A flood event intelligent identification method based on statistical threshold and multi-condition joint constraint

Abstract

The application discloses a flood stage intelligent identification method based on statistical threshold and multi-condition joint constraint, and comprises the following steps: step 1, data collection; step 2, flood discrimination threshold calculation; step 3, flood rising point identification based on gradient analysis; step 4, flood falling point determination; step 5, effective flood stage identification based on multi-condition joint constraint determination; step 6, flood parameter calculation; and step 7, visual result output. The application adopts a statistical method based on mean value and standard deviation to calculate the flood discrimination threshold, can automatically adapt to the hydrological characteristics of different river basins, adopts a multi-condition joint constraint determination method to identify the effective flood stage, integrates the water level and flow synchronism test into the constraint condition, and ensures the scientificity and reliability of the identification result; and the method can be applied to different types of hydrological stations, supports batch automatic processing, and greatly improves the work efficiency.

Description

Technical Field

[0001] This invention belongs to the field of disaster prevention and mitigation technology, and in particular relates to an intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints. Background Technology

[0002] Accurate identification and screening of flood events are the core foundation for hydrological analysis, flood forecasting, reservoir optimization, and flood control and disaster reduction decision-making. With the rapid development of hydrological monitoring networks, hydrological stations have accumulated massive amounts of high temporal resolution observation data, posing a serious challenge to traditional artificial flood event identification methods (including fixed threshold methods and single-condition discrimination methods).

[0003] The existing methods have the following problems: (1) The fixed threshold method cannot adapt to the differences in hydrological characteristics of different basins and the seasonal changes of the same basin. It generates a large number of misidentifications during the dry season and may miss small and medium floods during the wet season; (2) The single condition discrimination method is difficult to distinguish between real floods and abnormal situations such as data noise and instrument failure; (3) The identification of the rising point and the falling point lacks mathematical theoretical support and relies more on empirical thresholds; (4) Flood verification mostly uses simple logic and operation, and lacks consideration of the differences in the importance of each factor; (5) The accuracy of flood total calculation is insufficient and it is difficult to meet the needs of refined water resources management.

[0004] Therefore, there is an urgent need for an intelligent flood field identification method that integrates multidisciplinary theories, is highly adaptive, and has high scientific rigor, in order to solve the above-mentioned technical problems. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints, so as to solve the above-mentioned technical problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] This invention discloses an intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints. The method includes the following steps:

[0008] Step 1: Data Collection: Collect time-series observation data from hydrological stations in the target area, including station name, observation time t, flow rate Q(t), water level Z(t), and catchment area A, and arrange them in chronological order to form an ordered dataset;

[0009] Step 2: Calculation of flood discrimination threshold: Calculate the flood discrimination threshold based on the statistical characteristics of the flow time series. At a certain observation time Traffic satisfaction At that time, the moment Mark the moments exceeding the threshold; merge temporally consecutive or adjacent moments exceeding the threshold to form potential flood periods; denote the first moment exceeding the threshold in each potential flood period as... The last time the threshold was exceeded was ;

[0010] Step 3: Identification of flood initiation point based on gradient analysis: using the first threshold exceeding the threshold time during the potential flood period. Starting from the beginning of the search, the search proceeds backward along the time series, i.e., in the direction of decreasing time, and gradient analysis is used to locate the starting point of the flood. ;

[0011] Step 4, Determining the flood receding point: First, at the flood's starting point... In the subsequent period, find the moment corresponding to the maximum flow rate and record it as the peak flow moment. :

[0012] (7)

[0013] In the formula: The flow rate at the peak of the flood; This refers to the last time the threshold is exceeded, which is also the end of the potential flood period.

[0014] Then, at the peak of the flood Starting from the beginning of the search, the search proceeds progressively along the time series in the future direction, i.e., in increasing time, to determine the receding point of the flood. ;

[0015] Step 5: Identification of valid flood events based on multi-condition joint constraint judgment: Perform multi-condition joint constraint judgment on the identified potential flood periods. When all constraints are met simultaneously, the potential flood period is judged as a valid flood event; otherwise, it is rejected.

