Method for calculating wavefront water level through sudden flood forecast flow

Abstract

The invention belongs to the technical field of flood forecasting, and particularly relates to a method for calculating a wavefront water level through sudden flood forecast flow, which comprises the following steps: acquiring hydrological observation and forecast data of session flood in real time, and determining a fixed node for wavefront water level forecast calculation; determining whether the forecast initial flow needs to be corrected; determining an initial flow correction control node for initial flow correction; converting the forecast traffic into adaptive traffic; filtering the adaptive traffic to generate first adaptive traffic; converting the first adaptive traffic into second adaptive traffic; calculating a first water level; converting the first water level into a first corrected water level; calculating a second water level; converting the second water level into a second corrected water level; calculating a water surface gradient process; and ending the flood wavefront water level forecasting operation of the session, and outputting a wavefront water level forecasting calculation result. According to the method, two wavefront flood flow hydrographs and corresponding water level hydrographs can be generated, and the technical level of flood wavefront water level forecasting is effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of flood forecasting technology, specifically relating to a method for calculating wavefront water level in forecasting sudden floods. Background Technology

[0002] Floods have always been a major concern for the Chinese nation. Sudden floods refer to regional flash floods (with a confluence duration typically within 6 hours) caused by short-duration, high-intensity torrential rains, or floods formed downstream by dam breaks and the breaching of high-altitude glacial lakes. Affected by global climate change, extreme weather events are becoming increasingly frequent, leading to more frequent, high-incidence, and easily occurring sudden flood disasters, which are the most serious type of flood causing injury and death. According to statistics from the Planning and Programming Department of the Ministry of Water Resources, in normal years, sudden floods in small and medium-sized rivers account for approximately 80% of the total losses from floods nationwide, and in the past 10 years, more than two-thirds of flood-related deaths have occurred in small and medium-sized rivers. Small and medium-sized rivers remain a weak link in my country's flood control and disaster relief efforts, characterized by low accuracy in water level forecasts, short effective lead times, and a high rate of false alarms, making them a vulnerable aspect of hydrological forecasting services.

[0003] Commonly used flood forecasting methods include hydrological methods, hydraulic methods, and system mathematical models. Forecasting of sudden flood flow is relatively mature, and a practical and reliable method is to establish a rainfall-runoff calculation model based on hydrological mechanisms. Flood waves in the middle and lower reaches of small and medium-sized rivers are diffusion waves. The flood wave body before the maximum flow of a flood event is called the wavefront, and the flood wave body after the maximum flow is called the waveback. Influenced by factors such as fluctuating backwater and flood rise and fall, the water level-discharge relationship at any cross-section in the middle and lower reaches of small and medium-sized rivers exhibits a single or complex loop shape. The water conditions before the wavefront are determined by upstream rainfall and the initial water conditions at the cross-section, while the water conditions after the waveback are mainly controlled by the downstream boundary conditions of the river network. Typically, the single-valued online calculation technology for the water level-discharge relationship mainly uses the head index method, which calculates the cross-sectional flow process based on continuous water level observation data of the calculation cross-section and reference cross-section, and discrete flow observation data after simplified measurements of the calculation cross-section. However, water level forecasting is a water level-discharge inverse calculation problem, which is technically challenging, and the methods urgently need improvement. Summary of the Invention

[0004] The present invention addresses the above-mentioned technical problems by providing a method for calculating wavefront water levels in forecasting sudden floods.

[0005] A method for calculating wavefront water level based on forecast flow for sudden floods, the method comprising:

[0006] S1, real-time acquisition of hydrological observation and forecast data of flood events, determination of fixed nodes for wavefront water level forecast calculation, the hydrological observation and forecast data including observed water level, observed flow rate and forecast flow rate, the fixed nodes from front to back include starting point, flow rate-water level relationship control node and calculation end point;

[0007] S2, determine whether the initial forecast flow needs to be corrected. If no correction is needed, proceed to step S4; otherwise, proceed to step S3.

[0008] S3, determine the initial traffic correction control node for initial traffic correction, including the control node before initial traffic correction and the control node after initial traffic correction;

[0009] S4, starting from the starting point to the flow-water level relationship control node, several flow conversion control nodes on the forecast flow are located step by step from front to back. Two adjacent flow conversion control nodes form a segment, and the forecast flow is divided into one or several segments. The node type of the control node before each segment corresponds to a preset node type algorithm for converting the forecast flow to the appropriate flow. The forecast flow before the flow-water level relationship control node is converted into the appropriate flow through each node type algorithm.

[0010] S5, convert the predicted flow rate between the flow-level relationship control node and the calculation end node into an adapted flow rate;

[0011] S6, the adapted traffic is filtered and then converted into the first adapted traffic;

[0012] S7, convert the first adapted traffic into the second adapted traffic;

[0013] S8, calculate the first water level between the starting point and the flow-water level relationship control node based on the first adaptive flow rate;

[0014] S9, starting from the starting point to the flow-water-level relationship control node, several water level correction nodes on the first water level are located step by step from front to back. Two adjacent water level correction nodes form a segment, thus dividing the first water level into one or several segments. The water level correction type of the correction node before each segment corresponds to a preset correction algorithm for correcting the first water level, forming a first water level correction scheme; starting from the starting point to the flow-water-level relationship control node, the corresponding first corrected water level is obtained through the first water level correction calculation of each segment;

[0015] S10, using the first adaptive flow rate between the control node and the calculation end node based on the flow-water-level relationship, calculate the first corrected water level;

[0016] S11, calculate the second water level between the starting point and the flow-water level relationship control node based on the second adapted flow rate;

[0017] S12, from the starting point to the flow-water-level relationship control node, the first water level correction scheme compiled in step S9 is adopted, including all correction nodes, segments, water level correction types and correction algorithms of segments, and from the starting point to the flow-water-level relationship control node, the corresponding second corrected water level is obtained by calculating the second water level correction of each segment.

[0018] S13, using the second adaptive flow rate between the control node and the calculation end node based on the flow-water-level relationship, calculate the second corrected water level;

[0019] S14. Calculate the water surface gradient process based on the first adapted flow rate and the corresponding first corrected water level, the second adapted flow rate and the corresponding second corrected water level.

[0020] Optionally, step S1 includes:

[0021] S101, if the forecast flow reaches the preset minimum control flow earlier than the observed flow, then the time point when the forecast flow reaches the minimum control flow is taken as the starting point; otherwise, the time point when the observed flow is not less than the minimum control flow and the forecast flow reaches 50% of the minimum control flow is taken as the starting point.

[0022] S102, after the starting point, the maximum value of the predicted flow rate Q is taken as the calculation end point;

[0023] S103, starting from the end node of the calculation, the time point when the time-varying acceleration aQ of the predicted flow rate Q first turns from negative to positive and the value is closest to 0 is taken as the flow rate-water level relationship control node. When the time between the end node of the calculation and the flow rate-water level relationship control node is less than the preset minimum time for cross-sectional hydrological analysis, the flow rate-water level relationship control node is located according to the preset minimum time between the end node of the calculation and the flow rate-water level relationship control node.

[0024] The formula for calculating the time-varying velocity VQ of flow rate Q is:

[0025] (1)

[0026] The formula for calculating the time-varying acceleration aQ of flow rate Q is:

[0027] (2)

[0028] in, It is time-series flow Data sequence number starting from 1 (referred to as time point number); It is the time difference between two adjacent data points; The value of i is calculated starting from 2. The value of i is calculated starting from 3.

[0029] Optionally, in step S2, if the observed water level and the predicted flow rate have inconsistent trends, or if the observed flow rate and the predicted flow rate have inconsistent trends, then the predicted flow rate needs to be corrected; otherwise, the predicted flow rate does not need to be corrected.

[0030] Optionally, in step S3, the control node before the initial flow correction adopts the following positioning method:

[0031] During the period between the control node before the initial flow correction and the starting point, the changes and trends of the observed water level and observed flow are consistent; during the period between the control node before the initial flow correction and the starting point, the time-varying velocity VQ and time-varying acceleration aQ of the predicted flow Q remain unchanged; the duration between the control node before the initial flow correction and the starting point is not less than the preset minimum duration.

