Urban waterlogging evolution monitoring and early warning method based on multi-source data

Abstract

The present application relates to the technical field of data processing, in particular to a kind of urban waterlogging water evolution monitoring and early warning method based on multi-source data.The method generates the rainfall intensity and water depth of each cycle end of each catchment unit pre-divided;According to its catchment area, rainfall intensity and water depth at the end of the current cycle and water depth at the end of the last cycle, the actual inflow is determined by inversion;And identify whether to generate reverse flow identifier for catchment unit;If the current rainfall intensity and water depth meet the preset condition, compare the theoretical drainage capacity value determined based on the current water depth with the actual inflow, determine the drainage resistance index;From the catchment unit that meets the preset condition, the to-be-monitored unit is obtained by screening;According to the drainage resistance index difference of each to-be-monitored unit and its neighbor catchment unit, generate drainage anomaly label, label is used to indicate that there is local physical blockage or hydraulic jacking in the corresponding unit, thereby improving the accuracy and efficiency of emergency response.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method for monitoring and early warning of urban flooding evolution based on multi-source data. Background Technology

[0002] Urban flooding is a prominent hydrological disaster encountered during urbanization. Its rapid spread and wide-ranging impact pose serious threats to urban transportation, residents' lives, and the safety of public facilities. Accurate monitoring of flooding progression and timely dissemination of early warning information are core requirements for urban emergency management.

[0003] Currently, mainstream technologies fall into two main categories: one is point-based monitoring based on the Internet of Things (IoT), such as deploying electronic water gauges at flood-prone areas to obtain real-time surface water depth; the other is numerical simulation based on computer simulation, which involves constructing a hydrodynamic model of the urban drainage network and inputting rainfall data to simulate the water accumulation process. However, sensors such as electronic water gauges can only provide phenological data such as water depth and cannot answer the crucial question of the cause of water accumulation—whether it is due to partial blockage of storm drains by debris or systemic hydraulic backwater caused by rising water levels in downstream rivers. Without identifying the cause, it is difficult to provide accurate guidance for emergency response (such as clearing storm drains or activating pumping stations). Secondly, hydrodynamic models typically operate under fixed design parameters. If the cause of water accumulation is actual blockage of storm drains or downstream backwater, the physical boundary conditions have changed drastically, but the hydrodynamic model cannot perceive and adapt to this change, continuing to calculate based on ideal conditions. This leads to predictions that deviate significantly from reality, rendering early warnings ineffective. Summary of the Invention

[0004] To address the technical problem that existing technologies fail to identify the causes of waterlogging, leading to significant deviations in waterlogging predictions from reality, this invention provides a method for monitoring and early warning of urban waterlogging evolution based on multi-source data. The specific technical solution adopted is as follows:

[0005] This invention proposes a method for monitoring and early warning of urban flooding evolution based on multi-source data. The method includes:

[0006] Periodically and synchronously acquire meteorological radar grid data and video water accumulation data, and process them to generate rainfall intensity and water accumulation depth for each pre-divided catchment unit at the end of each cycle;

[0007] For each catchment unit, the actual inflow rate into the underground pipe network is determined by inversion based on its catchment area, the rainfall intensity and water depth at the end of the current cycle and the water depth at the end of the previous cycle; and whether a backflow indicator is generated for the catchment unit is identified.

[0008] If the current rainfall intensity and water depth meet the preset conditions, the drainage resistance index will be determined by comparing the theoretical drainage capacity value determined based on the current water depth with the actual inflow rate.

[0009] From the water collection units that meet the preset conditions, exclude units with backflow indicators or drainage blockage indicators that exceed the preset extreme threshold to obtain the units to be monitored; based on the differences in drainage blockage indicators between each unit to be monitored and its neighboring water collection units, generate drainage anomaly tags, which are used to indicate that the corresponding unit has local physical blockage or hydraulic backing.

[0010] The drainage hydrodynamic model is modified based on the labels; the modified model is used to predict water accumulation evolution and generate early warning information.

[0011] Furthermore, the process of pre-dividing each catchment unit includes:

[0012] Acquire geographic information system data of urban drainage pipe networks;

[0013] Based on the geographic information system data of the urban drainage network, the spatial location of all rainwater inlet nodes is extracted.

[0014] By combining digital elevation model and surface runoff path analysis, the corresponding surface catchment area is divided for each stormwater inlet node; theoretically, the surface runoff in each catchment area flows into the only stormwater inlet in the catchment area.

[0015] Each rainwater inlet node and its corresponding surface catchment area are defined as a catchment unit; the catchment area of ​​the catchment unit is the horizontal projected area of ​​the corresponding surface catchment area.

[0016] Furthermore, the process of generating rainfall intensity and water depth at the end of each cycle for each catchment unit includes:

[0017] Acquire radar reflectivity data published in grid format; and backtrack the raw water depth data collected within the current period at the end of the current period.

[0018] The radar reflectivity data is converted into rainfall intensity data; the raw water depth data is then low-pass filtered to obtain the processed water depth data.

[0019] For each catchment unit, the rainfall intensity allocated to the catchment unit is determined by interpolation calculation based on the spatial relationship between its geometric center location and the surrounding radar grid points.

[0020] The arithmetic mean of all water depths in the processed water depth data is used as the water depth of the catchment unit at the end of the current cycle.

[0021] Furthermore, the process of determining the actual inflow rate includes:

[0022] For each catchment unit, calculate the product of the water depth and the catchment area at the end of the current cycle as the current surface water volume; calculate the water depth and the catchment area at the end of the previous cycle as the surface water volume of the previous cycle.

[0023] Calculate the change in current surface water volume compared to the previous period's surface water volume per unit time, and use this as the change in surface water volume.

[0024] Calculate the product of the rainfall intensity at the end of the current cycle and the catchment area as the rainfall replenishment; subtract the change in surface water retention from the rainfall replenishment to obtain the theoretical inflow rate.

[0025] Set the theoretical inflow rate as the actual inflow rate.

[0026] Furthermore, the step of identifying whether a backflow identifier is generated for the water catchment unit includes:

[0027] If the theoretical inflow rate remains below the preset negative fault tolerance threshold within a preset time window, it is determined that backflow has occurred in the water collection unit, and a backflow flag is generated for it; wherein, the absolute value of the preset negative fault tolerance threshold is greater than zero.

[0028] Furthermore, the method also includes:

[0029] For each catchment unit, if the current water depth is greater than the preset depth threshold, it is further determined whether the current rainfall intensity is greater than the rainfall intensity threshold; if the current rainfall intensity is greater than the rainfall intensity threshold, the catchment unit is determined to meet the preset conditions and proceeds to the subsequent calculation process.

[0030] If the current water depth is greater than the preset depth threshold, and the current rainfall intensity is not greater than the rainfall intensity threshold, the catchment unit is determined to be in the post-rain dissipation period, and the subsequent calculation process is initiated in the first mode. In the first mode, when calculating the drainage obstruction index, if the change in surface water level is negative, the absolute value of the change in surface water level is taken as the actual inflow rate; if the change in surface water level is positive, the process is directly switched to the backflow determination process.