[0016] The joint constraint determination of multiple conditions uses the logical AND operation:

[0017] (8)

[0018] In the formula: For effective flood events; For duration constraints; For peak flood modulus constraints; Constraints on water level fluctuation range; Constrained by the time of flood rise; Constraints on water level-flow synchronization;

[0019] Step 6: Flood parameter calculation: For effective flood events, calculate various flood characteristic parameters, including peak flow, peak time, total flood volume, peak water level, initial rise water level, duration, and rise time;

[0020] Step 7: Output visualization results: Output visualization results, including a summary table of flood events and a flood process diagram.

[0021] Furthermore, the time series observation data mentioned in step 1 is at least the observation data of one flood season; the observation time t is in units of 10 minutes or 15 minutes, depending on the frequency of the hydrological station monitoring data.

[0022] Furthermore, in step 2, the flood discrimination threshold is calculated based on the statistical characteristics of the flow time series. The specific process is as follows:

[0023] First, calculate the mean and standard deviation of the flow series:

[0024] (1)

[0025] (2)

[0026] In the formula: The mean of the flow series, in units. ; The standard deviation of the flow series is given in units. ; For the first Observation time Corresponding flow rate, unit ; i is the sequence number of the observation time. N represents the total number of observation times in the flow sequence;

[0027] Then, the flood discrimination threshold is calculated as follows:

[0028] (3)

[0029] In the formula: Threshold for flood discrimination.

[0030] Furthermore, step 3 involves using gradient analysis to locate the starting point of the flood surge. The specific process is as follows:

[0031] (1) First-order gradient calculation:

[0032] (4)

[0033] In the formula: The time interval between adjacent observations, in seconds; for First-order gradient of flow at time step, in units ;

[0034] (2) Calculation of second-order gradient:

[0035] (5)

[0036] In the formula: for The second gradient of the flow rate at time t, in m³ / s³;

[0037] (3) Criteria for determining the starting point of the price increase:

[0038] ① The first-order gradient changes from non-positive to positive: and ;in, The point at which the price begins to rise, to be determined; for The first-order gradient of the flow at any given moment; for The previous observation time, i.e. ; for The first-order gradient of the flow at any given moment;

[0039] ② The second gradient is positive: ;in, for The second gradient of the flow at time step;

[0040] ③ The flow rate is near a local minimum within the search window, that is, during the search process from the moment to be determined... Take forward and backward respectively Each time step constitutes a local window, when Flow of time It is considered to be near a local minimum when the following conditions are met:

[0041] (6)

[0042] In the formula: Let be the window radius, and take a positive integer. ; This is the allowable fluctuation factor. ; This represents the time interval between adjacent observation times.

[0043] Furthermore, step 4 involves determining the receding point of the floodwaters. The specific process is as follows:

[0044] flood receding point The determination requires that one of the following conditions be met:

[0045] ① The flow rate drops back to near the starting point of the surge: ,in This is the tolerance factor, with a default value of 0.1.

[0046] ② Traffic volume drops below the threshold: ;

[0047] in, for Flow rate at any given moment; for Traffic flow at any given moment.

[0048] Furthermore, the specific process for determining whether each constraint condition in step 5 is satisfied is as follows:

[0049] ① Duration constraint The flood lasted for no less than 3 hours, that is:

[0050] (9)

[0051] In the formula: Duration of the flood; The moment when the flood recedes; This is the moment when the floodwaters begin to rise;

[0052] ② Peak flood modulus constraint The flood peak modulus, i.e., the peak flow rate per unit catchment area, is not less than 0.2 m³ / (s·km²), that is:

[0053] (10)

[0054] In the formula: The peak flow modulus is the peak flow rate per unit area of ​​the catchment area, expressed in m³ / (s·km²). Peak flow rate, in m³ / s; The drainage area is expressed in km².

