Method for judging urban inland inundation water accumulation degree based on dynamic adjustment of liquid position cell interval

Abstract

The invention discloses a method for judging the urban inland inundation degree based on a dynamic adjustment liquid position cell interval, and belongs to the technical field of urban inland inundation monitoring. The problem of dynamically and accurately judging the waterlogging water accumulation degree is solved. The method comprises the following steps: collecting multi-source real-time liquid level data; establishing triggering conditions of waterlogging judgment, including a sliding window triggering condition, an event triggering condition and a space partition triggering condition, and considering basic auxiliary triggering information; constructing a liquid position information interval basic model, and calculating a liquid position information interval; calculating a liquid position cell interval to obtain an adjusted dynamic liquid position cell interval, and executing the dynamic liquid position cell interval according to a sequence of an event triggering condition, a sliding window triggering condition and a space partition triggering condition; performing dynamic confidence interval driven ponding degree judgment to obtain a ponding degree judgment result; and outputting the corrected judgment result of the urban inland inundation water accumulation degree in the dynamic adjustment liquid position zone. According to the invention, the accuracy, real-time performance and pertinence of urban inland inundation ponding degree judgment are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of urban waterlogging monitoring, and particularly relates to a method for judging the waterlogging degree of urban waterlogging based on dynamic adjustment of the confidence interval of liquid level. BACKGROUND

[0002] With the acceleration of urbanization, the hardening rate of urban underlying surface is improved, and extreme rainfall events occur frequently. Urban waterlogging has become a prominent problem affecting the life of residents, traffic and the safety of municipal facilities. The current judgment of urban waterlogging degree mainly relies on three traditional methods, which have significant technical shortcomings.

[0003] 1. Static parameter modeling judgment: a fixed error parameter is used to construct a liquid level calculation model, without considering the fluctuation characteristics of the short-time sudden rise and fall of waterlogging liquid level, resulting in a large deviation in confidence interval calculation (error usually exceeds 15%), which cannot adapt to real-time liquid level changes;

[0004] 2. Single trigger adjustment method: only relying on single data fluctuation or event trigger adjustment, without considering the spatial heterogeneity of urban waterlogging (such as the difference in drainage between low-lying areas and highlands), the problem of "low-lying area adjustment lag, high area over-adjustment" occurs, and the judgment is not specific enough;

[0005] 3. Lack of closed-loop verification mechanism: after adjustment, the rationality of the confidence interval is not verified, or only simple data comparison is used for correction, which is difficult to find abnormal values of physical laws (such as the upper limit of confidence being higher than the terrain elevation), resulting in low reliability of early warning information;

[0006] 4. High dependence on manual assistance: some key parameters need to be adjusted manually, with a response lag of more than 30 minutes, which cannot meet the actual needs of "minute-level early warning" in waterlogging emergency scenarios.

[0007] Therefore, it is urgent to solve the problems of parameter fixation, one-sided adjustment, verification deficiency and response lag of traditional methods, and to provide precise and efficient technical support for urban waterlogging prevention and control. SUMMARY

[0008] The problem to be solved by the application is to realize dynamic and accurate judgment of waterlogging degree, and a method for judging the waterlogging degree of urban waterlogging based on dynamic adjustment of the confidence interval of liquid level is proposed.

[0009] To achieve the above purpose, the application realizes the following technical scheme:

[0010] A method for judging the waterlogging degree of urban waterlogging based on dynamic adjustment of the confidence interval of liquid level, comprising the following steps:

[0011] S1. Collecting multi-source real-time liquid level data, including collecting real-time liquid level data and historical liquid level data of multiple sites in a waterlogging monitoring area and preprocessing to obtain preprocessed multi-source real-time liquid level data;

[0012] S2. Establishing a trigger condition for waterlogging judgment, including a sliding window trigger condition, an event trigger condition, a spatial partition trigger condition, and considering basic auxiliary trigger information;

[0013] S3. Constructing a liquid position confidence interval basic model to calculate the liquid position confidence interval;

[0014] S4. Based on the preprocessed multi-source real-time liquid level data obtained in step S1, calculating the liquid position confidence interval for different trigger conditions of waterlogging judgment obtained in step S2, obtaining the adjusted dynamic liquid position confidence interval, and executing in the order of event trigger condition→ sliding window trigger condition→ spatial partition trigger condition;

[0015] S5. For the adjusted dynamic liquid position confidence interval obtained in step S4, performing waterlogging degree judgment driven by dynamic confidence interval to obtain waterlogging degree judgment result;

[0016] S6. Adjusting and verifying the waterlogging degree judgment result obtained in step S5 and correcting the abnormality to output the judgment result of the city waterlogging waterlogging degree of the adjusted dynamic adjustment liquid position confidence interval.

[0017] Further, the preprocessing method in step S1 includes removing jump values caused by sensor failure by 3σ principle, using linear interpolation to complete short-time data missing, and then uniformly converting the data to 2000 national geodetic coordinate system to obtain preprocessed multi-source real-time liquid level data.