[0032] Optionally, starting from the starting point, the time point at which the sign of either the time-varying velocity VQ or the time-varying acceleration aQ of the predicted flow Q changes is selected as the control node before the initial flow correction.

[0033] Optionally, in step S3, the control node bit after the initial flow correction is located at the last of the following three timing points:

[0034] When the predicted flow rate at the starting point is less than the preset minimum control flow rate, the first point where the predicted flow rate is closest to rising to the minimum control flow rate is taken as the control node after the initial flow rate correction.

[0035] After the starting point, the time-varying velocity VQ and time-varying acceleration aQ of the predicted flow Q remain unchanged. The time point at which either the time-varying velocity VQ or the time-varying acceleration aQ of the predicted flow first shows a positive-to-negative value transition point is used as the control node after the initial flow correction.

[0036] After the initial flow correction, the time between the control node and the starting point is not less than the preset minimum time.

[0037] Optionally, in step S4, the type of the flow conversion control node includes at least one of the following: steady type, fluctuating type, and rising type;

[0038] The stable advancement type includes at least one of a first stable advancement type and a second stable advancement type;

[0039] The method for determining the steady-progress type is as follows: starting from the starting point or the control node before the segment, the time point where the time-varying velocity VQ of the predicted flow Q changes from positive to negative and is closest to 0 is taken as the subsequent control node of the steady-progress type; when the segment duration of the steady-progress type exceeds the cross-sectional flow history and the predicted flow and observed water level and observed flow change trends within the segment are consistent, the steady-progress type is the first steady-progress type; when the segment duration of the steady-progress type exceeds 1.5 times the cross-sectional flow history and the observed water level and observed flow change trends within the segment are consistent, but the time-varying velocity VQ of the predicted flow Q is greater than a preset deviation value, the steady-progress type is the second steady-progress type.

[0040] The node type algorithm corresponding to the first stable type has the following adapted traffic calculation formula:

[0041] (4)

[0042] in, It is the timing point number starting from the control node before segmentation, and the control node before segmentation... =1, The time series index i is calculated starting from 2; It is the adapted flow of time series number i, and the adapted flow of the starting point is the observed flow; It is the time-varying velocity of the predicted flow rate at time point i;

[0043] The node type algorithm corresponding to the second stable type has the following adapted traffic calculation formula:

[0044] (5)

[0045] in, It is the timing point number starting from the control node before segmentation, and the control node before segmentation... =1; It is the predicted flow rate of the last time series point within the segment. It is the predicted flow rate of the first time series point within the segment; It is the adapted traffic for time sequence number i. The time series index i is calculated starting from 2; It is the time-varying velocity of the predicted flow rate at time point i; These are preset parameters;

[0046] The fluctuation type includes at least one of a first fluctuation type and a second fluctuation type;

[0047] The method for judging the first type of fluctuation is as follows: starting from the control node before segmentation, the time-varying velocity VQ value of the predicted flow Q changes from positive to negative and then back to positive, the time-varying acceleration aQ of the predicted flow Q changes from negative to positive and then back to negative, the predicted flow first rises and then falls and then rises again, and the absolute value of the flow difference in the falling segment is less than the flow difference between the two rising segments.

[0048] The post-control node of the first fluctuation type is located at the time point when the time-varying velocity VQ of the predicted flow Q falls to the point where the value is closest to 0.

[0049] The node type algorithm corresponding to the first fluctuation type is:

[0050] Using the time series index within a segment as the independent variable, the adapted flow rate for each time series index within the segment is calculated using an exponential function. The calculation formula is as follows:

[0051] (6)

[0052] in, The sequence number within the segment starts from 1, which is the sequence number of the control node before the segmentation. It is the traffic adapted for i6 timing point sequence numbers. The calculated time series index i6 starts from 2; and These are the two parameters of the exponential function;

[0053] Formula (4) is used to calculate the adaptive flow at the first two time points, the middle time point, and the last two time points within a segment. The adaptive flow at these five points is used to determine the two parameters of the adaptive flow exponential function for segmented calculation. and Then, calculate the adaptive flow corresponding to each time point within the segment according to formula (6);

[0054] The method for determining the second type of fluctuation is as follows: starting from the control node before the segment, the time-varying velocity VQ value of the predicted flow rate Q first decreases and turns negative, then gradually increases and turns positive, and the time-varying acceleration aQ of the predicted flow rate Q gradually increases and turns positive; preferably, if the control node after the next segment is the flow-water level relationship control node, then this segment is determined to be the second type of fluctuation.

[0055] The second type of fluctuation's post-control node is located at the last of the following three reference time series points: the time series point where the forecast flow rises after a decrease to the point where it first exceeds the control node before the segment; the time series point where the forecast flow rises after a decrease to the minimum value within the segment, located dTpf hours before the time-varying velocity VQ of the forecast flow Q turns positive, where dTpf is the hydrological characteristic value of the forecast section; and the time series point where the time-varying acceleration aQ of the forecast flow Q increases to a local maximum value that is positive.

[0056] The node type algorithm corresponding to the second fluctuation type has the following adapted traffic calculation formula:

[0057] (7)

[0058] in, The sequence number within the segment starts from the sequence number of the control node before the segmentation, which is 1. It's about adapting to traffic volume. The calculation of i starts from 2; It is a forecast of the flow rate; It is the time-varying velocity of the forecast flow; It is the minimum forecast flow within the segment; It is the maximum predicted flow rate within the segment; and These are preset parameters;

[0059] The upward movement type includes at least one of the following: first upward movement type, second upward movement type, third upward movement type, fourth upward movement type, and ninth upward movement type;

[0060] The method for determining the first, third, and ninth types of water level rise is as follows: starting from the control node before segmentation, the forecast flow rate continuously increases and the water surface gradient continues to increase.

[0061] The method for locating the post-control node for the first, third, and ninth rise types is as follows: if the time between the pre-segment control node and the calculation end node does not exceed twice the preset minimum time, then the node is located at the calculation end node; otherwise, the node is located at the time point where the predicted flow rate Q drops to its closest point to 0 after the maximum value of the time-varying velocity VQ. The segmentation time should exceed the cross-sectional flow duration; if it does not exceed this duration, other node types are selected.

[0062] The further determination method for the first, third, and ninth rise types is as follows: First, use formula (4) to calculate the initial value of the adapted flow of the control node after segmentation and the initial value of the adapted flow of the control node before segmentation. If the relative error of the initial value of the adapted flow of the control node before segmentation relative to the forecast flow exceeds 5%, then the segmentation is the ninth rise type; if the above conditions are not met, but the relative error of the initial value of the adapted flow of the control node after segmentation relative to the forecast flow does not exceed 5%, then the segmentation is the third rise type; the segmentation that does not meet the above conditions is the first rise type.

[0063] The node type algorithms corresponding to the first, third, and ninth rise types, and the adapted traffic calculation process are as follows:

[0064] First, determine 6 fitting points for the adaptive flow calculation. The method for determining these points is as follows: the control node before segmentation, a time series point after the control node before segmentation, and 4 consecutive time series points starting from the m-th point forward from the control node after segmentation.

[0065] The method for calculating m is as follows:

[0066] (8)

[0067] in, The sequence number of the last point of the timing data within the segment, starting from the timing point number of the control node before the segmentation, is 1.

[0068] Calculate the time-varying velocity adjustment ratio coefficients for the three types of flow increases mentioned above. :

[0069] (9)

[0070] in, It is the predicted flow of the control nodes before segmentation; It is the predicted flow of the control nodes after segmentation; It is the adaptation traffic of the control node before the segmentation control point; It is the adaptation traffic of the control node before segmentation;

[0071] according to Calculate the fitted flow at the fitting point for the 6 fitted flows within the segment:

[0072] (10)

[0073] The adaptive flow of the fitting point is calculated based on the time sequence number of the fitting point from the control node before segmentation and the formula (10). The adaptive flow of the fitting point is calculated based on the fitting point of the fitting point and the formula (6) is used to determine the two calculation parameters. Then, the adaptive flow of each time sequence number in the segment, except for the control node before segmentation, is recalculated based on the formula (6).