[0055] ③ Water level fluctuation range constraints Flood level fluctuations range from 0.2m to 10m, excluding invalid and abnormal data;

[0056] (11)

[0057] In the formula: The magnitude of the flood level change is expressed in meters (m). For the peak of the flood The corresponding water level value, in meters (m); The starting point of the rise The corresponding water level value, in meters (m);

[0058] ④ Time constraints for rising water levels The time from the start of the rise to the peak of the flood must be no less than 30 minutes; that is:

[0059] (12)

[0060] In the formula: The time it takes for the water level to rise, that is, the duration from the starting point of the rise to the peak of the flood; This is the peak of the flood season; This is the moment when the floodwaters begin to rise;

[0061] ⑤ Constraints on the synchronization of water level and flow rate The consistency ratio between the direction of water level and flow rate changes should not be less than 60%, calculated as follows:

[0062] a. Calculate the changes in flow rate and water level at adjacent time points:

[0063] (13)

[0064] In the formula: For the first The change in flow rate within a time step, in m³ / s; For the first Water level change within a time step, in meters; Current flood period Inner At each observation time, ; This represents the total number of observation times during the current flood period.

[0065] b. Define a symbolic function to extract the direction of change:

[0066] (14)

[0067] c. Define an indicator function to determine directional consistency:

[0068] (15)

[0069] d. Calculate the synchronization ratio:

[0070] (16)

[0071] In the formula: The ratio of the synchronicity of changes in water level and flow rate is dimensionless. The current flood period is the total number of intervals between adjacent moments within the current flood period.

[0072] Furthermore, the total flood volume in step 6 is calculated using the trapezoidal integral method:

[0073] (17)

[0074] In the formula: Total flood volume, in m³; Current flood period Inner The flow rate at each observation time, in m³ / s; For the first The flow rate at each observation time, in m³ / s; The time interval between adjacent observation times, in units of ; This represents the total number of observation times during the current flood period.

[0075] Peak flow Peak flood time Flood peak water level Rising water level Duration Flooding time All of these are determined in the preceding steps.

[0076] Furthermore, the flood event summary table mentioned in step 7 includes: flood number, start time, end time, peak flow, peak time, duration, and total flood volume; the flood process diagram uses a dual vertical axis to display the time changes of flow and water level, and marks the starting point, peak, and receding point.

[0077] The beneficial effects of this invention are as follows: This invention employs a statistical method based on mean and standard deviation to calculate the flood discrimination threshold, which can automatically adapt to the hydrological characteristics of different watersheds; it uses a gradient analysis-based method for identifying the starting point of the flood, judging the flow change trend through first- and second-order gradients and determining the receding point based on the flow value, resulting in high identification accuracy; it uses a multi-condition joint constraint judgment method to identify effective flood events, integrating the water level-flow synchronization test into the constraint conditions to ensure the scientific validity and reliability of the identification results; and it uses the trapezoidal integral method to calculate the total flood volume, which is simple, reliable, and computationally efficient. Furthermore, the method described in this invention is applicable to different types of hydrological stations, supports batch automatic processing, and significantly improves work efficiency.

[0078] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0079] Figure 1 This is a schematic diagram of the method flow described in this invention;

[0080] Figure 2 This is a schematic diagram of the multi-condition joint constraint determination process;

[0081] Figure 3This is a diagram of a flood event in Example 1. Detailed Implementation

[0082] This invention discloses an intelligent method for identifying flood events based on statistical thresholds and multi-condition joint constraints, such as... Figure 1 As shown, the method includes the following steps:

[0083] Step 1: Data Collection. Collect time-series observation data from hydrological stations in the target area, including station name, observation time t, flow rate Q(t), water level Z(t), and catchment area A, arranged in chronological order to form an ordered dataset. The time-series observation data should cover at least one flood season (approximately three months), generally one year or several consecutive years, depending on the actual situation. The observation time t is generally measured in 10-minute or 15-minute units, depending on the frequency of the hydrological station's monitoring data.