[0018] Further, the specific implementation method of step S2 includes the following steps:

[0019] S2.1. Setting the sliding window trigger condition includes that the real-time liquid level preset time length change exceeds the liquid level fluctuation threshold, and the liquid level error standard deviation change rate in the preset time length rolling window exceeds the change rate threshold, wherein the error standard deviation is calculated based on the measured data of the preset group number in the window;

[0020] S2.2. Setting the event trigger condition to obtain sudden scene information by connecting multiple department systems, including the rainstorm warning level of the meteorological department, the blocked warning flow rate of the pipe network monitoring system being lower than the preset flow rate threshold, the topographic reconstruction notice of the municipal department, and the slope change caused by construction around the monitoring point exceeding the preset slope threshold;

[0021] S2.3. Set the spatial partition trigger condition as dividing the monitoring area into low-lying water area, ordinary flat area, and high ground drainage area based on terrain elevation data and drainage capacity, the slope of the low-lying water area is lower than the preset low-slope threshold, the slope of the ordinary flat area is between the preset low-slope threshold and the preset high-slope threshold, and the slope of the high ground drainage area is higher than the preset high-slope threshold, and set the liquid level rising speed threshold of each partition;

[0022] S2.4. Set the basic auxiliary trigger information to include synchronously acquiring land use type data and historical waterlogging event records, the land use type data including green land, road, and building land, and the historical waterlogging event records including waterlogging depth, duration, and disaster-bearing body information.

[0023] Further, in step S3, error modeling is first performed, then the basic confidence interval is calculated, and finally the sliding window parameter updating port, event parameter mapping calling port, and spatial partition threshold adapting port are set, and the specific implementation method includes the following steps:

[0024] S3.1. Error modeling includes systematic error and random error;

[0025] The systematic error is compensated by a preset period of field calibration to establish a compensation formula:

[0026]

[0027] wherein, is the actual liquid level, is the measured liquid level, is a fixed bias value;

[0028] The random error is based on the distribution characteristics of historical data, a preset sample size threshold n0, and a large sample with n≥n0 adopts a normal distribution, and a small sample with n

[0029] The probability density function of the large sample normal distribution is:

[0030]

[0031] wherein, σ is the standard deviation of the random error, is the random error;

[0032]

[0033] wherein, is the true liquid level mean value;

[0034] The random error of the small sample follows a t distribution with n-1 degrees of freedom, and the probability density function is:

[0035]

[0036] wherein, Γ(·) is the gamma function, n is the sample size;

[0037] S3.2. Constructing the basic confidence interval based on the random error distribution combined with the confidence level The formula is divided into two categories:

[0038] The basic confidence interval of large sample is:

[0039]

[0040] wherein, is the basic confidence interval, is the sample mean of the compensated actual liquid level; is the two-sided quantile of the standard normal distribution, and α is the significance level; is the initial standard deviation;

[0041] The basic confidence interval of small sample is:

[0042]

[0043] wherein, is the two-sided quantile of the t distribution, is the sample standard deviation of the compensated actual liquid level;

[0044] ;

[0045] S3.3. Setting the sliding window parameter update port to support real-time replacement of the standard deviation, the event parameter mapping calling port to support preset rule triggering, and the spatial partition threshold adaptation port to support differentiated parameter input for different partitions.

[0046] Further, the updating method of the dynamic liquid position confidence interval satisfying the sliding window triggering condition in step S4 includes first calculating the mean of the actual liquid level in the sliding window, and the calculation formula is:

[0047]

[0048] wherein, is the total number of sliding windows, is the actual liquid level of the i-th sliding window, is the mean of the actual liquid level in the sliding window;

[0049] Then calculate the sample standard deviation of the random error in the sliding window :

[0050] ;

[0051] Then the parameter substitution is performed, and the initial standard deviation of the basic model is replaced with the newly calculated standard deviation, and the confidence interval is updated synchronously.

[0052] Further, the updating method of the dynamic liquid position confidence interval meeting the event trigger condition in step S4 comprises adjusting the standard deviation and the system error compensation value based on an event-adjustment coefficient mapping table;

[0053] The adjustment formula of the standard deviation is

[0054]

[0055] wherein, is the initial standard deviation, is the adjustment coefficient corresponding to the event, is the adjusted standard deviation triggered by the event;

[0056] The adjustment of the system error compensation value is to increase a fixed compensation amount Δb on the basis of the basic compensation value b base .

[0057] Further, the updating method of the dynamic liquid position confidence interval meeting the spatial partition trigger condition in step S4 comprises adjustment of the low-lying water area, the highland drainage area, and the ordinary flat area.

[0058] The adjustment of the standard deviation of the low-lying water area is to multiply the current standard deviation by the partition adjustment coefficient k low ; and the adjustment of the upper limit of the confidence interval is to increase a preset increment value ΔCI low on the basis of the current upper limit of the confidence interval.

[0059] The adjustment of the standard deviation of the highland drainage area is to multiply the current standard deviation by the partition adjustment coefficient k high .

[0060] The ordinary flat area uses the current standard deviation and the confidence interval.

[0061] Further, the waterlogging degree judgment method driven by the dynamic confidence interval in step S5 is as follows.

[0062] Mild waterlogging: the mean value of the confidence interval is in a preset mild waterlogging mean value interval, and the confidence upper limit is ≤ a preset mild waterlogging upper limit threshold × a partition weight;

[0063] Moderate waterlogging: the mean value of the confidence interval is in a preset moderate waterlogging mean value interval, and the confidence upper limit is ≤ a preset moderate waterlogging upper limit threshold × a partition weight;

[0064] Severe waterlogging: the mean value of the confidence interval is > a preset severe waterlogging mean value threshold, or the confidence upper limit is > a preset severe waterlogging upper limit threshold × a partition weight.