[0074] The method for determining the second type of rise is as follows: the control node before the segment is the outlet of the flow-water level curve loop, or the control node after the fluctuation type, which predicts that the flow will continue to rise.

[0075] The method for locating the post-control node of the second type of rise is as follows: if the time between the pre-control node and the calculation end node does not exceed twice the preset minimum time, then it is located at the calculation end node; otherwise, it is located at the time series point where the value of the time-varying velocity VQ of the forecast flow Q first appears after the pre-control node and then drops to the sign of the value closest to 0.

[0076] The node type algorithm corresponding to the second type of rise is: the flow difference of the adapted flow within the segment is equal to the flow difference of the predicted flow, the adapted flow of the control node before the segment is equal to the adapted flow of the control node after the segment of the previous segment, and the adapted flow is calculated using formula (4).

[0077] The fourth type of rise is a forecast initial flow correction algorithm, the front control node of the fourth type of rise is the starting point, and the back control node of the fourth type of rise is the back control node of the initial flow correction.

[0078] The node type algorithm corresponding to the fourth type of increase is as follows: According to step S3, the control node before the initial flow correction and the control node after the initial flow correction are determined. The observed flow from the control node before the initial flow correction to the starting point is used. An autoregressive moving average model is used to construct a cubic polynomial, and the adapted flow from the starting point to the segmented control node is extrapolated.

[0079] Optionally, step S5 includes: using formula (4) to calculate the adaptive flow rate between the flow-water level relationship control node and the calculation end node.

[0080] Optionally, step S6 includes: using a 3-point numerical filtering algorithm to filter the adaptive flow between the starting point and the end point of the calculation and convert it into the first adaptive flow.

[0081] Optionally, the calculation formula for the 3-point numerical filtering algorithm is as follows:

[0082] (11)

[0083] in, It is the first adaptive flow at the starting point; It is the adaptive traffic at the starting point; It is the sequence number of the time series starting from the starting point, and the starting point =1; It is the first adapted traffic for time sequence number i; It is the adapted traffic for time sequence number i; It is the first adapted traffic for the calculation of the end node; It calculates the adaptation traffic for the end node.

[0084] Optionally, in step S7, the calculation formula for converting the first adapted traffic into the second adapted traffic is as follows:

[0085] (12)

[0086] in, and These are the preset parameters; It is the sequence number of the timing point within the segment, starting from the timing point number of the control node before the segmentation, which is 1; It is the first adapted traffic for time sequence number i; It is the first adapted traffic for the calculation of the end node; It is the first adaptive flow at the starting point; It is the second adaptive flow at the starting point; It is the second adapted traffic for time sequence number i. The calculation of i starts from 2; It is the time-varying velocity of the first adapted flow of time sequence number i.

[0087] Optionally, step S8 includes:

[0088] Based on the preset flow boundary control value The flood magnitude of the flood event is determined, and the flood magnitude is divided into two levels: small to medium flow and large flow. Based on the flood magnitude, the first water level of the flow-water level control node is calculated.

[0089] (14)

[0090] Among them, QP1 99 The first adaptive flow rate for the flow-level relationship control node, H1 99 The first water level at the control node for the flow-water level relationship; , , , and These are the preset parameters;

[0091] Calculate the first water level between the starting point and the flow-water-level relationship control node:

[0092] (15)

[0093] in, It is the sequence number of the time series starting from the starting point, and the starting point =1; It is the first watermark of time point i. The calculation of i starts from 2; It is the first adapted traffic for time sequence number i; It is the observed flow rate at the starting point; It is the observed water level at the starting point.

[0094] Optionally, step S11 uses the same calculation method as step S8 to calculate the second water level between the starting point and the flow-water level relationship control node. When calculating the second water level in step S11, the first adaptive flow rate is replaced with the second adaptive flow rate, and the first water level is replaced with the second water level. The parameters used in formula (14) are different from those in step S8.

[0095] Optionally, in step S9, the water level correction type includes at least one of the basic type, the extension type, and the rope type;

[0096] The basic type includes at least one of the zero basic type, the first basic type, and the second basic type;

[0097] The extension types include the seventh extension type and the eighth extension type;

[0098] The method for determining the zeroth basic type is as follows: starting from the starting point or the pre-segment correction node, if the first adapted flow rate before the flow-water level relationship control node increases continuously, and the first adapted flow rate and the first water level change trend are the same within the preset shortest time before the pre-segment correction node, then the post-correction node of the zeroth basic type is located at the flow-water level relationship control node.

[0099] The correction algorithm corresponding to the zeroth basic type, the water level correction calculation formula is as follows:

[0100] (16)

[0101] in, It is the first corrected water level after the time series number i has been corrected. It is the first correction water level of the correction node before segmentation; It is the sequence number starting from the pre-segmentation correction node, and the pre-segmentation correction node's... The value is 1, and the water level correction calculation i starts from 2; It is the first watermark of time point i. It is the first water level of the pre-segmentation correction node; when the pre-segmentation correction node is the starting point. ;

[0102] The method for determining the first basic type and the seventh extended type is as follows: the first basic type and the seventh extended type are located by the sign change of the time-varying acceleration aQP1 of the first adaptive flow QP1 being closest to 0 or a local extreme value; after the first basic type and the seventh extended type continue to rise after the previous correction node, the first basic type and the seventh extended type are located by the local extreme value of the time-varying acceleration aQP1 of the first adaptive flow QP1; when the segmented previous correction node is within the preset shortest time, relative to the segmented previous correction node, the slope of the first adaptive flow and the first water level curve at each time point is greater than that in formula (15). When the value is 1.2 times, the segment is the seventh extended type; otherwise, the segment is the first basic type.

[0103] The correction algorithm corresponding to the first basic type has the following water level correction calculation formula:

[0104] (18)

[0105] in, It is the first corrected water level after the time series number i has been corrected. It is the first correction water level of the correction node before segmentation; It is the sequence number of the time series starting from the starting point, and the previous correction node. The value is 1, and the water level correction calculation i starts from 2; It is the time-varying acceleration of the first adapted flow of the segmented correction node; It is the time-varying acceleration of the first adaptive flow of the correction node before segmentation; It refers to the duration of each segment; It is the first adapted traffic of the correction node before segmentation; It is the first adapted traffic of the segmented and corrected node; It is the first water level of the correction node before segmentation; These are preset parameters;

[0106] The correction algorithm corresponding to the seventh extension type has the following water level correction calculation formula:

[0107] (19)

[0108] in, It is the first corrected water level after the time series number i has been corrected. It is the first correction water level of the correction node before segmentation; It is the sequence number starting from the pre-segmentation correction node, and the pre-segmentation correction node's... The value is 1, and the water level correction calculation i starts from 2; It is the time-varying acceleration of the first adapted flow of the segmented correction node; It is the time-varying acceleration of the first adaptive flow of the correction node before segmentation; It refers to the duration of each segment; It is the first adapted traffic of the correction node before segmentation; It is the first adapted traffic of the segmented and corrected node; It is the first water level of the correction node before segmentation; These are preset parameters;

[0109] The pre-correction node positioning method for the second basic type and the eighth extended type is as follows: the time-varying velocity VQP1 of the first adaptive flow QP1 continuously increases to a local maximum value and the time-varying acceleration aQP1 of the first adaptive flow QP1 is greater than 0, or the time-varying acceleration aQP1 of the first adaptive flow QP1 is a local maximum value and the time-varying velocity VQP1 of the first adaptive flow QP1 is greater than 0.

[0110] The modified node positioning method for the second basic type and the eighth extended type is as follows: if the time from the node before segmented correction to the flow-level relationship control node does not exceed the preset minimum time, then it is positioned as the flow-level relationship control node, or the time-varying velocity VQP1 of the first adapted flow QP1 is closest to 0 and the first adapted flow reaches a local maximum value for positioning.