[0084] Step 2: Flood discrimination threshold calculation: Calculate the flood discrimination threshold based on the statistical characteristics of the flow time series to identify potential flood periods.

[0085] First, the mean and standard deviation of the flow series are calculated as follows:

[0086] (1)

[0087] (2)

[0088] In the formula: The mean of the flow series, in units. ; The standard deviation of the flow series is given in units. ; For the first Observation time Corresponding flow rate, unit ; i is the sequence number of the observation time. N represents the total number of observation times in the flow sequence;

[0089] Then, the flood discrimination threshold is calculated as follows:

[0090] (3)

[0091] In the formula: Flood discrimination threshold;

[0092] At a certain observation time Traffic satisfaction At that time, the moment These moments are marked as exceeding the threshold. Timely consecutive or adjacent moments exceeding the threshold are merged to form potential flood periods. The first moment exceeding the threshold in each potential flood period is denoted as... The last time the threshold was exceeded was This threshold takes into account both the average level and the degree of fluctuation of the flow, and can adapt to the hydrological characteristics of different watersheds.

[0093] Step 3: Identification of flood initiation point based on gradient analysis: using the first threshold exceeding the threshold time during the potential flood period. Starting from the beginning of the search, the search proceeds backward along the time series (i.e., in the direction of decreasing time), and gradient analysis is used to accurately locate the starting point of the flood. The starting point is the inflection point where the flow rate changes from a stable or declining state to an increasing state.

[0094] (1) Calculation of first-order gradient (rate of change):

[0095] (4)

[0096] In the formula: The time interval between adjacent observations, in seconds; for First gradient (rate of change) of flow at time t, in units .

[0097] (2) Calculation of second-order gradient (acceleration):

[0098] (5)

[0099] In the formula: for The second gradient (acceleration) of the flow rate at time t, in m³ / s³.

[0100] (3) Criteria for determining the starting point of the price increase:

[0101] ① The first-order gradient changes from non-positive to positive: and .in, The point at which the price begins to rise, to be determined; for The first-order gradient of the flow at any given moment; for The previous observation time, i.e. ; for The first-order gradient of the flow at any given moment.

[0102] ② The second gradient is positive (flow rate increases rapidly): .in, for The second gradient of the flow at time step.

[0103] ③ The flow rate is near a local minimum within the search window, that is, during the search process from the moment to be determined... Take forward and backward respectively Each time step constitutes a local window, when Flow of time It is considered to be near a local minimum when the following conditions are met:

[0104] (6)

[0105] In the formula: The radius of the window, taken as a positive integer, defaults to... ; The default value is the allowable fluctuation factor. ; This represents the time interval between adjacent observation times.

[0106] Step 4, Determining the Flood Receding Point: First, at the point where the floodwaters begin to rise... In the subsequent period, find the moment corresponding to the maximum flow rate and record it as the peak flow moment. :

[0107] (7)

[0108] In the formula: The flow rate at the peak of the flood; This refers to the last time the threshold is exceeded, which is also the end of the potential flood period.

[0109] Then, at the peak of the flood Starting from this point, the search proceeds progressively forward along the time series (i.e., in ascending order of time) to determine the point where the flood recedes. The receding point is the location where the floodwaters recede and the flow returns to normal levels.

[0110] Criteria for determining the landing point (meeting any one of the following conditions is sufficient):

[0111] ① The flow rate drops back to near the starting point of the surge: ,in This is the tolerance factor, with a default value of 0.1.

[0112] ② Traffic volume drops below the threshold: .

[0113] in, for Flow rate at any given moment; for Traffic flow at any given moment.