[0065] Further, the specific implementation method of step S6 includes the following steps:

[0066] S6.1. Sliding window adjustment verification: the probability of the confidence interval covering the measured liquid level value after updating the sliding window is greater than or equal to the preset general area coverage probability threshold / the preset key area coverage probability threshold, and if it does not meet the standard, the window length is adjusted according to the preset rule to recalculate the standard deviation;

[0067] S6.2. Event mapping adjustment verification: compare the determination results before and after the event trigger with the actual waterlogging situation, and if the deviation rate is greater than the preset deviation rate threshold, the standard deviation adjustment coefficient in the mapping table is corrected;

[0068] S6.3. Spatial partition adjustment verification: cross-verification across partitions, and the model adjusted by the data of low-lying areas is used to verify the performance of ordinary areas, to ensure the migration of the partition adjustment rule, and if the coverage probability of a certain partition is less than the preset partition coverage probability threshold, the liquid level rising speed threshold of the partition is recalibrated;

[0069] S6.4. Abnormal correction: the adjustment result exceeding the physical law is replaced by the mean value of the confidence interval of a preset number of adjacent monitoring points in the same partition, or the adjustment logic of step S4 is triggered again.

[0070] The beneficial effects of the present application are:

[0071] The judgment method of the urban waterlogging water accumulation degree based on dynamic adjustment of liquid position confidence interval improves the determination accuracy; through the "three-dimensional dynamic adjustment + special verification" mechanism, the probability of the confidence interval covering the measured value meets the preset standard, the water accumulation level determination deviation rate is controlled within the preset range, and the method is significantly better than the traditional static model.

[0072] The judgment method of the urban waterlogging water accumulation degree based on dynamic adjustment of liquid position confidence interval enhances real-time responsiveness; the response time after triggering adjustment is less than or equal to the preset response time, the determination result is updated at a preset update frequency, the delay of early warning and pushing meets the preset delay requirement, and the emergency demand in the waterlogging burst scenario is met.

[0073] The judgment method of the urban waterlogging water accumulation degree based on dynamic adjustment of liquid position confidence interval strengthens the scene specificity; through spatial partition adaptation and event mapping rules, differential adjustment of "different regions and different events" is realized, and the adaptation problem of spatial heterogeneity and burst characteristics of urban waterlogging is solved.

[0074] The judgment method of the urban waterlogging water accumulation degree based on dynamic adjustment of liquid position confidence interval guarantees traceability of operation and maintenance; adjustment logs and verification results are recorded completely, data backtracking for more than a preset period of time is supported, optimization of adjustment rules and parameters is facilitated, and long-term operation stability of the system is improved. BRIEF DESCRIPTION OF DRAWINGS

[0075] Figure 1 A flow chart of a judgment method of urban waterlogging accumulation degree based on dynamic adjustment of liquid position signal interval according to the present application. DETAILED DESCRIPTION

[0076] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application, that is, the specific embodiments described are only a part of the embodiments of the present application, but not all the specific embodiments. The components of the specific embodiments of the present application generally described and shown in the drawings herein can be arranged and designed in various different configurations, and the present application can also have other embodiments.

[0077] Therefore, the detailed description of the specific embodiments of the present application provided below in the drawings is not intended to limit the scope of the claimed present application, but only represents selected specific embodiments of the present application. Based on the specific embodiments of the present application, all other specific embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of the present application.

[0078] In order to further understand the inventive content, characteristics and effects of the present application, the following specific embodiments are exemplified, and the drawings are attached Figure 1 The detailed description is as follows:

[0079] Example 1:

[0080] A judgment method of urban waterlogging accumulation degree based on dynamic adjustment of liquid position signal interval, comprising the following steps:

[0081] S1. Collecting multi-source real-time liquid level data, including collecting real-time liquid level data and historical liquid level data of multiple stations in a waterlogging monitoring area and preprocessing to obtain preprocessed multi-source real-time liquid level data;

[0082] Further, the preprocessing method in step S1 includes removing the jump value caused by sensor failure by 3σ principle, using linear interpolation to complete short-time data missing, and then uniformly converting the data into 2000 national geodetic coordinate system to obtain the preprocessed multi-source real-time liquid level data.

[0083] Further, the function of 3σ principle outlier elimination algorithm is to identify and remove the liquid level jump value caused by sensor failure, and the core is to determine a reasonable range based on the normal distribution characteristics of data.

[0084] Step and formula: ① Calculate the sample mean of liquid level data set (x ):

[0085] Suppose the original data of liquid level of a monitoring point in a period of time is y1, y2, … y m (m is the total amount of data, the number of data groups in the statistical period is determined according to the preset sampling frequency), then:

[0086]

[0087] Where y i is the liquid level measurement value at the i-th moment, is the average value of the liquid level in the period.

[0088] 2. Calculate the sample standard deviation (s y ):

[0089]

[0090] Where m−1 is the degree of freedom (the sample standard deviation needs to eliminate the bias of mean estimation), s y reflects the dispersion degree of liquid level data.

[0091] 3. Determine the threshold value of abnormal value: according to the 3σ principle, if a liquid level value y i satisfies the following conditions, it is determined as an abnormal value and is excluded:

[0092]

[0093] Application example: the liquid level value of a monitoring point at 10:15 is 28 cm, the average value of the period =15 cm, s y =4 cm, then the threshold range is

[0094] 15±12=[3,27] cm, 28 cm is out of range, and is determined as an abnormal value.