[0111] The method for determining the second basic type and the eighth extended type is as follows: if the difference between the first water level of the node before segmentation correction and the first correction water level is greater than the preset correction threshold, then the segmentation is the eighth extended type; otherwise, the segmentation is the second basic type.

[0112] The calculation process of the water level correction algorithm corresponding to the second basic type and the eighth extended type is as follows:

[0113] The two calculation parameters of formula (20) are determined by using three consecutive time points before the node before piecewise correction and the node after piecewise correction, for a total of four fitting points. and The first three points use the first adapted flow rate and the first corrected water level. The fourth point uses the first adapted flow rate and the first water level of the node after segmented correction, plus the water level difference before and after the segmented correction, to determine the parameters. and Then, the first corrected water level of each time point in the segment is calculated using formula (20);

[0114] (20)

[0115] Calculated based on the first corrected water level at the first and second points after the node before segmented correction. :

[0116] (twenty one)

[0117] in, and It is the first corrected water level of the first and second points after the correction node before segmentation;

[0118] If the first correction water level of the correction node before segmentation is corrected If the first corrected water level in the segment is not smooth, the solution is as follows:

[0119] Using the first corrected water level at three points—the node before the piecewise correction and the two points before the node before the piecewise correction—the following calculations are performed. :

[0120] (twenty two)

[0121] if ,

[0122] in, and It is the first corrected water level of the point before the node before the segmented correction and the second point before the node; It is the first corrected water level of the node before segmented correction;

[0123] Calculate the first corrected water level for the four water level fitting points:

[0124] (twenty three)

[0125] in, It is the first water level of the node before segmented correction; It is a preset correction threshold; It is the first corrected water level of the node before segmented correction; and It is the first corrected water level at the point one and two points before the node before the segmented correction; It is the first corrected water level of the node after segmented correction.

[0126] Using the water levels at four fitting points, the two parameters of formula (20) are recalculated. and Then, the first corrected water level of each time point in the segment is calculated using formula (20).

[0127] The types of rope loops, from front to back, include rope loop front type, rope loop inlet type, rope loop apex type, and rope loop outlet node;

[0128] The method for locating the rope loop nodes is as follows: First, determine the rope loop vertex node, where the time-varying acceleration aQP1 of the first adaptive flow rate QP1 changes from negative to positive and is closest to 0; then, determine the rope loop inlet node, where the time-varying velocity VQP1 of the first adaptive flow rate QP1 moving forward from the rope loop vertex node changes from positive to negative and is closest to 0; determine the rope loop front node, where the time-varying acceleration aQP1 of the first adaptive flow rate QP1 moving forward from the rope loop inlet node reaches a local maximum value and is greater than 0, and the time-varying acceleration aQP1 subsequently changes from positive to negative and gradually decreases, or the local maximum value point of the flow rate is predicted; finally, determine the rope loop outlet node, where the time-varying velocity VQP1 of the first adaptive flow rate QP1 moving backward from the rope loop vertex node changes from negative to positive and is closest to 0.

[0129] The segmentation of the rope loop type is from the rope loop inlet node to the rope loop entrance node. The correction algorithm corresponding to the rope loop type, and the water level correction calculation formula are as follows:

[0130] (twenty four)

[0131] in, It is the time-varying acceleration of the first adaptable flow at the rope loop inlet node; It is the time-varying acceleration of the first adaptive flow at the front node of the rope loop; It refers to the duration of each segment; It is the first adaptive flow of the node before the rope loop before the segmentation correction node; It is the first compatible flow rate at the rope loop inlet node; It is the first corrected water level at the node before the rope loop; It is the first water level at the node before the rope loop; These are preset parameters;

[0132] The segmentation of the rope loop inlet type is from the rope loop inlet node to the rope loop vertex node. The correction algorithm corresponding to the rope loop inlet type, and the water level correction calculation formula are as follows:

[0133] (25)

[0134] in, It is the time-varying acceleration of the first adapted flow at the vertex node of the loop; It is the first adaptive flow of the loop vertex node; It is the first corrected water level at the rope loop inlet node; It is the first water level at the rope loop inlet node; It refers to the duration of each segment; These are preset parameters;

[0135] The segmentation of the rope loop vertex type is from the rope loop vertex node to the rope loop exit node. The correction algorithm corresponding to the rope loop vertex type, and the water level correction calculation formula are as follows:

[0136] (26)

[0137] in, It is the time-varying acceleration of the first adaptable flow rate at the rope loop outlet node; It is the first compatible flow rate at the rope loop outlet node; It is the first corrected water level at the top node of the rope loop; It is the first water level at the top node of the rope loop; It refers to the duration of each segment; These are preset parameters.

[0138] Optionally, step S12 uses the same correction node positioning method, water level correction type, and correction algorithm as step S9. When calculating the second corrected water level in step S12, the first adapted flow rate is replaced with the second adapted flow rate, the time-varying velocity of the first adapted flow rate is replaced with the time-varying velocity of the second adapted flow rate, the time-varying acceleration of the first adapted flow rate is replaced with the time-varying acceleration of the second adapted flow rate, the first water level is replaced with the second water level, and the first corrected water level of the calculated result is replaced with the second corrected water level.

[0139] Optionally, in step S10, the formula for calculating the first corrected water level is:

[0140] (28)

[0141] in, It is the sequence number of the time point starting from the flow-water level relationship control node, and the flow-water level relationship control node... =1; It is the first correction water level of time series number i. The calculation of i starts from 2; It is the first adapted traffic for time sequence number i; It is the first adapted flow rate of the flow-water level relationship control node; It is the first corrected water level of the flow-water-level relationship control node; These are preset parameters.

[0142] Optionally, step S13 uses the same calculation formula as step S10. When calculating the second corrected water level in step S13, the first adapted flow rate is replaced with the second adapted flow rate, and the first water level is replaced with the second water level. The parameters used in formula (28) are different from those in step S10. The first corrected water level of the calculation result is replaced with the second corrected water level.

[0143] Optionally, step S14 includes:

[0144] Based on the first adapted flow rate and the corresponding first corrected water level, and the second adapted flow rate and the corresponding second corrected water level at each time point, the time series values ​​of the two wavefront water level gradient processes are obtained by back-calculating using the Manning formula. The Manning formula is as follows:

[0145] (29)

[0146] in, This refers to the corresponding traffic, which is either the first or second adapted traffic. It is the cross-sectional roughness; It is the cross-sectional water flow area, which is obtained from the relationship table between cross-sectional water level and water flow area based on the first or second corrected water level. It is the hydraulic radius of the cross section, which is obtained from the relationship table between the cross section water level and the hydraulic radius based on the first or second corrected water level. It is the water surface gradient of the corresponding flow rate and the corresponding corrected water level.

[0147] Beneficial Effects: This invention enables real-time acquisition of hydrological observation data and hydrological forecast flow data for forecast sections in the middle and lower reaches of small and medium-sized rivers. It rapidly calculates hourly water levels at the sections, and the forecast flow can adaptively generate two flood wave front-adaptive flow process lines based on a preset flow-water level relationship. Furthermore, it uses the time-varying acceleration of the flow as a key element to correct the water level and obtain two water level process lines, significantly improving the accuracy of water level forecasts. The method of this invention is easy to program and can perform calculations at the second level. Compared to existing methods for calculating water levels from forecast flow, it not only significantly improves the forecast accuracy of flood wave front water level processes but also outputs more data elements, which is beneficial for flood forecast consultation and effectively enhances the decision support capabilities for flood forecasting and early warning. Attached Figure Description

[0148] Figure 1 This is a flowchart of one method of the present invention;

[0149] Figure 2 This is a large cross-sectional view of the predicted section in Embodiment 1 of the present invention;

[0150] Figure 3 This is a diagram showing the relationship between the predicted cross-sectional water level and the water flow area in Embodiment 1 of the present invention;

[0151] Figure 4 This is a diagram showing the relationship between the predicted cross-sectional water level and the wetted perimeter in Embodiment 1 of the present invention;

[0152] Figure 5 This is a diagram showing the relationship between the predicted cross-sectional water level and the hydraulic radius in Embodiment 1 of the present invention. Detailed Implementation

[0153] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, so as to better understand the purpose, features and advantages of the present invention. It should be understood that the embodiments shown in the drawings are not intended to limit the scope of the present invention, but are only for illustrating the essential spirit of the technical solution of the present invention.