[0114] Step 5: Identification of Valid Flood Periods Based on Multi-Condition Joint Constraint Judgment: For the identified potential flood periods, a multi-condition joint constraint judgment is performed. All constraints must be satisfied simultaneously for a period to be considered a valid flood period. The judgment process is as follows: Figure 2 As shown.

[0115] (1) The joint constraint determination criterion is: using logical AND operation, if all conditions are met simultaneously, it is determined to be a valid flood event:

[0116] (8)

[0117] In the formula: For effective flood events; For duration constraints; For peak flood modulus constraints; Constraints on water level fluctuation range; Constrained by the time of flood rise; This is a constraint on the synchronization of water level and flow rate.

[0118] (2) Specific constraints include:

[0119] ① Duration constraint The flood lasted for no less than 3 hours, that is:

[0120] (9)

[0121] In the formula: Duration of the flood; The moment when the flood recedes; This is the moment when the floodwaters begin to rise.

[0122] ② Peak flood modulus constraint The flood peak modulus, i.e., the peak flow rate per unit catchment area, is not less than 0.2 m³ / (s·km²), that is:

[0123] (10)

[0124] In the formula: The peak flow modulus is the peak flow rate per unit area of ​​the catchment area, expressed in m³ / (s·km²). Peak flow rate, in m³ / s; The watershed area is expressed in km².

[0125] ③ Water level fluctuation range constraints The flood level fluctuation is within a reasonable range (0.2m to 10m), and invalid and abnormal data are excluded;

[0126] (11)

[0127] In the formula: The magnitude of the flood level change is expressed in meters (m). For the peak of the flood The corresponding water level value, in meters (m); The starting point of the rise The corresponding water level value, in meters (m).

[0128] ④ Time constraints for rising water levels The time from the start of the rise to the peak of the flood must be no less than 30 minutes; that is:

[0129] (12)

[0130] In the formula: The time it takes for the water level to rise, that is, the duration from the starting point of the rise to the peak of the flood; This is the peak of the flood season; This is the moment when the floodwaters begin to rise.

[0131] ⑤ Constraints on the synchronization of water level and flow rate The consistency ratio between the direction of water level and flow rate changes should not be less than 60%, calculated as follows:

[0132] a. Calculate the changes in flow rate and water level at adjacent time points:

[0133] (13)

[0134] In the formula: For the first The change in flow rate within a time step, in m³ / s; For the first Water level change within a time step, in meters; Current flood period Inner At each observation time, ; This represents the total number of observation times during the current flood period.

[0135] b. Define a symbolic function to extract the direction of change:

[0136] (14)

[0137] c. Define an indicator function to determine directional consistency:

[0138] (15)

[0139] d. Calculate the synchronization ratio:

[0140] (16)

[0141] In the formula: The ratio of the synchronicity of changes in water level and flow rate is dimensionless. The current flood period is the total number of intervals between adjacent moments within the current flood period.

[0142] (3) Judgment result: The potential flood period is confirmed as a valid flood event if and only if the above 5 constraints are met at the same time; otherwise, it is excluded.

[0143] Step 6: Flood Parameter Calculation: For valid flood events that have passed verification, calculate various flood characteristic parameters, including peak flow, peak time, total flood volume, peak water level, initial rise water level, duration, and rise time.

[0144] (1) Peak flow and peak time: Peak flow and the peak of the flood This has been determined in step 4.

[0145] (2) Calculation of total flood volume: The trapezoidal integral method is used for calculation.

[0146] (17)

[0147] In the formula: Total flood volume, in m³ (final result converted to million cubic meters). Current flood period Inner The flow rate at each observation time, in m³ / s; For the first The flow rate at each observation time, in m³ / s; The time interval between adjacent observation times, in units of ; This represents the total number of observation times during the current flood period.

[0148] (3) Other parameters: including peak water level and peak time Corresponding water level value Unit: meters; Starting water level: the moment the water level begins to rise. Corresponding water level value , Unit: meters; Duration ; time of flooding wait.