[0095] Further, the function of linear interpolation missing value completion algorithm: complete short-time data missing (such as sensor signal interruption), based on the effective data points before and after the missing period to construct a linear relationship. Steps and formulas: suppose the missing value corresponds to the moment t0, the previous effective data point is (t1, y1) (t1 < t0), and the next effective data point is (t2, y2) (t2 > t0), then the liquid level value y0 at the missing moment is calculated by linear interpolation:

[0096]

[0097] Where t1, t0, t2 are time (converted to the preset time measurement unit for calculation), y1, y2 are the effective liquid level values at the corresponding moments, and y0 is the completed missing liquid level value.

[0098] S2. Establish the trigger condition of waterlogging judgment, including sliding window trigger condition, event trigger condition, spatial partition trigger condition, and consider the basic auxiliary trigger information;

[0099] Further, the specific implementation method of step S2 includes the following steps:

[0100] S2.1. Set the sliding window trigger condition to include the real-time liquid level preset time length change exceeding the liquid level fluctuation threshold, and the liquid level error standard deviation change rate in the preset time length rolling window exceeding the change rate threshold, wherein the error standard deviation is calculated based on the measured data of the preset group number in the window;

[0101] S2.2. Set the event trigger condition to obtain the sudden scene information from the interface multi-department system, including the rainstorm warning level of the meteorological department, the blockage warning flow rate of the pipe network monitoring system lower than the preset flow rate threshold, the terrain reconstruction notice of the municipal department, and the slope change caused by the construction around the monitoring point exceeding the preset slope threshold;

[0102] S2.3. Set the spatial partition trigger condition to divide the monitoring area into low-lying water area, ordinary flat area, and high ground drainage area based on the terrain elevation data and drainage capacity, the slope of the low-lying water area is lower than the preset low-slope threshold, the slope of the ordinary flat area is between the preset low-slope threshold and the preset high-slope threshold, and the slope of the high ground drainage area is higher than the preset high-slope threshold, and set the liquid level rising speed threshold of each partition;

[0103] S2.4. Set the basic auxiliary trigger information to include synchronously obtaining land use type data and historical waterlogging event records, the land use type data includes green land, road, and building land, and the historical waterlogging event records include water accumulation depth, duration, and disaster-bearing body information.

[0104] S3. Build a liquid position confidence interval model to calculate the liquid position confidence interval;

[0105] Further, in step S3, first, error modeling is performed, then the basic confidence interval is calculated, and finally, the sliding window parameter update port, the event parameter mapping calling port, and the spatial partition threshold adaptation port are set, and the specific implementation method includes the following steps:

[0106] S3.1. Error modeling includes systematic error and random error;

[0107] The systematic error is compensated by a preset period of field calibration to establish a compensation formula:

[0108]

[0109] Wherein, is the actual liquid level, is the measured liquid level, is a fixed bias value;

[0110] The random error is based on the distribution characteristics of historical data, a preset sample size threshold n0, and a normal distribution for large samples with n≥n0, and a t distribution for small samples with n

[0111] The probability density function of the normal distribution of large samples is:

[0112]

[0113] where σ is the standard deviation of the random error, is the random error;

[0114]

[0115] where, is the true liquid level mean;

[0116] The random error of small samples follows a t distribution with n-1 degrees of freedom, and the probability density function is:

[0117]

[0118] where Г(·) is the gamma function, and n is the sample size;

[0119] S3.2. Based on the distribution of random errors, combined with the confidence level, the basic confidence interval is constructed The formula is divided into two categories:

[0120] The basic confidence interval for large samples is:

[0121]

[0122] where, is the basic confidence interval, is the sample mean of the compensated actual liquid level; is the two-sided quantile of the standard normal distribution, and α is the significance level; is the initial standard deviation;

[0123] The basic confidence interval for small samples is:

[0124]

[0125] where, is the two-sided quantile of the t distribution, is the sample standard deviation of the compensated actual liquid level;

[0126] ;

[0127] S3.3. Set the sliding window parameter update port to support real-time standard deviation replacement, the event parameter mapping call port to support preset rule triggering, and the spatial partition threshold adaptation port to support differentiated parameter input for different partitions.

[0128] S4. Based on the preprocessed multi-source real-time liquid level data obtained in step S1, calculate the liquid level confidence interval for different triggering conditions of waterlogging judgment obtained in step S2, obtain the adjusted dynamic liquid level confidence interval, and execute in the order of event triggering condition, sliding window triggering condition, and spatial partition triggering condition.

[0129] Further, the updating method of the dynamic liquid level confidence interval that meets the sliding window triggering condition in step S4 includes first calculating the mean value of the actual liquid level in the sliding window, and the calculation formula is:

[0130]

[0131] wherein, is the total number of sliding windows, is the actual liquid level of the i-th sliding window, is the mean value of the actual liquid level in the sliding window;

[0132] Then calculate the sample standard deviation of random error in the sliding window :

[0133] ;

[0134] Then perform parameter replacement, replace the initial standard deviation of the basic model with the newly calculated standard deviation, and update the confidence interval synchronously. For example, replace the current standard deviation (such as the event-adjusted σ window ) of the basic model with σ event , and update the confidence interval:

[0135] ;

[0136] Further, the triggering condition is to meet any condition of the “sliding window triggering information”;

[0137] Further, the updating method of the dynamic liquid level confidence interval that meets the event triggering condition in step S4 includes adjusting the standard deviation and the system error compensation value based on the event-adjustment coefficient mapping table;

[0138] The adjustment formula for the standard deviation is

[0139]

[0140] wherein, is the initial standard deviation, is the adjustment coefficient corresponding to the event, the adjusted standard deviation after the event trigger;

[0141] The adjustment of the system error compensation value is for pipe network blockage early warning, and the basic compensation value b base is increased by a fixed compensation amount Δb to obtain

[0142]

[0143] wherein Δb is an additional compensation amount corresponding to a specific event, b event is the system error compensation value after the event trigger.