[0031] If the current water depth is not greater than the preset depth threshold, the water collection unit is determined not to meet the preset conditions, will not proceed to the subsequent calculation process, and the drainage obstruction index will be set to the preset default value.

[0032] Furthermore, the process for determining the theoretical drainage capacity value includes:

[0033] Obtain the design flow area of ​​the rainwater inlet of the water collection unit, and obtain the preset reference flow coefficient used to characterize the water flow efficiency;

[0034] Based on the water pressure generated by the current water depth, and combined with the design flow area of ​​the rainwater inlet and the preset benchmark flow coefficient, the maximum drainage flow of the rainwater inlet under ideal unobstructed operating conditions is calculated as the theoretical drainage capacity value.

[0035] Furthermore, the process for determining the drainage resistance index includes:

[0036] Calculate the sum of the actual inflow rate and the preset minimum positive number, and use it as the first sum value;

[0037] The ratio of the theoretical drainage capacity value to the first sum value is calculated and used as a drainage resistance index.

[0038] Furthermore, the neighboring water collection unit is used to represent other water collection units in the drainage network topology that have a direct or indirect hydraulic connection with the rainwater inlet node corresponding to the current water collection unit to be monitored through underground pipes.

[0039] The step of generating drainage anomaly labels based on the differences in drainage obstruction indices between each monitored unit and its neighboring catchment units includes:

[0040] A hydraulic correlation matrix is ​​constructed based on the drainage network topology. The hydraulic correlation matrix is ​​used to quantitatively describe the intensity of hydraulic influence between any two catchment units due to underground pipe connections. The more direct the underground connection between the two units and the shorter the path, the higher the intensity of the hydraulic influence.

[0041] Calculate the arithmetic mean and standard deviation of the drainage resistance index of all monitored units; for each monitored unit, obtain the standardized resistance index based on the arithmetic mean and standard deviation; obtain other units that are hydraulically related to the monitored unit from the hydraulic correlation matrix, as well as the hydraulic influence intensity of all other units; use the hydraulic influence intensity as the hydraulic correlation weight.

[0042] For each unit to be monitored, the standardized resistance index of other units is multiplied by the hydraulic correlation weight of each unit to obtain a weighted value; the sum of the weighted values ​​of the unit to be monitored and all other units is calculated as the neighbor average resistance index of the unit to be monitored.

[0043] Based on the standardized obstruction index and the neighbor's average obstruction index, corresponding drainage anomaly labels are generated.

[0044] Furthermore, the generation of corresponding drainage anomaly labels based on standardized obstruction indices and neighboring average obstruction indices includes:

[0045] For each monitoring unit, if the standardized obstruction index of the monitoring unit is higher than a preset first threshold and the average obstruction index of the neighbors is not higher than a preset second threshold, a first drainage anomaly label is generated; wherein, the first drainage anomaly label is used to indicate that there is a local physical blockage in the space represented by the monitoring unit.

[0046] If the standardized resistance index of the monitored unit is higher than the preset first threshold, and the average resistance index of the neighboring unit is also higher than the preset second threshold, a second drainage anomaly label is generated; wherein, the second drainage anomaly label is used to indicate that there is system hydraulic backing in the space represented by the monitored unit.

[0047] The present invention has the following beneficial effects:

[0048] This invention periodically and synchronously acquires meteorological radar grid data and video water accumulation data. Through standardized processing, it accurately matches multi-source data with pre-divided catchment units, generating rainfall intensity and water accumulation depth for each catchment unit at the end of each cycle. This effectively solves the problem of insufficient multi-source data fusion in existing methods, ensuring that the monitoring data can accurately reflect the real-time water accumulation status of each area, and providing high-quality data support for subsequent flow rate inversion and retention index calculation. For each catchment unit, combining the catchment area, water accumulation depth in previous and subsequent cycles, and current rainfall intensity, the actual inflow rate is inverted by the difference between the change in retained water volume and the rainfall replenishment. Furthermore, it utilizes drainage retention indices to intelligently zone... The system combines localized physical blockage with systemic hydraulic backing, enabling the output warning information to not only include "when and where water accumulation may occur" but also clearly indicate "the cause of water accumulation," thereby greatly improving the accuracy and efficiency of emergency response. Real-time correction of the hydrodynamic model based on drainage anomaly labels eliminates the deviation between fixed model parameters and actual drainage anomalies. The corrected model is then used to predict water accumulation evolution and generate warning information, allowing the hydrodynamic model to approximate real, dynamically changing physical conditions in real time. This fundamentally eliminates model prediction bias caused by changes in boundary conditions, significantly improving the accuracy and reliability of water accumulation evolution prediction in complex and realistic rainstorm scenarios. Attached Figure Description

[0049] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a flowchart of a method for monitoring and early warning of urban flooding evolution based on multi-source data, provided in one embodiment of the present invention.

[0051] Figure 2 This is an example diagram illustrating the actual inflow rate determination process provided in one embodiment of the present invention;

[0052] Figure 3 This is an example diagram illustrating the drainage anomaly label generation process provided in one embodiment of the present invention. Detailed Implementation

[0053] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method for monitoring and early warning of urban flooding evolution based on multi-source data proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0055] The following description, in conjunction with the accompanying drawings, details a specific scheme for a method for monitoring and early warning of urban flooding evolution based on multi-source data, provided by the present invention.

[0056] Please see Figure 1 The diagram illustrates a flowchart of a method for monitoring and early warning of urban flooding evolution based on multi-source data, according to an embodiment of the present invention. The method includes:

[0057] S101: Periodically and synchronously acquire meteorological radar grid data and video water accumulation data, and process them to generate the rainfall intensity and water accumulation depth of each pre-divided catchment unit at the end of each cycle.

[0058] Weather radar grid data is a digital product of radar echo intensity (i.e. reflectivity factor) recorded at each grid point after dividing the atmospheric space into countless regular three-dimensional grid points.

[0059] For example, if it is known that meteorological radar grid data is processed and published in real time by national or regional meteorological operational departments, then grid data covering the area of ​​interest can be obtained periodically (e.g., every 5-10 minutes) through dedicated data interfaces, meteorological data service platforms, or APIs in standard meteorological data formats (e.g., NetCDF, GRIB2).

[0060] Video water accumulation data is obtained by analyzing real-time video streams captured by surveillance cameras deployed on the road surface, and using computer vision algorithms to automatically identify and extract information on the depth of water accumulation on the ground, expressed in pixel coordinates or physical units.

[0061] For example, the acquisition of video water accumulation data includes: real-time processing of video streams from designated cameras, and estimating the water accumulation depth frame by frame in a pre-defined area of ​​interest (such as a low-lying area in a lane) using water level detection, texture segmentation, or deep learning models to form a high-frequency raw water depth data stream.

[0062] It should be noted that the methods for acquiring meteorological radar grid data and road video monitoring data involved in this invention are both existing mature technologies in this field.

[0063] It should be noted that, in order to coordinate the acquisition frequency of the two types of data (i.e., weather radar grid data and video water accumulation data) and meet the timeliness requirements of real-time early warning, this invention sets a fixed calculation cycle (e.g., 5 minutes), and all subsequent core calculation steps are triggered based on this cycle. At the end of each cycle, the latest radar grid data available at the end of the cycle is acquired synchronously, as well as the accumulated video water accumulation data within the cycle is processed retrospectively.