[0149] Step 7: Output Visualization Results: Output visualization results, including a flood event summary table and a flood process diagram. The flood event summary table includes information such as: flood number, start time, end time, peak flow, peak time, duration, and total flood volume. The flood process diagram uses a dual vertical axis to display the time changes in flow and water level, and marks key feature points such as the starting point, peak, and receding point.

[0150] Example 1

[0151] This embodiment uses the Jiuduhe Hydrological Station in Beijing as an example to illustrate the specific application of the above method. Water level and flow observation data from the station between July 21, 2022, and December 31, 2024 were selected, and the above method was used to intelligently identify flood events across the entire time period.

[0152] First, time-series observation data from the Jiuduhe hydrological station for the aforementioned period were collected, including flow rate. and water level The data is arranged chronologically to form an ordered dataset. Then, based on the mean and standard deviation of the station's flow sequence, the flood discrimination threshold is calculated as follows: m³ / s. After steps 2 to 5, including screening for exceeding the threshold, identifying the starting point of the flood, identifying the flood peak and determining the receding point, and verifying the joint constraints of multiple conditions, three valid flood events were finally identified from the data of this period.

[0153] Taking one flood as an example, the execution process and results of each step are explained in detail. This flood occurred between 18:00 on August 9, 2024 and 19:00 on August 11, 2024, and its flood process is as follows: Figure 3 As shown.

[0154] (1) Identification of the flood initiation point: Based on the gradient analysis method in step 3, the flood initiation point was identified as approximately 18:00 on August 9, 2024, with a corresponding initiation flow rate of [missing information]. m³ / s, in Figure 3 The graph is marked with a green triangle. At this point, the first gradient of the flow rate changes from non-positive to positive, and the second gradient is positive, indicating that the flow rate is beginning to accelerate.

[0155] (2) Peak flow identification and receding point determination: Based on the method in step 4, the peak flow time was identified as approximately 00:00 on August 10, 2024, and the peak flow rate was... m³ / s, in Figure 3 Marked with a red five-pointed star. The search then proceeded in ascending time, determining the relocation point to be approximately 19:00 on August 11, 2024. Figure 3 It is marked with a black inverted triangle.

[0156] (3) Verification of multiple constraints: Based on step 5, verify the five constraints of this flood event:

[0157] ① Duration constraint: h h, satisfies;

[0158] ② Peak flow modulus constraint: based on the drainage area of ​​the station. Calculate the peak flood modulus Verify that it is not less than 0.2 m³ / (s·km²), which satisfies the requirement;

[0159] ③ Water level fluctuation range constraints: The initial rise in water level is approximately 176.45 m, and the peak water level is approximately 176.71 m. m is within a reasonable range of 0.2 m to 10 m, which satisfies the requirement;

[0160] ④ Time constraint for rising water level: from the starting point of the rise (18:00 on August 9) to the flood peak (00:00 on August 10). h min, satisfied;

[0161] ⑤ Constraints on the synchronicity of water level and flow rate: by Figure 3 It can be seen that the water level curve and the flow rate curve show the same overall trend, both exhibiting a rapid initial rise followed by a gradual decline, demonstrating a high degree of synchronicity. ,satisfy.

[0162] If all five constraints are met, the flood event is determined to be a valid flood event.

[0163] (4) Flood parameter calculation: Based on step 6, the following flood characteristic parameters are calculated: peak flow The flood flow rate is m³ / s, with the peak time around 00:00 on August 10, 2024, and the duration is... h, the total flood volume is calculated using the trapezoidal integral method. One million cubic meters. Figure 3 The orange-filled area visually represents the area corresponding to the total flood volume.

[0164] Finally, output the visualization results: Figure 3 The graph depicts the process of this flood event, using a dual vertical axis format. The left vertical axis represents water level (in meters), and the right vertical axis represents flow rate (in m³ / s). The horizontal axis represents time. The blue curve represents the water level change, the red curve represents the flow rate change, the horizontal dashed line represents the flood identification threshold, and the green triangle, red pentagram, and black inverted triangle mark the starting point, peak, and receding point, respectively. The identification process for the other two valid flood events is the same as described above and will not be shown here.

