incorrect, a verification error file is returned. The user can make online modifications based on the error content and re-overwrite the modified content and save it to the model library. At the same time, the modified content is recorded in the log. After the modification, the data continues to be verified until the verification is successful. The model runs the simulation. After running the model simulation, check the model again to see if there is an error. If there is an error, the error content will pop up and the user can modify it online based on the error content. After the modification is completed, submit it online for further verification in the background until there is no error and the model result is returned to the server model result library.

8. The method for realizing online model integration of urban drainage system according to claim 7, characterized in that: The data verification method before model operation is to perform compliance check and topological relationship check on the input data. The main check content is to check whether the type and parameters of the verification data and the topological relationship are correct.

9. The method for realizing online model integration of urban drainage system according to claim 7, characterized in that: Verifying whether the model reports an error mainly relies on analyzing the changes in the loss function during model training. This involves observing the changes in the loss during model training and verification to determine whether the model is convergent, overfitting, underfitting, or diverging. This can be determined by the following situations: 1) Model convergence: When the training loss (train loss) decreases during training and the validation loss (valloss) on the validation set also decreases, and both tend to be stable, it means that the model has converged; 2) Model overfitting: When the training loss (train loss) decreases during training but the validation loss (valloss) on the validation set increases instead of decreasing, or when the train loss decreases but the val loss remains unchanged, the model is overfitting; 3) Model underfitting: When the training loss during training and the validation loss on the validation set remain at a relatively high level, or neither of them has a clear downward trend, it indicates that the model is underfitting; 4) Model divergence: When both the training loss (trainloss) during training and the validation loss (val loss) on the validation set increase or fluctuate greatly, it indicates that the model is diverging.

10. The method for realizing online model integration of urban drainage system according to claim 1, characterized in that: In step 4, the model online system includes a model input interface, a model parameter modification interface and a model result display interface; the model input interface is for inputting model range, model simulation data and boundary condition data; The model parameter modification interface is used to modify the data after the model data verification error and the model simulation operation error. The modification interface is manually modified and submitted to complete the modification; the result display interface renders and displays the model operation results according to the operation result fields and different categories of field values; The data transmission processes involved in the input interface, model parameter modification interface and result display interface are all online transmission, real-time synchronous dynamic update, and the modification of the model database does not affect the drainage system dynamic database, but the modification of the drainage system dynamic database will synchronously affect the modification of the model database.

Images