[0144] Further, the event-adjustment coefficient mapping table is shown in Table 1:

[0145] Table 1

[0146]

[0147] Further, the updating method of the dynamic liquid level confidence interval that meets the spatial partition trigger condition in step S4 includes adjustment of low-lying water accumulation areas, highland drainage areas, and ordinary flat areas.

[0148] The adjustment of the standard deviation of the low-lying water accumulation area is to multiply the current standard deviation by the partition adjustment coefficient k low ; the upper limit of the confidence interval is adjusted by increasing the preset increment value ΔCI low to the current upper limit of the confidence interval.

[0149] The adjustment of the standard deviation of the highland drainage area is to multiply the current standard deviation by the partition adjustment coefficient k high .

[0150] The ordinary flat area uses the current standard deviation and confidence interval.

[0151] Further, the formula is as follows:

[0152] (1) Low-lying water accumulation area (slope lower than the preset low slope threshold, liquid level rising speed exceeding the corresponding partition threshold)

[0153] ① Standard deviation adjustment: multiply the current standard deviation by the partition adjustment coefficient k low :

[0154]

[0155] ② Confidence interval upper limit adjustment: increase the preset increment value ΔCI low to the current upper limit of the confidence interval.

[0156]

[0157] (2) High drainage area (slope is higher than the preset high slope threshold, liquid level rising speed exceeds the corresponding partition threshold) standard deviation adjustment: multiply the current standard deviation by the partition adjustment coefficient k high :

[0158]

[0159] (3) General flat area (slope between the preset low slope threshold and the preset high slope threshold) no additional adjustment, use the current standard deviation and confidence interval:

[0160]

[0161] ;

[0162] S5. Adjust the dynamic liquid position confidence interval obtained in step S4 to obtain the waterlogging degree judgment result.

[0163] Further, the dynamic parameter input: obtain the adjusted confidence interval mean, confidence upper limit, confidence level and monitoring point partition identifier;

[0164] Further, the partition weight setting: according to the regional risk sensitivity, the low-lying area, the general area and the high area are respectively set with the corresponding partition weight ;

[0165] Further, the dynamic confidence interval driven waterlogging degree judgment method of step S5 is:

[0166] Mild waterlogging: the confidence interval mean is in the preset mild waterlogging mean interval and the confidence upper limit is less than or equal to the preset mild waterlogging upper limit threshold multiplied by the partition weight;

[0167] Moderate waterlogging: the confidence interval mean is in the preset moderate waterlogging mean interval and the confidence upper limit is less than or equal to the preset moderate waterlogging upper limit threshold multiplied by the partition weight;

[0168] Severe waterlogging: the confidence interval mean is greater than the preset severe waterlogging mean threshold or the confidence upper limit is greater than the preset severe waterlogging upper limit threshold multiplied by the partition weight.

[0169] S6. Adjust the waterlogging degree judgment result obtained in step S5 to obtain the judgment result of the urban waterlogging degree of the adjusted dynamic liquid position confidence interval.

[0170] Further, the specific implementation method of step S6 includes the following steps:

[0171] S6.1. Sliding window adjustment verification: the probability of the confidence interval covering the measured liquid level value after updating the statistical sliding window is ≥ the preset ordinary area coverage probability threshold / the preset key area coverage probability threshold. If it does not meet the standard, adjust the window length according to the preset rules and recalculate the standard deviation.

[0172] S6.2. Event mapping adjustment verification: compare the determination results before and after the event trigger with the actual waterlogging situation. If the deviation rate is > the preset deviation rate threshold, modify the standard deviation adjustment coefficient in the mapping table.

[0173] S6.3. Spatial partition adjustment verification: cross-verification across partitions, use the model adjusted by the data of low-lying areas to verify the performance of ordinary areas, and ensure the transferability of the partition adjustment rules. If the coverage probability of a certain partition is < the preset partition coverage probability threshold, recalibrate the liquid level rise rate threshold of that partition.

[0174] S6.4. Abnormal correction: replace the confidence interval mean of a preset number of adjacent monitoring points in the same partition or re-trigger the adjustment logic of step S4 for adjustment results that exceed the physical law.

[0175] Further, dynamic visualization: generate "event-triggered historical process line", "corrected monitoring data and early warning diagnosis information", "visualized water accumulation map and risk periphery prompt";

[0176] Hierarchical early warning push: push the adapted early warning information to different partitions, and the push frequency is synchronized with the adjustment frequency of S4;

[0177] Standardized interface: provide a JSON format data interface containing key parameters of the adjustment method and final determination results for the city flood control dispatch platform to call, and the interface update frequency is consistent with the liquid level collection frequency;

[0178] Adjustment log storage: automatically record the trigger conditions, adjustment method, parameters before and after adjustment, and verification results of each adjustment, support traceability and review.

[0179] Further, in the embodiment, the system is deployed on the municipal flood control cloud platform, and the hardware configuration is: CPU Intel Xeon Gold 6420 (2 cores), memory 128GB, hard disk 4TB SSD; the software environment is: operating system Linux CentOS9.0, model calculation engine uses Python (depends on NumPy, Pandas library), visualization uses ECharts, and early warning push is realized through SMS gateway +APP. The 7 modules are integrated through micro-service architecture, and the data transmission between modules uses Kafka message queue to ensure real-time and stability.