[0064] It should be noted that after acquiring the latest radar grid data and the original video water accumulation data, these data also need to be preprocessed, such as filtering (e.g., taking the average of the last 3 cycles) to filter out high-frequency numerical noise caused by sensor errors; unit unification and dimension checking: unify the unit of rainfall intensity to meters per second, the unit of water accumulation depth to meters, and the unit of catchment area to square meters.

[0065] In this embodiment, geographic information system (GIS) data of the urban drainage network is acquired; based on the GIS data, the spatial locations of all stormwater inlet nodes are extracted; combining digital elevation model (DEM) and surface runoff path analysis, a corresponding surface catchment area is defined for each stormwater inlet node; theoretically, surface runoff within each catchment area flows into a single stormwater inlet within that catchment area; each stormwater inlet node and its corresponding surface catchment area are defined as a catchment unit; the catchment area of ​​the catchment unit is the horizontal projection area of ​​the corresponding surface catchment area.

[0066] The Urban Drainage Network Geographic Information System (Drainage Network GIS) is a core digital archive of urban infrastructure, typically maintained by municipal drainage management departments. It contains the following key structured data: Node data: Records the spatial location (coordinates) and attributes of all drainage facilities, especially storm drain nodes, whose attributes include number, type, ground elevation, and road location; Pipeline data: Records information about the pipelines connecting each node, including start point, end point, pipe diameter, length, and slope. Data from the Urban Drainage Network GIS can be obtained through a data sharing interface, which will not be detailed in this embodiment.

[0067] Digital Elevation Model (DEM): High-precision urban ground elevation grid data used to reflect surface undulations.

[0068] It should be noted that each catchment unit has a unique identifier, usually associated with its storm drain number, and a key static attribute of each catchment unit is its catchment area. For example, the catchment area can usually be automatically calculated by GIS software to determine the horizontal projected area of ​​its corresponding surface water catchment area, typically in square meters. This is a mature existing technology in the field and will not be elaborated further in this embodiment.

[0069] It should be noted that basin delineation based on DEM and storm drain nodes is a standard and well-known analytical method in geographic information science and hydrology, and has been integrated into mainstream GIS software platforms. The inventiveness of this invention lies not in the delineation method itself, but in the creative combination of this static, geographic information-based "catchment unit" grid with dynamic, multi-source monitoring data (i.e., radar rainfall, video water accumulation).

[0070] In this embodiment, radar reflectivity data published in a grid format is acquired; and raw water depth data collected during the current cycle is backtracked to the end of the current cycle. The radar reflectivity data is converted into rainfall intensity data. The raw water depth data is low-pass filtered to obtain processed water depth data. For each catchment unit, the rainfall intensity allocated to the catchment unit is determined by interpolation calculation based on the spatial relationship between its geometric center position and the surrounding radar grid points. The arithmetic mean of all water depths in the processed water depth data is calculated as the water depth of the catchment unit at the end of the current cycle.

[0071] It should be noted that, using the well-known ZR relationship in meteorology (i.e., the empirical relationship between radar reflectivity factor Z and rainfall intensity R), the acquired radar reflectivity grid data was converted into rainfall intensity grid data point by point. Rainfall intensity grid data refers to the rainfall intensity recorded at each grid point.

[0072] It should be noted that since radar rainfall data is located at grid points, while catchment cells are irregular polygons, spatial interpolation is needed to "assign" the rainfall intensity grid data to the cells. A common method is inverse distance weighted (IDW) interpolation.

[0073] The location of the geometric center point is determined by directly calculating the centroid coordinates of any polygonal surface (i.e., a catchment unit) using the "computational geometry" tool in GIS software.

[0074] For example, first, the geometric center of the catchment unit is located, and then all radar grid points within a certain range around it are searched. The closer a grid point is to the center point, the greater its weight in the calculation result. Finally, the rainfall intensity values ​​of all grid points falling within the search range are weighted and averaged according to their distance weights. The result is then used as the rainfall intensity allocated to the catchment unit. The specific search range is determined based on the actual situation, and this embodiment does not impose a specific limitation. For example, it may be directly specified that the four nearest grid points are required to participate in the calculation.

[0075] It should be noted that, to ensure the accuracy of the actual inflow rate in subsequent physical calculations based on water balance, it is also necessary to check the unit system and dimensional consistency of the rainfall intensity and water depth at the end of each cycle for each generated catchment unit. Specifically, this includes unifying the rainfall intensity unit: the initial unit of rainfall intensity obtained from radar data is usually millimeters per hour (mm / h). Before being used in subsequent calculations, it needs to be converted to meters per second (m / s) in the International System of Units (SI). The conversion formula is: 1 mm / h ≈ 2.77778 × 10^-7 m / s. This conversion also ensures that the rainfall intensity and catchment area ( After multiplying, the volumetric flow rate can be directly obtained. / s), unit matching; water depth unit uniformity: the water depth H obtained from video analysis should be uniformly in meters (m), which is the standard unit of length in the International System of Units; catchment area unit: the catchment area of ​​a catchment unit is fixed in square meters (m²). ).

[0076] S102: For each catchment unit, based on its catchment area, the rainfall intensity and water depth at the end of the current cycle and the water depth at the end of the previous cycle, the actual inflow rate into the underground pipe network is determined by inversion; and whether a backflow indicator is generated for the catchment unit.

[0077] It is important to understand that the principle of water balance is a fundamental conservation law in fluid mechanics and hydrology. Its physical meaning can be simply described as follows: for any defined and fixed surface space area (i.e., a catchment unit), the total amount of water flowing into the area during any given time period is equal to the total amount of water flowing out of the area plus the change in the water storage within the area.

[0078] The process of determining the actual inflow rate is as follows: Figure 2 As shown, it includes:

[0079] S102-1: For each catchment unit, calculate the product of the water depth and the catchment area at the end of the current cycle as the current surface water volume; calculate the water depth and the catchment area at the end of the previous cycle as the surface water volume of the previous cycle.

[0080] Perched water volume is the estimated total volume of all water accumulated on the surface of a catchment unit at the end of a certain period, before it is discharged into the underground pipe network. A larger perched water volume at the end of a period indicates more severe waterlogging in that catchment unit during that period, likely due to recent heavy rainfall and insufficient drainage system capacity (e.g., blockage or backwater). Conversely, a smaller perched water volume indicates less or very shallow waterlogging in that catchment unit during that period, reflecting a more normal drainage system operation or that rainfall has not yet formed effective runoff.

[0081] S102-2: Calculate the change in the current surface water volume compared to the previous period's surface water volume per unit time, and use this as the change in surface water volume.

[0082] It should be noted that the change in surface water volume = (current surface water volume - previous period's surface water volume) ÷ period duration.

[0083] It is important to understand that if the change in surface perched water in a certain catchment unit between the current cycle and the previous cycle is positive and large, it indicates that the water level is rising rapidly. This means that the amount of water supplied by rainfall or backflow is much greater than the drainage capacity of the underground pipe network. If the change in surface perched water is negative and large in absolute value, it indicates that the water level is receding rapidly. This means that the drainage capacity (or evaporation and infiltration) of the underground pipe network is very strong and greater than the rainfall supply. If the change in surface perched water is close to zero, it indicates that the water depth is basically stable and in a dynamic equilibrium state, that is, the inflow rate is approximately equal to the outflow rate.