[0165] Finally, it should be noted that the above description is only used to illustrate the technical solution of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred arrangement, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for intelligent identification of flood events based on statistical thresholds and multi-condition joint constraints, characterized in that, The method includes the following steps: Step 1: Data Collection: Collect time-series observation data from hydrological stations in the target area, including station name, observation time t, flow rate Q(t), water level Z(t), and catchment area A, and arrange them in chronological order to form an ordered dataset; Step 2: Calculation of flood discrimination threshold: Calculate the flood discrimination threshold based on the statistical characteristics of the flow time series. At a certain observation time Traffic satisfaction At that time, the moment Mark the moments exceeding the threshold; merge temporally consecutive or adjacent moments exceeding the threshold to form potential flood periods; denote the first moment exceeding the threshold in each potential flood period as... The last time the threshold was exceeded was ; Step 3: Identification of flood initiation point based on gradient analysis: using the first threshold exceeding the threshold time during the potential flood period. Starting from the beginning of the search, the search proceeds backward along the time series, i.e., in the direction of decreasing time, and gradient analysis is used to locate the starting point of the flood. ; Step 4, Determining the flood receding point: First, at the flood's starting point... In the subsequent period, find the moment corresponding to the maximum flow rate and record it as the peak flow moment. : （7） In the formula: The flow rate at the peak of the flood; This refers to the last time the threshold is exceeded, which is also the end of the potential flood period. Then, at the peak of the flood Starting from the beginning of the search, the search proceeds progressively along the time series in the future direction, i.e., in increasing time, to determine the receding point of the flood. ; Step 5: Identification of valid flood events based on multi-condition joint constraint judgment: Perform multi-condition joint constraint judgment on the identified potential flood periods. When all constraints are met simultaneously, the potential flood period is judged as a valid flood event; otherwise, it is rejected. The joint constraint determination of multiple conditions uses the logical AND operation: （8） In the formula: For effective flood events; For duration constraints; For peak flood modulus constraints; Constraints on water level fluctuation range; Constrained by the time of flood rise; Constraints on water level-flow synchronization; Step 6: Flood parameter calculation: For effective flood events, calculate various flood characteristic parameters, including peak flow, peak time, total flood volume, peak water level, initial rise water level, duration, and rise time; Step 7: Output visualization results: Output visualization results, including a summary table of flood events and a flood process diagram.

2. The intelligent flood field identification method based on statistical threshold and multi-condition joint constraints according to claim 1, characterized in that, The time series observation data mentioned in step 1 is at least the observation data of one flood season; the observation time t is in units of 10 minutes or 15 minutes, depending on the frequency of the hydrological station monitoring data.

3. The intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints according to claim 1, characterized in that, Step 2 involves calculating the flood discrimination threshold based on the statistical characteristics of the flow time series. The specific process is as follows: First, calculate the mean and standard deviation of the flow series: （1） （2） In the formula: The mean of the flow series, in units. ; The standard deviation of the flow series is given in units. ; For the first Observation time Corresponding flow rate, unit ; i is the sequence number of the observation time. N represents the total number of observation times in the flow sequence; Then, the flood discrimination threshold is calculated as follows: （3） In the formula: Threshold for flood discrimination.