[0180] The actual application examples of the embodiment are as follows:

[0181] Real-time data: 30 liquid level sensors (L1-L30) were deployed in the target area with a sampling frequency of 1 / 5 minutes. Real-time liquid level data was collected from 9:00 to 13:00 on July 20, 2024. At L15 station (low-lying area), the liquid level rose by 4cm from 9:30 to 9:35. The latitude, longitude, 85 elevation reference, and sensor signal strength (all ≥-70dBm, normal operation) were recorded simultaneously.

[0182] Historical data: Liquid level data from 2021 to 2023 was obtained from the municipal meteorological bureau and municipal departments, including different rainfall intensities and waterlogging levels. The average initial standard deviation of the low-lying area under heavy rain scenarios was 1.2cm.

[0183] Data preprocessing: The "3σ rule" was used to remove the abnormal value at L8 station at 10:15 (liquid level rising by 8cm, exceeding the reasonable range). Linear interpolation was used to complete the data of L7 and L9 stations. The data was converted to the national 2000 coordinate system to form a standardized data set.

[0184] Dynamic adjustment trigger information acquisition;

[0185] Sliding window trigger information: The liquid level change of L15 station from 9:30 to 9:35 was 4cm (>3cm), triggering the sliding window adjustment.

[0186] Event trigger information: At 10:00, the city meteorological bureau issued a heavy rain red alert (predicted 24-hour rainfall>100mm), and the flow rate of 3 pipe network monitoring points was less than 0.2m / s (pipe network blockage warning).

[0187] Spatial partition trigger information: Based on the DEM data (precision 1:2000) obtained by unmanned aerial vehicle survey in January 2024, the target area was divided into low-lying water area (including L15, L16, etc. 10 stations), ordinary flat area (including L5, L6, etc. 12 stations), and highland drainage area (including L25, L26, etc. 8 stations). The liquid level rising speed threshold of low-lying area was set to 2cm / 10 minutes.

[0188] Basic auxiliary information: Extract the 2024 national space planning data. The road area ratio in the target area is 32%, the building land is 38%, the green land is 25%, and the water body is 5%. The same period of heavy rain waterlogging event record in 2023 was obtained synchronously (the maximum water depth in the low-lying area was 42cm).

[0189] Liquid position interval basic model construction;

[0190] Error modeling: Establish system error compensation formula (actual liquid level = measured liquid level - 0.3 cm) through on-site calibration on July 1, 2024; based on historical data verification, random error conforms to normal distribution (n = 120 > 30, large sample);

[0191] Basic confidence interval calculation: Set ordinary area confidence level 90%, key area (including 3 schools and 2 hospitals) 95%, initial standard deviation: ordinary area 1.0 cm, low-lying area 1.2 cm, high area 0.8 cm, initial formula: confidence interval = liquid level mean ± 1.96 × initial standard deviation (Zα / 2 = 1.96, corresponding to 95% confidence level);

[0192] Adjustment interface reservation: Enable sliding window parameter update port, event parameter mapping call port, and spatial partition threshold adaptation port to ensure seamless connection with Module 4.

[0193] Dynamic adjustment;

[0194] Event-triggered adjustment: Receive heavy rain red alert at 10:00, adjust according to mapping table rules: initial standard deviation × 1.8 (low-lying area 1.2 cm × 1.8 = 2.16 cm), confidence level of the whole area is raised to 95%; At the same time, receive pipe network blockage warning, system error compensation coefficient + 0.5 cm (adjusted compensation formula: actual liquid level = measured liquid level - 0.3 cm + 0.5 cm = measured liquid level + 0.2 cm);

[0195] Sliding window adjustment: L15 station triggers sliding window condition, constructs 9:35-10:05 30-minute rolling window (including 6 groups of data), calculates liquid level error standard deviation σ = 1.6 cm, replaces event-adjusted standard deviation (2.16 cm → 1.6 cm), and updates confidence interval;

[0196] Spatial partition adaptation: L15 station belongs to low-lying area, liquid level rising speed 4 cm / 10 minutes (> 2 cm / 10 minutes), triggers partition adjustment: standard deviation 1.6 cm × 1.1 = 1.76 cm, confidence interval upper limit + 1.0 cm;

[0197] Collaborative logic: Execute in the order of "event trigger → sliding window → spatial partition", final L15 station adjusted standard deviation 1.76 cm, confidence level 95%, system error compensation coefficient + 0.2 cm.

[0198] Water accumulation degree determination;

[0199] Dynamic parameter input: After adjustment, the mean confidence interval of station L15 is 32cm, the upper confidence limit is 38cm, and the zone is identified as a low-lying area (weight 1.2).

[0200] Judgment process: 31.2cm≤45×1.2=54cm, and the average value of 32cm is within the range of 15-35cm, so it is judged as moderate waterlogging.

[0201] Multiple methods were used to adjust, verify, and correct the effects.

[0202] Validation of sliding window adjustment: After updating the sliding window at station L15, the probability of the confidence interval covering the measured value was 93% (≥95% key area standard, not met). When the window was reduced to 20 minutes (4 sets of data) and recalculated, the standard deviation was 1.8cm, and the coverage probability increased to 96%.

[0203] Event mapping adjustment verification: Comparing the judgment results of 10 low-lying area stations under the red rainstorm warning with the manual patrol data, the accuracy rate of severe waterlogging judgment is 94% (≥92%, meeting the standard).