[0084] S102-3: Calculate the product of the rainfall intensity and the catchment area at the end of the current cycle as the rainfall replenishment; subtract the change in surface water from the rainfall replenishment to obtain the theoretical inflow.

[0085] It's important to understand that, based on the principle of water balance, we have the following conservation equation: Within a given period, the total water flowing into the catchment unit equals the total water flowing out of the catchment unit plus the change in water storage within the catchment unit. Specifically, in this scenario: the total water flowing into the catchment unit comes from rainfall, specifically rainfall replenishment multiplied by the period duration; the total water flowing out of the catchment unit comes from the water flow entering the ground through the storm drains, specifically the flow rate entering the ground through the storm drains multiplied by the period duration; the change in water storage within the catchment unit comes from the change in surface water volume, specifically the change in surface perched water. Substituting these values ​​into the conservation equation, we can deduce that the flow rate entering the ground through the storm drains equals the rainfall replenishment minus the change in surface perched water. Therefore, the flow rate entering the ground through the storm drains is the theoretical inflow rate.

[0086] Theoretical inflow rate: Based on the observed changes in rainfall and surface water, this amount of water should enter the ground, according to the principle of water balance.

[0087] It should be noted that if the theoretical inflow is negative, it means that even without rainfall (i.e., the rainfall supply is 0), the right side of the equation still needs a negative change in surface water retention to hold true. This is physically equivalent to "extra water gushing out from underground, leading to an increase in surface water retention," which is conclusive evidence of backflow.

[0088] S102-4: Set the theoretical inflow rate to the actual inflow rate.

[0089] In this embodiment, if the theoretical inflow rate is continuously lower than the preset negative fault tolerance threshold within a preset time window, it is determined that water backflow has occurred in the water collection unit, and a backflow identifier is generated for it; wherein, the absolute value of the preset negative fault tolerance threshold is greater than zero.

[0090] It should be noted that the physical information of the backflow has been fully recorded by the independent status flag "backflow identifier".

[0091] It should be noted that the specific value of the preset time window is determined based on the general physical time scale of hydraulic phenomena, and this embodiment does not impose a specific limitation. For example, it is known that real water backflow is usually a continuous process, so the preset time window is usually set to 2 to 3 cycles.

[0092] It should be noted that since the theoretical inflow rate is calculated from rainfall and water accumulation data that contain measurement errors, its value inherently fluctuates randomly around the true physical zero value. If zero is used as the boundary directly, these tiny negative fluctuations are easily misjudged as backflow, generating a large number of false alarms. Therefore, a negative fault tolerance threshold is preset.

[0093] The specific value of the preset negative fault tolerance threshold is determined by comprehensively considering both measurement and calculation errors, and this embodiment does not impose a specific limitation. For example, the specific determination process involves evaluating the accuracy of the sensor used (such as radar or video) under typical operating conditions, estimating the maximum possible random error range of the theoretical inflow rate caused by this, and setting the preset negative fault tolerance threshold to a typical value (such as a lower limit) of this error range, for example, for a catchment area of ​​5000... The catchment unit typically has a value of -0.005. / s.

[0094] It should be noted that backflow indicators are typically represented as logical (or Boolean) state variables. Specifically, a dedicated Boolean field is maintained in memory for each backflow unit; if a continuous backflow is detected in that unit, the value of this field is set to True (or 1); otherwise, it is False (or 0).

[0095] It should be noted that if a cold start occurs or there is no data from the previous cycle, the present invention defaults to the water depth of the previous cycle being 0, or skips the calculation of the first cycle and waits for data accumulation.

[0096] S103: If the current rainfall intensity and water depth meet the preset conditions, the drainage resistance index will be determined by comparing the theoretical drainage capacity value determined based on the current water depth with the actual inflow.

[0097] In this embodiment, for each catchment unit, if the current water depth is greater than a preset depth threshold, it is further determined whether the current rainfall intensity is greater than a rainfall intensity threshold. If the current rainfall intensity is greater than the rainfall intensity threshold, the catchment unit is determined to meet the preset conditions and enters the subsequent calculation process. If the current water depth is greater than the preset depth threshold, but the current rainfall intensity is not greater than the rainfall intensity threshold, the catchment unit is determined to be in the post-rain dissipation period and enters the subsequent calculation process in the first mode. In the first mode, when calculating the drainage obstruction index, if the change in surface water level is negative, the absolute value of the change in surface water level is taken as the actual inflow rate. If the change in surface water level is positive, the process directly proceeds to the backflow determination process. If the current water depth is not greater than the preset depth threshold, the catchment unit is determined not to meet the preset conditions, does not enter the subsequent calculation process, and the drainage obstruction index is set to a preset default value.

[0098] It should be noted that the example value of the preset depth main threshold is determined comprehensively based on the detection error level of the video water level recognition technology and the physical requirements for the formation of surface runoff. This embodiment does not impose specific limitations. For example, the value of the preset depth main threshold needs to be higher than the typical error range of the water level recognition algorithm (e.g., 1-2 cm). Moreover, it also needs to be understood through industry experience as the minimum physical depth of water accumulation (a few centimeters) required to form continuous surface runoff and generate sufficient hydrostatic pressure to drive drainage. Based on the above, the value of the preset depth main threshold can be obtained. For example, the preset depth main threshold is usually set between 0.02 meters and 0.05 meters (i.e., 2 to 5 centimeters).

[0099] It should be noted that the specific value of the rainfall intensity threshold can be determined based on publicly available meteorological definitions. For example, in meteorology, rainfall of less than 1 mm / hour is usually considered "drizzle," and the runoff it generates is negligible. Moreover, considering that the water may be absorbed by the surface or evaporated, it cannot form a stable runoff input that dominates the water balance process. Therefore, the typical range of rainfall intensity threshold values ​​is set between 0.5 mm / hour and 2.0 mm / hour.

[0100] It is understandable that if the current water depth of a certain catchment unit is greater than the preset main threshold, and the current rainfall intensity is also greater than the rainfall intensity threshold, it means that the catchment unit has both significant water accumulation and effective rainfall, and is in a "strongly active rain" state. In this case, the calculated value can be used for subsequent calculation processes.

[0101] It should be noted that if the current water depth of a certain catchment unit is greater than the preset depth threshold and the current rainfall intensity is not greater than the rainfall intensity threshold, it indicates that the catchment unit has significant water accumulation but no effective rainfall, and is in a state of "weak activity or post-rain dissipation period". At this time, the rainfall intensity is approximately 0, and the rate of decrease in surface water volume (i.e., negative surface water percolation change, which is positive) is physically equivalent to the actual water flow entering the underground pipe network. Therefore, if the surface water percolation change is negative, the absolute value of the surface water percolation change is taken as the actual inflow flow; if the surface water percolation change is positive, the process directly proceeds to the backflow judgment process.