[0154] In the following description, certain specific details are set forth for the purpose of illustrating various disclosed embodiments in order to provide a thorough understanding of the various disclosed embodiments. However, those skilled in the art will recognize that embodiments may be practiced without one or more of these specific details. In other instances, well-known techniques associated with this application may not have been shown or described in detail to avoid unnecessarily obscuring the description of the embodiments.

[0155] Throughout this specification, references to "an embodiment" or "an embodiment" indicate that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Therefore, the appearance of "in an embodiment" or "an embodiment" in various places throughout the specification does not necessarily refer to the same embodiment. Furthermore, a particular feature, structure, or characteristic may be combined in any manner in one or more embodiments.

[0156] In the following description, in order to clearly demonstrate the structure and operation of the present invention, a number of directional terms will be used. However, terms such as "front", "back", "left", "right", "outside", "inside", "outward", "inward", "up", and "down" should be understood as convenient terms and not as limiting terms.

[0157] Reference Figure 1 This invention provides a method for calculating wavefront water level in forecasting sudden floods, which includes the following steps:

[0158] S1, real-time acquisition of hydrological observation and forecast data, and determination of fixed nodes for wavefront water level forecast calculation: real-time acquisition of hydrological observation and forecast data for each flood event, and determination of fixed nodes for wavefront water level forecast calculation. The hydrological observation and forecast data include observed water level, observed flow rate and forecasted flow rate. The fixed nodes, from front to back, include the starting point, the flow rate-water level relationship control node and the calculation end point.

[0159] The timing for initiating real-time water level forecasting operations is determined by rainfall, water level, and flow rate observation data. Based on the forecast flow rate for each flood event, the hourly forecast flow rate Q (VQ and aQ) is calculated. Following the method for determining the starting point (node ​​0), the calculation ending point (node ​​98), and the flow-water level relationship control node (node ​​99), the start time, end time, and flow-water level relationship control node for the flood wave front water level forecasting calculation can be determined. Therefore, this invention first determines the forecast calculation start time, i.e., the starting point; then determines the forecast calculation end time, which is also the time when the maximum flow rate occurs before the wave front, i.e., the calculation ending node; and finally determines the flow-water level relationship control node.

[0160] In one embodiment, after acquiring hydrological observation and forecast data for a flood event, the time-varying velocity and time-varying acceleration of the hydraulic elements of the flow rate are first calculated.

[0161] The formula for calculating the time-varying velocity VQ of flow rate Q is:

[0162] (1)

[0163] The formula for calculating the time-varying acceleration aQ of flow rate Q is:

[0164] (2)

[0165] in, It is time-series flow Data sequence number starting from 1 (referred to as time point number); It is the time difference between two adjacent data points; The value of i is calculated starting from 2. The value of i is calculated starting from 3.

[0166] The specific methods for locating the starting point, the calculation end point, and the control points for the flow-water level relationship are as follows:

[0167] S101, the starting point (node ​​0) is determined by the observed flow and the forecast flow. If the forecast flow reaches the preset minimum control flow earlier than the observed flow, the time point when the forecast flow reaches the minimum control flow is taken as the starting point. Otherwise, the time point when the observed flow is not less than the minimum control flow and the forecast flow reaches 50% of the minimum control flow is taken as the starting point.

[0168] S102, after the starting point, the maximum value of the predicted flow rate Q is taken as the end node of the calculation (node ​​98). At this time, the time-varying velocity VQ of the predicted flow rate changes from negative to positive and is closest to the value of 0.

[0169] S103, starting from the end of the calculation and moving backward, the time point when the time-varying acceleration aQ of the predicted flow rate Q first turns positive and is closest to 0 is taken as the flow-water level relationship control node (node ​​99). If the time between the end of the calculation and the flow-water level relationship control node is less than the preset minimum time for cross-sectional hydrological analysis, then the flow-water level relationship control node is located according to the preset minimum time between the end of the calculation and the control node. In other words, the time between the end of the calculation and the control node must not be less than the preset minimum time. The preset minimum time is the hydrological characteristic value of the predicted cross-section.

[0170] S2, determine whether the initial forecast flow needs to be corrected: if no correction is needed, proceed to step S4; otherwise, proceed to step S3.

[0171] When a new flood front reaches the forecast section, the flow-water level relationship exhibits polymorphism due to the combined effects of downstream flood backwater and backwater fluctuations; that is, it does not have a fixed, single-value correspondence. Therefore, the initial flow correction is a correction for the flow rate during the initial stage of polymorphic instability of the additional gradient at the flood front.

[0172] The time elapsed between the initial hydrological assessment and the starting point is determined by the hydrological characteristics of the forecast section. This time should not be shorter than the shortest time elapsed for the hydrological trend analysis of the section, dTinit. dTinit is a constant determined by the underlying surface conditions of the catchment area of ​​the forecast section and the channel size of the forecast river segment. Based on the functional relationship between flow rate and water level, and considering the changing trends of the observed flow rate process line, observed water level process line, and forecast flow rate process line within the above two time periods, in step S2, if the changing trends of the observed water level and the forecast flow rate are inconsistent (deviation), or if the changing trends of the observed flow rate and the forecast flow rate are inconsistent (deviation), then the forecast flow rate needs to be corrected to ensure that the initial forecast flow rate conforms to the functional relationship between flow rate and water level; otherwise, no correction is required.

[0173] S3, Determine the control node for initial traffic correction: The control node for initial traffic correction includes the control node before initial traffic correction and the control node after initial traffic correction.

[0174] Starting from the starting point, locate several control nodes on the forecast flow step by step from front to back. When the first node type is the fourth rising type, the ID of the corresponding node type algorithm is 34. The control node after this node type is the control node after the initial flow correction. At this time, it is necessary to locate the control node before the initial flow correction. In other cases, it is not necessary to locate the control node before the initial flow correction.

[0175] In one embodiment, in step S3, the control node uses the following positioning method before the initial flow correction:

[0176] 1) During the period between the control node and the starting point before the initial flow correction, the changes and trends of the observed water level and observed flow are consistent;

[0177] 2) During the period between the control node and the starting point before the initial flow correction, the time-varying velocity VQ and time-varying acceleration aQ of the forecast flow Q remain unchanged, that is, neither of the two elements has local extreme values;

[0178] Among them, the time-varying velocity VQ and the time-varying acceleration aQ of the predicted flow rate Q have many peaks and troughs. This local extremum refers to the absence of peaks and / or troughs in the time period between the control node and the starting point before the initial flow correction.

[0179] 3) The time between the control node and the starting point before the initial flow correction shall not be less than the preset minimum time dTinit;

[0180] 4) Starting from the starting point, prioritize the time point at which the sign of either the time-varying velocity VQ or the time-varying acceleration aQ of the predicted flow Q changes as the control node before the initial flow correction.

[0181] In one embodiment, in step S3, the control node after initial flow correction is located at the last of the following three time points. That is, the control node after initial flow correction is located at whichever of the following time points is at the end of the time:

[0182] 1) When the predicted flow at the starting point is less than the preset minimum control flow, the first point where the predicted flow is closest to the minimum control flow is taken as the control node after the initial flow correction; where the minimum control flow at the starting point, Qkmin, is a constant determined by the underlying surface conditions of the catchment area of ​​the predicted section and the scale of the predicted river section. Affected by factors such as the river regulation and storage effect, the predicted flow is less than Qkmin data, which is unreliable.

[0183] 2) After the starting point, the time-varying velocity VQ and time-varying acceleration aQ of the predicted flow Q remain unchanged. The time point at which either the time-varying velocity VQ or the time-varying acceleration aQ of the predicted flow Q first shows a positive-to-negative value transition point is used as the control node after the initial flow correction.