[0204] Spatial zoning adjustment verification: The model adjusted using low-lying area data was used to verify the performance of L6 stations in ordinary areas, with a coverage probability of 91% (≥88%, meeting the standard).

[0205] Anomaly correction: No adjustment results that exceed the laws of physics, no additional correction is required.

[0206] Dynamic determination result output and application;

[0207] Dynamic visualization: Generates an "Event Trigger Change Timeline" (weather rainfall changes from light rain to moderate rain to heavy rain), "Corrected Monitoring Data and Early Warning Diagnostic Information" (Early warning information: Flooding warning has been issued under the Hongxing East Road Expressway Bridge. The current water level is 31.2cm, the threshold is 32.5cm, and the surface water is 45mm. This has exceeded the threshold by 10% and will affect the area under the Hongxing East Road Expressway Bridge. Please dispatch personnel to handle the situation promptly), and "Visualized Flooding Map and Risk Surrounding Area Warnings" (The map uses colors to distinguish between flooding, with red indicating flooding, and provides risk area warnings for key users in the surrounding area).

[0208] Tiered early warning push: Send early warning information to businesses and health service centers in low-lying areas: "Roads are flooded, please detour", push frequency 1 time / 5 minutes;

[0209] Standardized Interface: The API interface is now online, returning JSON format data, including the standard deviation of 1.8cm within the window, the event type "Red Rainstorm Warning + Urban Flooding", the zone identifier "Low-lying Area", and the judgment result "Severe Urban Flooding", for navigation apps to use;

[0210] Adjust log storage: Record the triggering conditions, adjustment method (event mapping + sliding window + spatial partitioning), parameters before and after adjustment (standard deviation 1.2cm → 1.8cm) and verification results (coverage probability 96%). The log retention period is set to 3 years.

[0211] In this embodiment, the real-time judgment results on July 29, 2025, were compared with the actual waterlogging situation: the hit rate for mild waterlogging areas was 93%, the hit rate for moderate waterlogging areas was 94%, and the hit rate for severe waterlogging areas was 97%. The time for the early warning information to be pushed to relevant departments was less than 3 minutes, which provided 1.5 hours of preparation time for traffic management and emergency rescue. No major traffic congestion or safety accidents caused by water accumulation occurred.

[0212] Statistics show that the confidence interval determined by the system covers the measured value on average 92.5%, the deviation rate of water accumulation level determination is only 6.8%, and the response delay is ≤5 seconds, which fully meets the actual needs of urban waterlogging prevention and control, and verifies the effectiveness and practicality of this method and system.

[0213] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0214] Although this application has been described above with reference to specific embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of this application. In particular, as long as there is no structural conflict, the features in the specific embodiments disclosed in this application can be combined with each other in any way. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, this application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals, characterized in that, Includes the following steps: S1. Collect multi-source real-time liquid level data, including real-time liquid level data and historical liquid level data from multiple stations within the urban flood monitoring area, and preprocess the data to obtain preprocessed multi-source real-time liquid level data; S2. Establish triggering conditions for waterlogging judgment, including sliding window triggering conditions, event triggering conditions, and spatial partitioning triggering conditions, and consider basic auxiliary triggering information; S3. Construct a basic model of the liquid position confidence interval and calculate the liquid position confidence interval; S4. Based on the preprocessed multi-source real-time liquid level data obtained in step S1, calculate the liquid level information interval for different triggering conditions of the waterlogging judgment obtained in step S2, and obtain the adjusted dynamic liquid level information interval. Execute the following sequence: event triggering condition → sliding window triggering condition → spatial partitioning triggering condition. S5. For the adjusted dynamic liquid position confidence interval obtained in step S4, perform dynamic confidence interval driven water accumulation degree judgment to obtain water accumulation degree judgment result; S6. Adjust, verify, and correct any anomalies in the water accumulation level judgment results obtained in step S5, and output the corrected judgment results of the urban waterlogging level in the dynamically adjusted liquid position information interval.

2. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 1, characterized in that, The preprocessing method in step S1 includes removing jump values ​​caused by sensor failures using the 3σ principle, using linear interpolation to fill in short-term data gaps, and then uniformly converting the data to the 2000 National Geodetic Coordinate System to obtain preprocessed multi-source real-time liquid level data.

3. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 2, characterized in that, The specific implementation method of step S2 includes the following steps: S2.

1. Set the sliding window trigger conditions, including the real-time liquid level change over a preset time exceeds the liquid level fluctuation threshold, and the change rate of the standard deviation of the liquid level error within the preset time scrolling window exceeds the change rate threshold, wherein the standard deviation of the error is calculated based on the measured data of a preset number of groups within the window; S2.

2. Set the event triggering conditions to obtain emergency scenario information by connecting to the multi-department system, including the rainstorm warning level of the meteorological department, the blockage warning flow rate of the pipeline monitoring system being lower than the preset flow rate threshold, the terrain modification notice of the municipal department, and the slope change caused by construction around the monitoring point exceeding the preset slope threshold. S2.

3. Set the spatial zoning trigger condition to be based on terrain elevation data and drainage capacity, divide the monitoring area into low-lying water accumulation area, ordinary flat area, and high-lying drainage area. The slope of the low-lying water accumulation area is lower than the preset low slope threshold, the slope of the ordinary flat area is between the preset low slope threshold and the preset high slope threshold, and the slope of the high-lying drainage area is higher than the preset high slope threshold. Set the liquid level rise rate threshold for each zone. S2.