[0102] It should be noted that if the change in surface water is positive in the absence of effective rainfall, it means that the surface water is still increasing. The only reasonable physical reason is that "backflow" has occurred. In this case, the corresponding unit is directly identified as having backflow, a backflow identifier with the highest priority (such as "EXTREME_BACKFLOW") is generated for it, and the subsequent drainage obstruction index calculation is skipped, directly entering the model correction and early warning process for backflow.

[0103] It should be noted that since the conclusion of backflow itself clearly points to the physical cause of "extremely high back pressure downstream," the downstream outlet node of the affected pipeline network can be located directly based on this conclusion, and a targeted "downstream pressure field calibration" operation can be performed. For example, the boundary water level conditions of the downstream outlet node of the affected pipeline network can be significantly increased. The specific operation is a common technical means for those skilled in the art, and will not be described in detail in this embodiment.

[0104] It should be noted that the default value is determined based on the physical definition of the drainage resistance index. Specifically, when the catchment unit does not meet the calculation conditions (e.g., no significant water accumulation), there is a lack of other data indicating abnormal drainage. Therefore, the most neutral and stable assumption is adopted: that is, its drainage is considered to be unobstructed, and the actual drainage capacity is equal to the theoretical drainage capacity, and the default value is set to 1.

[0105] It's important to understand that the actual inflow capacity of a drainage system changes in real time due to various dynamic faults such as blocked storm drains and downstream backflow. Therefore, an objective benchmark is needed to quantify the degree of degradation of "actual performance" relative to "ideal conditions." The theoretical drainage capacity value serves as such a benchmark: based on the currently measured water depth and the inherent design dimensions of the storm drains, it calculates the maximum drainage flow achievable under ideal conditions assuming complete unobstructed flow and no downstream resistance. This allows for subsequent comparison between the calculated actual inflow flow and the theoretical drainage capacity value, thereby achieving a standardized and quantitative assessment of the drainage efficiency of each catchment unit.

[0106] In this embodiment, the design flow area of ​​the rainwater inlet of the water collection unit is obtained, and the preset reference flow coefficient used to characterize the water flow efficiency is obtained; based on the water pressure generated by the current water depth, and combined with the design flow area of ​​the rainwater inlet and the preset reference flow coefficient, the maximum drainage flow of the rainwater inlet under ideal unobstructed working conditions is calculated as the theoretical drainage capacity value.

[0107] The design flow area is obtained directly from the attribute database of the urban drainage network geographic information system (GIS) for the unique stormwater inlet corresponding to each catchment unit. The design flow area is a fixed geometric attribute, typically expressed in square meters (m²). (in units of )

[0108] It should be noted that the preset reference flow rate coefficient not only represents orifice contraction but also comprehensively includes the friction loss along the connecting pipe. The specific value of the preset reference flow rate coefficient is determined based on industry experience, and this embodiment does not impose a specific limitation. For example, the preset reference flow rate coefficient is usually taken between 0.5 and 0.7.

[0109] This invention calculates the theoretical drainage capacity based on the free outflow principle of orifices in fluid dynamics. The specific process is as follows: The current water depth generates hydrostatic pressure at the rainwater inlet. This hydrostatic pressure is the fundamental driving force for water flow through the rainwater inlet. According to Bernoulli's principle, this hydrostatic pressure can be converted into a theoretical flow velocity, the magnitude of which is proportional to the square root of the water depth. Then, the theoretical flow velocity determined by the water depth is combined with the designed flow area of ​​the rainwater inlet, and the efficiency reduction when the water flows through is considered (i.e., multiplied by a preset reference flow coefficient). The maximum drainage flow rate that the rainwater inlet can pass through under ideal unobstructed flow conditions and without downstream backflow can then be calculated. Therefore, the theoretical drainage capacity can be expressed by the following formula:

[0110]

[0111] Where Q represents the theoretical drainage capacity; C represents the preset baseline flow coefficient; S represents the design flow area; and g represents the gravitational acceleration constant (e.g., 9.80665 m / s²). H represents the depth of the water.

[0112] The theoretical drainage capacity value characterizes the maximum water flow that a storm drain can theoretically pass through at the end of a certain cycle, based on the currently measured water depth and assuming that the corresponding storm drain is in a completely unobstructed state and there is no backflow in the downstream pipe network. Specifically, a larger theoretical drainage capacity value for a water collection unit at the end of a cycle indicates a deeper water depth at that time, or a larger designed flow area for the storm drain within the unit. Under ideal conditions, deeper water generates greater hydrostatic pressure, which should drive a larger flow rate. Therefore, for the same actual inflow rate, a larger theoretical drainage capacity value generally means a greater theoretical drainage potential.

[0113] In this embodiment, if the actual inflow rate is negative, the corresponding unit is directly determined to be in an abnormal state, and subsequent ratio calculations are skipped; if the actual inflow rate is positive, the sum of the actual inflow rate and the preset minimum positive number is calculated as the first sum; the ratio of the theoretical drainage capacity value to the first sum is calculated as the drainage obstruction index.

[0114] It should be noted that the specific value of the preset minimum positive number should be much smaller than the typical value of the actual inflow under normal drainage conditions. This embodiment does not impose a specific limitation. For example, the historical actual inflow can be calculated from historical data, and the statistical value of the historical actual inflow (such as the median or arithmetic mean) can be taken as the typical actual inflow. Then, a value that is two orders of magnitude smaller than the typical actual inflow can be taken as the minimum positive number.

[0115] It should be noted that if the actual inflow rate of a unit is negative, that unit is considered to be in an abnormal state, indicating that backflow is more likely to have occurred. In this case, to ensure the effectiveness of subsequent drainage obstruction index calculations, a preset maximum abnormal value should be assigned to the drainage obstruction index. The preset maximum abnormal value indicates that "theoretically there is drainage capacity, but the actual flow rate is not only zero, but also negative (backflow)," and its obstruction level far exceeds that of ordinary blockage or backflow, belonging to the highest level of drainage failure, which can be represented by a very large floating-point number.

[0116] The drainage obstruction index quantifies the reduction in actual drainage performance relative to the theoretically maximum possible performance at the current water depth. Specifically, if the drainage obstruction index of a catchment unit at the end of the current cycle is closer to 1, it indicates that the actual inflow is close to the theoretical drainage capacity, meaning the actual drainage flow has almost reached its theoretical maximum potential, and drainage is very smooth with virtually no obstruction. Conversely, if the drainage obstruction index of a catchment unit at the end of the current cycle is larger (significantly greater than 1), it indicates that the actual inflow is much smaller than the theoretical drainage capacity. This means that while theoretically there is the capacity to discharge a lot of water, very little actually enters the ground, indicating severe drainage obstruction, possibly due to blocked storm drains, downstream backwater, or a combination of both. If the drainage obstruction index of a catchment unit at the end of the current cycle is less than 1 (theoretically possible, but rare in practice), it indicates that the actual inflow exceeds the theoretical drainage capacity, which violates the physical principle that actual flow cannot exceed the maximum ideal free outflow, usually indicating a significant error in measurement or calculation.

[0117] S104: From the water collection units that meet the preset conditions, exclude the units with backflow indicators or drainage blockage indicators that exceed the preset extreme threshold to obtain the units to be monitored; generate drainage anomaly labels based on the differences in drainage blockage indicators between each unit to be monitored and its neighboring water collection units. The labels are used to indicate that the corresponding unit has local physical blockage or hydraulic backing.