[0184] 3) After the initial flow correction, the time between the control node and the starting point shall not be less than the preset minimum time dTinit.

[0185] S4, Calculate the suitable flow before the flow-water level relationship control node: Starting from the starting point to the flow-water level relationship control node, locate several flow conversion control nodes on the forecast flow step by step from front to back. Two adjacent flow conversion control nodes form a segment, and the forecast flow is divided into one or several segments. The node type of the control node before each segment corresponds to a preset node type algorithm for converting the forecast flow to the suitable flow. The forecast flow Q before the flow-water level relationship control node is converted into the suitable flow QP through the node type algorithm.

[0186] Different node types correspond to corresponding node type algorithms. There is a one-to-one correspondence between node types and algorithms. The corresponding algorithm ID is a two-dimensional positive integer, which is marked on the control node before the segmentation of the traffic conversion segment.

[0187] The flow conversion control nodes include three fixed nodes: the starting point, the calculation end point, and the flow-level relationship control node, all of which participate in the segmentation. Each segment has a pre-segment control node and a post-segment control node. Segments are divided using these pre-segment and post-segment control nodes, and the pre-segment control node of the current segment is the post-segment control node of the previous segment. The pre-segment control node of the first segment is the starting point. Starting from the starting point, the post-segment control nodes of each segment are located progressively, and the node type is determined.

[0188] In addition to the three fixed nodes, the traffic conversion control nodes also include the following types of control nodes and their corresponding algorithms: 1) Initial traffic correction control node, no corresponding algorithm; 2) Steady progress node, corresponding to algorithms ID11 and ID12; 3) Fluctuation node, corresponding to algorithms ID21 and ID22; 4) Rising node, corresponding to algorithms ID31, ID32, ID33, ID34, and ID39.

[0189] Due to the influence of river channel storage and other factors, the VQ of the predicted flow rate Q is usually systematically overestimated. Although the VQ and aQ trends of the predicted flow rate Q within each segment are relatively stable, the trends of predicted flow rate and water level changes, as well as the flow-water level relationship, do not completely conform to the hydrological characteristics of the predicted cross-section. The algorithm for adapting the flow rate conversion corrects the time-varying velocity of the predicted flow rate using a proportionality coefficient k (multiplied by a constant, 0.2 < k ≤ 1), and then calculates the adapting flow rate for this segment based on the adapting flow rate of the preceding control node. The method for calculating the adapting flow rate for each segment is based on correcting the VQ of the predicted flow rate Q according to the flow-water level relationship. For each segment, the predicted flow rate is corrected... Calculate the appropriate traffic The general calculation formula is:

[0190] (3)

[0191] in, It is the sequence number of the time series starting from the node before the segmentation, and the starting point =1; It is the adapted flow of time series number i, and the adapted flow of the starting point is the observed flow; It is the time-varying velocity of the predicted flow rate at time point i; It is the proportional coefficient for adjusting the time-varying velocity of the point-by-point forecast flow.

[0192] In one embodiment, the type of the flow conversion control node includes at least one of the following: steady type, fluctuating type, and rising type.

[0193] 1. Steady-progressive type: Primarily used for flow correction during periods when water level forecasts have not undergone initial flow correction. Within the steady-progressive segment, the VQ and aQ trends of the forecast flow Q are relatively stable. However, the trends of forecast flow and water level changes do not perfectly match the hydrological characteristics of the forecast section. The algorithm for adapting the flow is based on the time-varying velocity of the forecast flow according to the proportionality coefficient k. i After making corrections, the adaptive traffic is calculated forward based on the adaptive traffic of the control nodes before segmentation.

[0194] The steady-progress type includes at least one of the first steady-progress type and the second steady-progress type.

[0195] The method for determining the steady-progression type is as follows: starting from the starting point or the control node before segmentation, the time point where the time-varying velocity VQ value of the predicted flow Q changes from positive to negative and is closest to 0 is taken as the subsequent control node for the steady-progression type. This change in positive to negative value includes both negative to positive and positive to negative.

[0196] The first method for judging the steady-progress type is: when the duration of the segment in which the steady-progress type is located exceeds the cross-sectional runoff history (the hydrological characteristics of the forecast cross-section, a constant) and the forecast flow rate within the segment is consistent with the changing trends of the observed water level and observed flow rate.

[0197] The node type algorithm ID11 corresponding to the first steady-progress type is the time-varying velocity adjustment ratio coefficient of the predicted flow within the segment. The formula for calculating the adaptive traffic is:

[0198] (4)

[0199] in, It is the timing point number starting from the control node before segmentation, and the control node before segmentation... =1, The time series index i is calculated starting from 2; It is the adapted flow of time series number i, and the adapted flow of the starting point is the observed flow; It is the time-varying velocity of the predicted flow rate at time point i.

[0200] The second method for judging the steady-progress type is as follows: When the duration of the segment containing the steady-progress type exceeds 1.5 times the cross-sectional confluence time and the observed water level and flow rate within the segment show consistent trends, but the time-varying velocity VQ of the predicted flow rate Q is systematically larger than the flow rate-water level function relationship, the systematic deviation is determined by the predicted flow rate for the corresponding time period. The flow rate adjustment ratio can be adaptively determined based on the predicted flow rates of the control nodes before and after the forecast. The flow rate adjustment ratio for each point within the segment is... equal.

[0201] The second stable type corresponds to node type algorithm ID12, which calculates the segmented time-varying velocity adjustment ratio coefficient based on the forecasted flow. Then, the adaptive traffic is calculated point by point. The calculation formula is as follows:

[0202] (5)

[0203] in, It is the timing point number starting from the control node before segmentation, and the control node before segmentation... =1; It is the predicted flow rate of the last time series point within the segment. It is the predicted flow rate of the first time series point within the segment; It is the adapted traffic for time sequence number i. The time series index i is calculated starting from 2; It is the time-varying velocity of the predicted flow rate at time point i; These are preset parameters.

[0204] 2. Fluctuation Type: The flow rate during a flood rise is determined by factors such as rainfall runoff, river storage capacity, reservoir and other water conservancy project scheduling, and downstream boundary conditions. The wave flow rate during a flood rise cannot always rise unidirectionally; there are periods of stable flow and periods of decreasing flow. Whether the flow rate and water level are in a "loop" is determined by the wave flow rate. When the wave flow rate decreases, the decrease in the flow rate adapted to the compensation operation for changes in the water surface gradient should be systematically reduced relative to the forecast flow rate. In fact, if the flow rate and water level trend in the previous section of the control node is rising, the trend of water level change in this section may not even be reversed compared to the previous section. Therefore, the adapted flow rate for calculating the water level requires corresponding matching algorithms and model parameters to calculate the water level using the flow rate to calculate the water level function relationship.

[0205] The volatility type includes at least one of the first volatility type and the second volatility type.

[0206] The method for determining the first type of fluctuation is as follows: within a segment, the predicted flow first increases, then decreases, and then increases again, while the adaptive flow maintains an upward trend within the segment. The hydraulic characteristics begin from the control node before the segment: the time-varying velocity VQ of the predicted flow Q changes from positive to negative and then back to positive; the time-varying acceleration aQ of the predicted flow Q changes from negative to positive and then back to negative; the predicted flow first increases, then decreases, and then increases again; the absolute value of the flow difference in the decreasing segment is less than the flow difference between the two increasing segments.

[0207] The post-control node for the first wave type is located at the time point when the time-varying velocity VQ of the forecast flow Q falls to the point closest to 0.

[0208] The node type algorithm ID21 corresponding to the first fluctuation type uses the time series point number within the segment as the independent variable and employs an exponential function to calculate the adaptive flow of each time series point within the segment. The calculation formula is as follows:

[0209] (6)

[0210] in, The sequence number within the segment starts from 1, which is the sequence number of the control node before the segmentation. It is the traffic adapted for i6 timing point sequence numbers. The calculated time series index i6 starts from 2; and These are the two parameters of the exponential function.