4. Set basic auxiliary trigger information including synchronously acquiring land use type data and historical waterlogging event records. Land use type data includes green space, roads, and building land. Historical waterlogging event records include water depth, duration, and disaster-bearing body information.

4. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 3, characterized in that, Step S3 first performs error modeling, then calculates the basic confidence interval, and finally sets the sliding window parameter update port, event parameter mapping call port, and spatial partition threshold adaptation port. The specific implementation method includes the following steps: S3.

1. Error modeling includes systematic errors and random errors; System errors are compensated for using a formula established through on-site calibration at preset intervals. ； in, This represents the actual liquid level. To measure the liquid level, This is a fixed deviation value; Random errors are tested based on historical data to examine distribution characteristics. A preset sample size threshold n0 is used. Large samples with n≥n0 adopt a normal distribution, while small samples with n<n0 adopt a t-distribution. The probability density function of a large sample normal distribution is: ； Where σ is the standard deviation of the random error. This is random error; ； in, This represents the average actual liquid level. The random error of a small sample follows a t-distribution with n-1 degrees of freedom, and its probability density function is: ； Where Г(·) is the gamma function and n is the sample size; S3.

2. Based on the random error distribution and combined with the confidence level, construct the basic confidence interval. The formulas are divided into two categories: The baseline confidence interval for a large sample is: ； in, Based on the basic confidence interval, The average value of the actual liquid level after compensation; α represents the two-tailed quantile of the standard normal distribution, and α is the significance level. The initial standard deviation; The baseline confidence interval for a small sample is: ； in, These are the two-tailed quantiles of the t-distribution. To compensate for the sample standard deviation of the actual liquid level; ； S3.

3. Set the sliding window parameter update port to support real-time replacement of standard deviation, the event parameter mapping call port to support preset rule triggering, and the spatial partition threshold adaptation port to support partition differential parameter input.

5. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 4, characterized in that, The method for updating the dynamic liquid level information interval that meets the sliding window triggering condition in step S4 includes first calculating the average value of the actual liquid level within the sliding window, using the following formula: ； in, This represents the total number of sliding windows. Let i be the actual liquid level of the i-th sliding window. This represents the average actual liquid level within the sliding window; Then calculate the sample standard deviation of the random error within the sliding window. : ； Then, parameter replacement is performed, replacing the initial standard deviation of the base model with the newly calculated standard deviation, and the confidence interval is updated synchronously.

6. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 5, characterized in that, The method for updating the dynamic liquid position information interval that meets the event triggering conditions in step S4 includes adjusting the standard deviation and the system error compensation value based on the event-adjustment coefficient mapping table. The adjustment formula for standard deviation is: ； in, The initial standard deviation, This is the adjustment factor corresponding to the event. The adjusted standard deviation for the event trigger; The system error compensation value is adjusted to target pipeline blockage early warning, based on the basic compensation value b. base A fixed compensation amount △b is added to the base.

7. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 6, characterized in that, The update method for the dynamic liquid position information interval that meets the spatial partition triggering conditions in step S4 includes the adjustment of low-lying water accumulation area, high-lying drainage area, and ordinary flat area; The standard deviation of low-lying, waterlogged areas is adjusted by multiplying the current standard deviation by the zoning adjustment factor k. low The upper limit of the confidence interval is adjusted to be the current upper limit of the confidence interval plus a preset increment value ΔCI. low ; The standard deviation of the highland drainage area is adjusted by multiplying the current standard deviation by the zoning adjustment factor k. high ; The standard deviation and confidence interval for the ordinary flat region will continue to be used.

8. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 7, characterized in that, The dynamic confidence interval-driven method for determining the degree of water accumulation in step S5 is as follows: Mild flooding: The confidence interval mean is within the preset mild flooding mean interval and the upper confidence limit is ≤ the preset mild flooding upper limit threshold × the partition weight; Moderate flooding: The confidence interval mean is within the preset moderate flooding mean interval and the upper confidence limit is ≤ the preset moderate flooding upper limit threshold × the partition weight; Severe flooding: Confidence interval mean > preset severe flooding mean threshold or confidence upper limit > preset severe flooding upper limit threshold × zone weight.

9. The method for determining the degree of urban flooding based on dynamically adjusted liquid position information intervals according to claim 8, characterized in that, The specific implementation method of step S6 includes the following steps: S6.

1. Sliding window adjustment verification: After the sliding window is updated, the probability that the confidence interval covers the measured liquid level value is greater than or equal to the preset coverage probability threshold for ordinary areas / the preset coverage probability threshold for key areas. If the threshold is not met, the window duration is adjusted according to the preset rules and the standard deviation is recalculated. S6.

2. Event Mapping Adjustment Verification: Compare the judgment results before and after the event is triggered with the actual flooding situation. If the deviation rate is greater than the preset deviation rate threshold, adjust the standard deviation adjustment coefficient in the mapping table. S6.

3. Spatial zoning adjustment verification: Cross-zoning cross-validation, using the low-lying area data adjustment model to verify the performance of the ordinary area, ensuring the transferability of zoning adjustment rules. If the coverage probability of a certain zoning is less than the preset zoning coverage probability threshold, the liquid level rise rate threshold of that zoning is recalibrated. S6.

4. Anomaly Correction: For adjustment results that exceed physical laws, replace them with the mean of the confidence interval of a preset number of adjacent monitoring points in the same zone, or re-trigger the adjustment logic of step S4.

Images