[0118] It should be noted that the example values ​​for the preset extreme thresholds are determined through statistical analysis of historical data, and this embodiment does not impose specific limitations. For example, in a large amount of historical data or simulation tests, the typical range of drainage resistance indicators when the drainage system experiences severe blockage or backflow is observed, and the extreme threshold is set near the lower limit of this range (e.g., exceeding the 90th percentile). For instance, the preset extreme threshold is usually set to a relatively large value, typically 20.

[0119] It is understandable that the catchment unit with the backflow indicator has been identified as having experienced the most severe hydraulic back pressure caused by extremely high back pressure downstream, and a "hydraulic back pressure" label has been immediately generated for it.

[0120] It should be noted that units whose drainage resistance index exceeds the preset extreme threshold indicate that their actual drainage capacity is negligible. Although the direct cause may be complete blockage or severe backing, as a conservative approach in engineering, it is also regarded as suspected severe backing and a corresponding "hydraulic backing" label is generated.

[0121] It is important to understand that the different causes of urban flooding have drastically different spatial propagation characteristics. Localized physical blockages (such as debris covering) are random, isolated point faults that only affect a single storm drain. In contrast, hydraulic backwater (such as high water levels downstream) is a planar fault that propagates upstream along the pressure pipeline network, affecting multiple hydraulically connected storm drains in a cluster. Therefore, by analyzing the spatial relationship between the degree of drainage obstruction of a "monitored unit" with drainage anomalies and the degree of obstruction of its neighboring units directly connected through the underground pipe network, and further analyzing whether "the monitored unit itself is abnormally high while neighboring units are normal" or "both the monitored unit and its neighbors are abnormally high," a one-to-one mapping relationship can be established between the spatial propagation pattern and the fault mechanism in fluid dynamics (i.e., whether it is localized blockage or hydraulic backwater). This allows for the generation of diagnostic labels indicating the specific physical cause for each monitored catchment unit.

[0122] It should be noted that the neighboring water catchment unit is used to represent other water catchment units in the drainage network topology that have a direct or indirect hydraulic connection with the rainwater inlet node corresponding to the current water catchment unit being monitored through underground pipes.

[0123] Direct hydraulic connection: Two rainwater inlets are connected to the same pipe or directly connected through a pipe.

[0124] Indirect hydraulic connection: Two storm drains are connected by sharing a downstream main pipe or via a series of intermediate pipes and manholes, so that the water flow entering the pipe network from one storm drain may affect the pressure or flow rate at the other storm drain.

[0125] It should be noted that the core technology and data foundation upon which the search for "neighboring catchment units" relies—the Geographic Information System (GIS) for urban drainage networks and its network topology analysis function—are all existing mature technologies in this field, and this embodiment can be directly applied. Specifically, the GIS for urban drainage networks stores complete network node and pipeline connection data, and has built-in standard spatial analysis tools such as network tracing and connectivity analysis. By calling the topology connection table exported from the existing GIS, a hydraulic connectivity list for each stormwater inlet node is automatically established, thereby efficiently and accurately determining the set of "neighboring catchment units" for each catchment unit.

[0126] The process of generating drainage anomaly tags is as follows: Figure 3 As shown, it includes:

[0127] S104-1: Construct a hydraulic correlation matrix based on the drainage network topology. The hydraulic correlation matrix is ​​used to quantitatively describe the intensity of hydraulic influence between any two catchment units due to underground pipe connections. The more direct the underground connection between the two units and the shorter the path, the higher the intensity of the hydraulic influence.

[0128] The hydraulic correlation matrix specifically refers to the matrix of the intensity of hydraulic influence between catchment units (i.e., hydraulic correlation weights).

[0129] For example, each row and column of the matrix corresponds to a water catchment unit, and the matrix elements... This represents the intensity of the hydraulic influence of water collection unit j on unit i, where if the rainwater inlets corresponding to unit i and unit j are directly or indirectly hydraulically connected in the pipe network, then... A value greater than 0 indicates a more direct connection path, fewer pipes involved, and a shorter distance. The higher the value, the stronger the hydraulic influence. If there is no connection, then... It equals 0.

[0130] It should be noted that the matrix usually needs to be normalized (that is, each element in each row of the matrix is ​​divided by the sum of all elements in that row), so that the sum of all elements in each row is 1, making the matrix normalized. This can be interpreted as "the relative proportion of the hydraulic influence of unit j on unit i".

[0131] S104-2: Calculate the arithmetic mean and standard deviation of the drainage resistance index of all monitored units; for each monitored unit, obtain the standardized resistance index of the monitored unit based on the arithmetic mean and standard deviation; obtain other units that are hydraulically related to the monitored unit from the hydraulic correlation matrix, as well as the hydraulic influence intensity of all other units; use the hydraulic influence intensity as the hydraulic correlation weight.

[0132] The standardized stagnation index quantifies the degree to which the drainage stagnation of a monitored unit deviates from the current average stagnation level of all units. Specifically, a higher standardized stagnation index for a monitored unit indicates that its drainage stagnation is significantly higher than the overall average, directly suggesting a more severe degree of drainage obstruction in that unit. Therefore, the standardized stagnation index can be expressed by the following formula: Standardized Stagnation Index = .

[0133] Other units with hydraulic connections refer to the neighboring catchment units of the unit to be monitored.

[0134] For example, suppose there are four catchment units A, B, C, and D. Catchment unit A and catchment unit B are directly connected by pipe 1, with no other storm drains in between. Catchment unit A and catchment unit C are indirectly connected via the path A→B→C. Catchment unit A and catchment unit D are not connected by any pipes. The hydraulic association weight assignment principle between catchment unit A and its neighboring catchment units is as follows: Since catchment unit A and catchment unit B are directly connected, the hydraulic association weight between them will be assigned a higher value (e.g., set to 1). Since catchment unit A and catchment unit C are indirectly connected, with a longer pipe path, the hydraulic association weight between them will be lower than the value for a direct connection. Typically, the weight decreases as the number of pipes in the path increases. Therefore, the hydraulic association weight between catchment unit A and catchment unit C = ... p represents the attenuation coefficient; since there is no pipe connection between unit A and unit D, the hydraulic correlation weight will be assigned to zero.

[0135] It should be noted that the specific value of the attenuation coefficient is determined based on the physical damping characteristics of hydraulic waves propagating in the pipe network, and this embodiment does not impose a specific limitation. For example, engineering experience values ​​are usually used: p=1 indicates that the influence decreases linearly with the pipe length, which is a relatively mild assumption; p=2 simulates the case where the influence decreases sharply with the square of the number of pipes.

[0136] S104-3: For each unit to be monitored, multiply the standardized resistance index of other units by the hydraulic correlation weights of each other unit to obtain a weighted value; calculate the sum of the weighted values ​​of the unit to be monitored and all other units to obtain the neighbor average resistance index of the unit to be monitored.