[0211] Formula (4) is used to calculate the adaptive flow at the first two time points, the middle time point, and the last two time points within a segment. The adaptive flow at these five points is used to determine the two parameters of the adaptive flow exponential function for segmented calculation. and Then, calculate the appropriate flow rate at each point in the segment according to formula (6). .

[0212] The method for determining the second type of fluctuation is as follows: the adaptive flow rate within the segment gradually decreases relative to the flow rate of the preceding node. The hydraulic characteristics are as follows: starting from the control node before the segment, the time-varying velocity VQ of the predicted flow rate Q first decreases and turns negative, then gradually increases and turns positive; the time-varying acceleration aQ of the predicted flow rate Q gradually increases and turns positive. Preferably, if the subsequent control node of the next segment is a flow-level relationship control node, then this segment can only be classified as the second type of fluctuation and cannot be treated as a multi-peak flood flow-level loop.

[0213] The second type of wave's post-control node is located at the last of the following three reference timing points. In other words, whichever timing point is at the end of the time frame is the post-control node:

[0214] 1) Flow location: The time point at which the forecast flow, after decreasing, rebounds to the point where it first exceeds the control node before the segmentation;

[0215] 2) VQ positioning: After the forecast flow drops to the minimum value (negative number) within the segment and then rises, the position is located at the time series point where the time-varying velocity VQ of the forecast flow Q is dTpf hours before the sign turns positive; dTpf is the hydrological characteristic value of the forecast section;

[0216] 3) aQ location: the time series point where the time-varying acceleration aQ of the forecast flow Q increases to a positive local maximum; among them, the time-varying acceleration aQ process line of the forecast flow Q has many peaks and troughs, and the local maximum here is the time series point where the peak of the time-varying acceleration aQ is closest to when the time-varying acceleration aQ increases to a positive value.

[0217] The node type algorithm ID22 corresponding to the second fluctuation type, and the flow time-varying velocity adjustment ratio coefficient for each point within the segment. They are different. Based on the maximum forecast flow within each segment. Minimum flow Calculate the time-varying velocity adjustment coefficient of the flow rate point by point. The formula for calculating the appropriate traffic volume is as follows:

[0218] (7)

[0219] in, It is the timing point number starting from the control node before segmentation, the control node before segmentation. =1; It's about adapting to traffic volume. The calculation of i starts from 2; It is a forecast of the flow rate; It is the time-varying velocity of the forecast flow; It is the minimum forecast flow within the segment; It is the maximum predicted flow rate within the segment; and These are preset parameters;

[0220] 3. Type of Rise: The forecast flow, the adapted flow, and the forecast water level all rise continuously within the segment. The type of rise includes at least one of the following: Type 1, Type 2, Type 3, Type 4, and Type 9.

[0221] The methods for determining the first, third, and ninth types of water level rises are as follows: starting from the control node before the segmentation, the forecasted flow rate continuously increases, and the water surface gradient continues to widen. Because the forecasted flow rate continuously increases and the water surface gradient continues to widen within the segment, the water level calculated based on the flow-water level relationship does not match the upward trend of the flow rate and water level. Continuous rise refers to a situation where there is no decline forming a peak.

[0222] The methods for locating the post-control nodes of the first, third, and ninth upward trends are as follows:

[0223] 1) If the time between the control node before segmentation and the calculation end node does not exceed twice the preset minimum time dTinit, then the location is at the calculation end node;

[0224] 2) Otherwise, the location is determined by the time series point where the time-varying velocity VQ of the predicted flow rate Q drops to its maximum value (positive value) and is closest to 0.

[0225] 3) The segment duration should exceed the cross-sectional flow duration. If the segment duration does not meet this requirement, other node types and corresponding algorithms should be selected.

[0226] The further judgment method for the first, third, and ninth rise types is as follows: First, use formula (4) to calculate the initial value of the adapted flow of the control node after segmentation and the initial value of the adapted flow of the control node before segmentation. Then, determine the node type based on the forecast flow and the initial value of the adapted flow. 1) If the relative error of the initial value of the adapted flow of the control node before segmentation to the forecast flow exceeds 5%, the segmentation is the ninth rise type; 2) If the above conditions are not met, but the relative error of the initial value of the adapted flow of the control node after segmentation to the forecast flow does not exceed 5%, the segmentation is the third rise type; 3) The segmentation that does not meet condition 2) is the first rise type.

[0227] The node type algorithm IDs are 31 for the first type of increase, 33 for the third type, and 39 for the ninth type. The flow calculation process for these three types of algorithms is as follows:

[0228] First, determine 6 fitting points for the adaptive flow calculation. The determination method is as follows: the control node before segmentation, a time series point after the control node before segmentation, and 4 consecutive time series points starting from the m-th point forward from the control node after segmentation.

[0229] The method for calculating m is as follows:

[0230] (8)

[0231] in, It is the time sequence number of the last point of the time sequence data starting from the control node time sequence number before the segmentation within the segment, which is the segment duration within the segment plus 1;

[0232] Calculate the time-varying velocity adjustment ratio coefficients for the three types of flow increases mentioned above. :

[0233] (9)

[0234] in, It is the predicted flow of the control nodes before segmentation; It is the predicted flow of the control nodes after segmentation; It is the adaptation traffic of the control node before the segmentation control point; It is the adaptation traffic of the control node before segmentation, that is, the known adaptation traffic of the control node after the previous segmentation.

[0235] according to Calculate the fitted flow at the fitting point for the 6 fitted flows within the segment:

[0236] (10)

[0237] Finally, based on the time sequence number of the fitting point from the control node before segmentation and the formula (10), the fitting flow of the fitting point is calculated. After determining the two parameters of the fitting flow calculation formula (6) based on the fitting point, the fitting flow of each time sequence point in the segment, except for the control node before segmentation, is recalculated based on the formula (6).

[0238] The second method for judging the type of rise is: the control node before the segment is the outlet of the flow-water level curve loop, or the control node after the fluctuation type, which predicts that the flow will continue to rise.

[0239] The second type of rise type of post-control node positioning method is as follows: 1) If the time between the pre-control node and the calculation end node does not exceed twice the preset minimum time, then position it at the calculation end node; 2) Otherwise, position it at the time series point where the value of the time-varying velocity VQ of the forecast flow first appears (positive) after the pre-control node and then drops to the sign change of the value closest to 0.

[0240] The node type algorithm ID32 corresponding to the second type of rise is that the flow difference of the adapted flow within the segment is equal to the flow difference of the predicted flow, the adapted flow of the control node before the segment is equal to the adapted flow of the control node after the segment of the previous segment, and the adapted flow is calculated using formula (4).

[0241] The fourth type of rise is the forecast initial flow correction algorithm. The control node before segmentation is the starting point. According to step S3, the control node before initial flow correction 19 and the control node after initial flow correction are determined. The observed flow from the control node before segmentation 19 to the starting point is used. An autoregressive moving average model is used to construct a cubic polynomial and extrapolate to calculate the adapted flow from the starting point to the control node after segmentation.

[0242] Among them, the autoregressive moving average model is an autoregressive moving average model used to predict the trend and change pattern of time series, and the ARIMA (Autoregressive Integrated Moving Average) model is preferred.

[0243] Step S4, based on the above method, determines the control node, node type, calculation segments, and corresponding algorithms for the forecast flow to adapt flow calculation, thus forming the calculation scheme for forecast flow Q to adapt flow QP. Since the adapt flow at the starting point is an observed value, the adapt flow QP can be calculated segment by segment from the starting point to the end.

[0244] S5, Calculate the adapted flow between the flow-level control node and the end node: Convert the forecast flow between the flow-level control node and the end node into an adapted flow. The time-varying rate of the adapted flow between the flow-level control node and the end node is the same as that of the forecast flow, that is, the adjustment ratio coefficient of the time-varying rate of the forecast flow is k=1. Therefore, for the corresponding node type algorithm ID11, the adapted flow between the flow-level control node and the end node can be calculated using formula (4).