4. The intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints according to claim 1, characterized in that, Step 3 describes using gradient analysis to locate the starting point of the flood. The specific process is as follows: (1) First-order gradient calculation: （4） In the formula: The time interval between adjacent observations, in seconds; for First-order gradient of flow at time step, in units ; (2) Calculation of second-order gradient: （5） In the formula: for The second gradient of the flow rate at time t, in m³ / s³; (3) Criteria for determining the starting point of the price increase: ① The first-order gradient changes from non-positive to positive: and ;in, The point at which the price begins to rise, to be determined; for The first-order gradient of the flow at any given moment; for The previous observation time, i.e. ; for The first-order gradient of the flow at any given moment; ② The second gradient is positive: ;in, for The second gradient of the flow at time step; ③ The flow rate is near a local minimum within the search window, that is, during the search process from the moment to be determined... Take forward and backward respectively Each time step constitutes a local window, when Flow of time It is considered to be near a local minimum when the following conditions are met: （6） In the formula: Let be the window radius, and take a positive integer. ; This is the allowable fluctuation factor. ; This represents the time interval between adjacent observation times.

5. The intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints according to claim 1, characterized in that, Step 4 describes determining the receding point of the floodwaters. The specific process is as follows: flood receding point The determination requires that one of the following conditions be met: ① The flow rate drops back to near the starting point of the surge: ,in This is the tolerance factor, with a default value of 0.

1. ② Traffic volume drops below the threshold: ; in, for Flow rate at any given moment; for Traffic flow at any given moment.

6. The intelligent flood field identification method based on statistical threshold and multi-condition joint constraints according to claim 1, characterized in that, The specific process for determining whether each constraint condition is satisfied in step 5 is as follows: ① Duration constraint The flood lasted for no less than 3 hours, that is: （9） In the formula: Duration of the flood; The moment when the flood recedes; This is the moment when the floodwaters begin to rise; ② Peak flood modulus constraint The flood peak modulus, i.e., the peak flow rate per unit catchment area, is not less than 0.2 m³ / (s·km²), that is: （10） In the formula: The peak flow modulus is the peak flow rate per unit area of ​​the catchment area, expressed in m³ / (s·km²). Peak flow rate, in m³ / s; The drainage area is expressed in km². ③ Water level fluctuation range constraints Flood level fluctuations range from 0.2m to 10m, excluding invalid and abnormal data; （11） In the formula: The magnitude of the flood level change is expressed in meters (m). For the peak of the flood The corresponding water level value, in meters (m); The starting point of the rise The corresponding water level value, in meters (m); ④ Time constraints for rising water levels The time from the start of the rise to the peak of the flood must be no less than 30 minutes; that is: （12） In the formula: The time it takes for the water level to rise, that is, the duration from the starting point of the rise to the peak of the flood; This is the peak of the flood season; This is the moment when the floodwaters begin to rise; ⑤ Constraints on the synchronization of water level and flow rate The consistency ratio between the direction of water level and flow rate changes should not be less than 60%, calculated as follows: a. Calculate the changes in flow rate and water level at adjacent time points: （13） In the formula: For the first The change in flow rate within a time step, in m³ / s; For the first Water level change within a time step, in meters; Current flood period Inner At each observation time, ; This represents the total number of observation times during the current flood period. b. Define a symbolic function to extract the direction of change: （14） c. Define an indicator function to determine directional consistency: （15） d. Calculate the synchronization ratio: （16） In the formula: The ratio of the synchronicity of changes in water level and flow rate is dimensionless. The current flood period is the total number of intervals between adjacent moments within the current flood period.

7. The intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints according to claim 6, characterized in that, The total flood volume in step 6 is calculated using the trapezoidal integral method: （17） In the formula: Total flood volume, in m³; Current flood period Inner The flow rate at each observation time, in m³ / s; For the first The flow rate at each observation time, in m³ / s; The time interval between adjacent observation times, in units of ; This represents the total number of observation times during the current flood period. Peak flow Peak flood time Flood peak water level Rising water level Duration Flooding time All of these are determined in the preceding steps.

8. The intelligent identification method for flood events based on statistical thresholds and multi-condition joint constraints according to claim 1, characterized in that, The flood event summary table mentioned in step 7 includes: flood number, start time, end time, peak flow, peak time, duration, and total flood volume; the flood process diagram uses a dual vertical axis to display the time change process of flow and water level, and marks the starting point, peak and receding point.

Images