[0137] The neighbor average obstruction index represents the average level of drainage anomalies among directly or indirectly connected units in the local hydraulic environment of the underground pipe network where a monitored unit is located. A higher neighbor average obstruction index indicates that neighboring units generally also experience high obstruction (i.e., the standardized obstruction indices of neighboring units are mostly positive). This suggests that the overall drainage of the local pipe network area where the monitored unit is located is poor, exhibiting a "contiguous obstruction" characteristic, a typical spatial manifestation of systemic hydraulic backwater (such as pressure waves propagating upstream due to high downstream water levels). Conversely, a lower neighbor average obstruction index indicates that the hydraulic neighbors of the monitored unit generally have relatively smooth or even abnormally smooth drainage (i.e., the standardized obstruction indices of neighboring units are mostly negative). This suggests that the overall drainage status of the local area where the monitored unit is located is better. If the standardized obstruction index of the monitored unit itself is very high, it further highlights its "isolated anomaly" characteristic, and is more likely to be a local blockage.

[0138] S104-4: Generate corresponding drainage anomaly labels based on standardized obstruction indices and neighboring average obstruction indices.

[0139] To accurately generate corresponding drainage anomaly labels, as an example, for each monitoring unit, if the standardized obstruction index of the monitoring unit is higher than a preset first threshold and the average obstruction index of its neighbors is not higher than a preset second threshold, a first drainage anomaly label is generated; wherein, the first drainage anomaly label is used to indicate that there is a local physical blockage in the space represented by the monitoring unit; if the standardized obstruction index of the monitoring unit is higher than the preset first threshold and the average obstruction index of its neighbors is also higher than the preset second threshold, a second drainage anomaly label is generated; wherein, the second drainage anomaly label is used to indicate that there is system hydraulic backlash in the space represented by the monitoring unit.

[0140] It should be noted that the specific values ​​of the preset first threshold and the preset second threshold are determined based on engineering experience, and this embodiment does not impose specific limitations. For example, by calculating the historical standardized hindrance index using historical data and analyzing the numerical distribution range of the standardized hindrance index (e.g., the 95th percentile), a typical range for the preset first threshold is between 1.0 and 2.0. The preset second threshold is often set to a value lower than the first threshold; for example, if the first threshold is equal to 1.5, the preset second threshold could be 0.6.

[0141] It should be noted that the first and second drainage anomaly labels are usually represented in the form of enumeration or integer codes. For example, a "fault type" status field is maintained for each water catchment unit; if it is determined to be a local physical blockage, the field value is set to "1" or the string "BLOCKAGE"; if it is determined to be hydraulic backlash, the field value is set to "2" or the string "BACKWATER"; if there is no anomaly, it is set to "0" or "NORMAL".

[0142] S105: Correct the drainage hydrodynamic model based on the label; use the corrected model to predict water accumulation evolution and generate early warning information.

[0143] It should be noted that the drainage hydrodynamic model of this invention refers to a computer simulation model that is constructed based on urban drainage network geographic information system (GIS) data and using mature numerical methods of fluid dynamics (such as solving the Saint-Venant equations), capable of simulating the water flow state within the network and the surface water accumulation process. The construction methods of such models (such as using professional software like SWMM, InfoWorks, and Mike Urban), parameter calibration, and verification processes are all existing, well-known, and mature technologies in the field of urban hydrology and hydrodynamics, and have been widely applied in urban planning, design, and risk assessment. These will not be elaborated upon in this embodiment and can be directly applied.

[0144] As an example, if the first drainage anomaly label is used, the first type of calibration operation is performed: the node in the model that represents the storm drain inlet of the corresponding catchment unit is located, and the inlet flow coefficient of the node is reduced to simulate the physical effect of the reduced flow area of ​​the storm drain in the model.

[0145] The original inlet flow coefficient is an empirical, dimensionless parameter used in hydrodynamic models to quantify the efficiency of surface runoff entering underground pipe networks through storm drains. It is not directly measured, but rather predetermined and input into the model during the model building and calibration phase, based on design specifications and experimental data.

[0146] It should be noted that the specific operation to reduce the inlet flow coefficient is as follows: obtain the original inlet flow coefficient of the node, divide it by the drainage resistance index of the water collection unit, and obtain the calibrated inlet flow coefficient value.

[0147] It should be noted that, to ensure the stability of the model's numerical calculations, the calculated calibrated inlet flow coefficient values ​​also need to be subject to minimum and maximum value constraints. For example, assuming the minimum constraint is 0.05 and the maximum constraint is 0.95, if the calibrated inlet flow coefficient value is greater than the maximum value, its value is forcibly set to 0.95; if the calibrated inlet flow coefficient value is less than the minimum value, its value is forcibly set to 0.05; if the calibrated inlet flow coefficient value is neither greater than the maximum value nor less than the minimum value, its original value is retained.

[0148] It should be noted that the specific values ​​of the minimum and maximum values ​​can be determined based on industry experience, and this embodiment does not impose any specific limitations.

[0149] As an example, if the second drainage anomaly label is used, a second type of calibration operation is performed: the downstream outlet node of the pipeline branch affected by the back pressure is located in the model, and the boundary water level condition of the node is increased to simulate the back pressure effect caused by the high downstream water level in the model.

[0150] The specific process of locating the downstream outlet node is as follows: All monitoring units that generate the second drainage anomaly label of "hydraulic backing" are mapped to their corresponding rainwater inlet model nodes; then, the common drainage path of these nodes is traced downstream along the pipeline topology until the first model node that converges these paths and discharges them to external water bodies (such as rivers, lakes) or the main system is found. This node is the downstream outlet node of the affected pipeline branch.

[0151] In drainage hydrodynamic models, boundary water level conditions refer to the external hydraulic constraints that must be specified to define the outlet of the computational domain. Specifically, for an outlet node, setting its boundary type to "constant head" means that the water level at that outlet node is forced to remain at a fixed value throughout the simulation.

[0152] It should be noted that the specific operation to improve the boundary water level condition is as follows: set the boundary type of the outlet node to "constant head"; obtain the ground elevation of the outlet node, add the current water depth of the upstream area of ​​the pipeline branch affected by it, and use the calculation result as the boundary water level value of the node.

[0153] It is important to understand that, to ensure the model can return to normal after troubleshooting or when diagnostic signals disappear, and to avoid prolonged operation with outdated calibration parameters, this invention introduces a timeliness and reset mechanism for model parameters. Specifically, each calibrated parameter (such as the inlet flow coefficient or downstream boundary water level) is associated with an effective duration timer. If no drainage anomaly tag triggering the calibration is detected again within several consecutive cycles (such as 3 cycles), the corresponding parameter will be automatically and gradually reset to its original design value.

[0154] It is important to understand that after the model is dynamically corrected, high-resolution future rainfall forecast data released by the meteorological department can be loaded immediately to drive the corrected drainage hydrodynamic model to perform rapid numerical simulation. Based on the current accurate reproduction of the actual situation (i.e., water level, flow rate, corrected parameters and boundary conditions), the model can deduce the spatiotemporal changes in surface water depth of each catchment unit over a period of time (e.g., 1 to 2 hours). If it is predicted that the water depth at a certain location will exceed the warning line, the early warning process will be automatically triggered, generating and releasing intelligent early warning information that includes the precise spatiotemporal location, the depth exceeding the standard, the diagnosed cause (local blockage or hydraulic backwater), and targeted treatment suggestions.

[0155] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0156] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for monitoring and early warning of urban flooding evolution based on multi-source data, characterized in that, The method includes: Periodically and synchronously acquire meteorological radar grid data and video water accumulation data, and process them to generate rainfall intensity and water accumulation depth for each pre-divided catchment unit at the end of each cycle; For each catchment unit, the actual inflow rate into the underground pipe network is determined by inversion based on its catchment area, the rainfall intensity and water depth at the end of the current cycle and the water depth at the end of the previous cycle; and whether a backflow indicator is generated for the catchment unit is identified. If the current rainfall intensity and water depth meet the preset conditions, the drainage resistance index will be determined by comparing the theoretical drainage capacity value determined based on the current water depth with the actual inflow rate. From the water catchment units that meet the preset conditions, exclude units with backflow indicators or drainage blockage indicators that exceed the preset extreme thresholds to obtain the units to be monitored. Obtain the neighboring water catchment units of each unit to be monitored. The neighboring water catchment units are used to represent other water catchment units in the drainage network topology that have direct or indirect hydraulic connection with the rainwater inlet node corresponding to the current water catchment unit through underground pipes. A hydraulic correlation matrix is ​​constructed based on the drainage network topology. The hydraulic correlation matrix is ​​used to quantitatively describe the intensity of hydraulic influence between any two catchment units due to the connection of underground pipelines. Calculate the arithmetic mean and standard deviation of the drainage resistance index of all monitored units; for each monitored unit, obtain the standardized resistance index based on the arithmetic mean and standard deviation; obtain other units that are hydraulically related to the monitored unit from the hydraulic correlation matrix, as well as the hydraulic influence intensity of all other units; use the hydraulic influence intensity as the hydraulic correlation weight. For each unit to be monitored, the standardized resistance index of other units is multiplied by the hydraulic correlation weight of each unit to obtain a weighted value; the sum of the weighted values ​​of the unit to be monitored and all other units is calculated as the neighbor average resistance index of the unit to be monitored. Based on standardized obstruction indices and neighboring average obstruction indices, corresponding drainage anomaly labels are generated. These labels are used to indicate that there is local physical blockage or hydraulic backing in the corresponding unit. The drainage hydrodynamic model is modified based on the labels; the modified model is used to predict water accumulation evolution and generate early warning information.

2. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 1, characterized in that, The process of dividing the pre-divided catchment units includes: Acquire geographic information system data of urban drainage pipe networks; Based on the geographic information system data of the urban drainage network, the spatial location of all rainwater inlet nodes is extracted. By combining digital elevation model and surface runoff path analysis, the corresponding surface catchment area is divided for each stormwater inlet node; theoretically, the surface runoff in each catchment area flows into the only stormwater inlet in the catchment area. Each rainwater inlet node and its corresponding surface catchment area are defined as a catchment unit; the catchment area of ​​the catchment unit is the horizontal projected area of ​​the corresponding surface catchment area.

3. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 2, characterized in that, The process of generating rainfall intensity and water depth at the end of each cycle for each catchment unit includes: Acquire radar reflectivity data published in grid format; and backtrack the raw water depth data collected within the current period at the end of the current period. The radar reflectivity data is converted into rainfall intensity data; the raw water depth data is then low-pass filtered to obtain the processed water depth data. For each catchment unit, the rainfall intensity allocated to the catchment unit is determined by interpolation calculation based on the spatial relationship between its geometric center location and the surrounding radar grid points. The arithmetic mean of all water depths in the processed water depth data is used as the water depth of the catchment unit at the end of the current cycle.

4. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 1, characterized in that, The process of determining the actual inflow rate includes: For each catchment unit, calculate the product of the water depth and the catchment area at the end of the current cycle as the current surface water volume; calculate the water depth and the catchment area at the end of the previous cycle as the surface water volume of the previous cycle. Calculate the change in current surface water volume compared to the previous period's surface water volume per unit time, and use this as the change in surface water volume. Calculate the product of the rainfall intensity at the end of the current cycle and the catchment area as the rainfall replenishment; subtract the change in surface water retention from the rainfall replenishment to obtain the theoretical inflow rate. Set the theoretical inflow rate as the actual inflow rate.

5. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 4, characterized in that, The process of identifying whether a backflow identifier is generated for a water catchment unit includes: If the theoretical inflow rate remains below the preset negative fault tolerance threshold within a preset time window, it is determined that backflow has occurred in the water collection unit, and a backflow flag is generated for it; wherein, the absolute value of the preset negative fault tolerance threshold is greater than zero.

6. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 4, characterized in that, The method further includes: For each catchment unit, if the current water depth is greater than the preset depth threshold, it is further determined whether the current rainfall intensity is greater than the rainfall intensity threshold; if the current rainfall intensity is greater than the rainfall intensity threshold, the catchment unit is determined to meet the preset conditions and proceeds to the subsequent calculation process. If the current water depth is greater than the preset depth threshold, and the current rainfall intensity is not greater than the rainfall intensity threshold, the catchment unit is determined to be in the post-rain dissipation period and enters the subsequent calculation process in the first mode. In the first mode, when calculating the drainage obstruction index, if the change in surface water level is negative, the absolute value of the change in surface water level is taken as the actual inflow; if the change in surface water level is positive, the process directly proceeds to the backflow judgment process. If the current water depth is not greater than the preset depth threshold, the water collection unit is determined not to meet the preset conditions, will not proceed to the subsequent calculation process, and the drainage obstruction index will be set to the preset default value.

7. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 6, characterized in that, The process for determining the theoretical drainage capacity value includes: Obtain the design flow area of ​​the rainwater inlet of the water collection unit, and obtain the preset reference flow coefficient used to characterize the water flow efficiency; Based on the water pressure generated by the current water depth, and combined with the design flow area of ​​the rainwater inlet and the preset benchmark flow coefficient, the maximum drainage flow of the rainwater inlet under ideal unobstructed operating conditions is calculated as the theoretical drainage capacity value.

8. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 7, characterized in that, The process for determining the drainage resistance index includes: If the actual inflow rate is negative, the corresponding unit is directly determined to be in an abnormal state, and subsequent ratio calculations are skipped. If the actual inflow rate is positive, calculate the sum of the actual inflow rate and the preset minimum positive number, and use it as the first sum value; The ratio of the theoretical drainage capacity value to the first sum value is calculated and used as a drainage resistance index.

9. The method for monitoring and early warning of urban flooding evolution based on multi-source data according to claim 1, characterized in that, The process of generating corresponding drainage anomaly labels based on standardized obstruction indices and neighboring average obstruction indices includes: For each monitoring unit, if the standardized obstruction index of the monitoring unit is higher than a preset first threshold and the average obstruction index of the neighbors is not higher than a preset second threshold, a first drainage anomaly label is generated; wherein, the first drainage anomaly label is used to indicate that there is a local physical blockage in the space represented by the monitoring unit. If the standardized resistance index of the monitored unit is higher than the preset first threshold, and the average resistance index of the neighboring unit is also higher than the preset second threshold, a second drainage anomaly label is generated; wherein, the second drainage anomaly label is used to indicate that there is system hydraulic backing in the space represented by the monitored unit.

Images