[0245] S6 filters the adapted traffic and converts it into the first adapted traffic.

[0246] In one embodiment, the adaptive flow QP may lose its smoothness at the flow segmentation node. Therefore, a 3-point numerical filtering algorithm is used to filter the adaptive flow QP and convert it into a first adaptive flow QP1, and the corresponding time-varying velocity VQP1 and time-varying acceleration aQP1 are calculated based on the first adaptive flow QP1.

[0247] The calculation formula for the 3-point numerical filtering algorithm is as follows:

[0248] (11)

[0249] in, It is the first adaptive flow at the starting point; It is the adaptive traffic at the starting point; It is the sequence number of the time series starting from the starting point, and the starting point =1; It is the first adapted traffic for time sequence number i; It is the adapted traffic for time sequence number i; It is the first adapted traffic for the calculation of the end node; It calculates the adaptation traffic for the end node.

[0250] S7 converts the first adaptation traffic to the second adaptation traffic.

[0251] Analysis of historical flood observation data shows that the smaller the cross-sectional flow in the middle and lower reaches of small and medium-sized rivers, or the smaller the water surface gradient of the river channel, the closer the flow-water level relationship is to a linear relationship, and the flood characteristics are close to steady flow. As the cross-sectional flow increases and the water surface gradient increases, the water level-flow relationship may exhibit a counterclockwise loop shape at the peak of the flood. From cross-sectional forecasting records, there is usually a systematic deviation between the maximum flood peak flow, the compiled flow rate, and the forecast flow rate. Generally, the compiled flow rate is smaller than the forecast flow rate. Since the maximum flood peak flow rate of the first adapted flow rate QP1 is basically the same as the forecast flow rate Q, it is necessary and practical to calculate another new wavefront flood indication flow process line QP2 based on the first adapted flow rate QP1.

[0252] The basic requirements for converting the first adapted flow rate QP1 to the second adapted flow rate QP2 are as follows: (1) The starting point flow rates QP2 and QP1 are the same; (2) The maximum peak flow rate of QP2 is closer to the total flow rate than that of QP1; (3) Further reduce the influence of the additional gradient of the water surface on the calculated water level.

[0253] In one embodiment, a logarithmic function is used, with the maximum peak flow of the flood forecast for each flood event as the control factor and the normalized QP1 flow as the independent variable, to calculate the point-by-point correction ratio coefficient of VQP1. The first adapted flow rate QP1 and the second adapted flow rate QP2 at the starting point are equal. The time-varying flow rate adjustment coefficient for the first adapted flow rate QP1 is calculated accordingly. The second adaptive flow rate QP2 is calculated by forward propagation. The calculation formula is:

[0254] (12)

[0255] in, and These are the preset parameters; It is the sequence number of the timing point within the segment, starting from the timing point number of the control node before the segmentation, which is 1; It is the first adapted traffic for time sequence number i; It is the first adapted traffic for the calculation of the end node; It is the first adapted flow at the starting point, which is also the observed flow; It is the second adaptive flow at the starting point; It is the second adapted traffic for time sequence number i. The calculation of i starts from 2; It is the time-varying velocity of the first adapted flow of time sequence number i.

[0256] S8, Calculate the first water level between the starting point and the flow-water level relationship control node: Calculate the first water level H1 between the starting point and the flow-water level relationship control node based on the first adapted flow rate.

[0257] Determining the flow-water level relationship requires at least two points. Many functions can be chosen to establish the relationship between forecast sections QP1 and QP2 and the corresponding single-valued water levels, among which establishing a linear relationship using two points is the simplest and most practical method. Since the starting point has hydrological flow and water level observation data, it serves as the control node for flow-water level calculation. Analysis of compiled hydrological data from forecast sections in the middle and lower reaches of small and medium-sized rivers shows that the point where the peak flow occurs is a convex point on the loop curve, making it unsuitable as a control node for water level calculation. The aQ value of the flow-water level relationship control node is close to zero before the peak flow occurs, and the flow-water level relationship for floods of different magnitudes is relatively stable. Therefore, using the flow-water level relationship control node as the control node for flow-water level calculation is superior to using the end point of the maximum flow calculation for a given flood event.

[0258] The functions for calculating the corresponding water levels of the flow-water-level relationship control nodes are the same for QP1 and QP2, but the parameters for calculating the corresponding water levels for QP1 and QP2 are different and related to the flood magnitude of each flood event. It is necessary to calculate QP1 and QP2 for the flow-water-level relationship control nodes of each historical flood event, and combine the integrated water levels of the flow-water-level relationship control nodes to determine the parameters of the formula for calculating the water levels of the flow-water-level relationship control nodes using the least squares method.

[0259] In one embodiment, the flood level is determined based on a preset flow boundary control value. The floods are divided into two magnitudes: for floods with a maximum forecast flow rate less than dQk, the flow-water-level relationship control node is calculated using a quadratic polynomial, while for floods with a flow-water-level relationship greater than dQk, the flow-water-level relationship control node is calculated using a power function.

[0260] The formula for the first water level of the control node in the QP1 calculation of the flow-water-level relationship is:

[0261] (14)

[0262] Among them, QP1 99 The first adaptive flow rate for the water level control node is calculated based on the flow rate, H1. 99 The corresponding first water level H1 for the flow rate calculation water level control node. 99 . , , , and They are QP1 99 Calculate the corresponding water level H1 99 The preset parameters of the calculation formula (14) are selected based on the first adaptive flow of node 99.

[0263] Based on the observed flow rate and water level at the starting point, and the flow rate, the appropriate flow rate QP1 for the water level control node is calculated. 99 and corresponding water level H1 99 By using a linear function, an adaptive function for calculating the water level relationship for a given flood event can be established. The formula for calculating the first water level H1 between the starting point and the control node for the flow-water level relationship is as follows:

[0264] (15)

[0265] Since step S8 calculates the first water level H1 based on the first adapted flow rate QP1 between the starting point and the flow-water level relationship control node, therefore It is the first watermark of time point i. It is the sequence number of the time series starting from the starting point, and the starting point =1, The calculation of i starts from 2; It is the adapted traffic for time sequence number i; It is the observed flow rate at the starting point; This refers to the observed water level at the starting point. Based on the starting point and the flow rate, the first adapted flow rate QP1 and the first water level H1 at the water level control node are calculated. This allows for the calculation of the first water level H1 between the starting point and the flow-water level relationship control node. i .

[0266] S9, the first water level between the starting point and the flow-water level relationship control node is corrected to generate the first corrected water level: from the starting point to the flow-water level relationship control node, several water level correction nodes are located step by step from front to back. Two adjacent correction nodes form a segment, thus dividing the calculated water level into one or several segments. The water level correction type of each segment's correction node corresponds to a preset correction algorithm for correcting the water level. The calculation method for segmented water level correction is marked on the first node of the segment, forming the first water level correction scheme; from the starting point to the flow-water level relationship control node, the corresponding corrected water level is obtained through the first water level correction of each segment.

[0267] The location of several water level correction nodes between the starting point and the flow-level relationship control node is based on the hydraulic element of the time-varying acceleration of the adapted flow rate. Since the time-varying accelerations of the two adapted flows, QP1 and QP2, are highly correlated, the correction nodes and algorithms for calculating the water level for adapted flows QP1 and QP2 are consistent. Therefore, only the location method, node type, and algorithm of the first water level H1 correction node need to be explained.

[0268] The types and matching algorithms (abbreviated IDs) of the flood wave front water level hydraulic factor correction nodes are marked on the preceding nodes of the water level correction calculation segment control. The following nodes are the first nodes of the algorithm for the next segment change and are the dividing points of the algorithm change. Fixed nodes are all water level correction nodes; starting from the starting point and ending at the flow-water level relationship control node, water level correction nodes are located step by step, and the calculation is pushed forward after segmentation.

[0269] The water level correction for a flood event is divided into single-peak floods and multi-peak floods. For a single-peak flood, two adjacent correction nodes form a segment, but for a multi-peak flood, the local flow extreme value loop segment is a multi-point combined integrated segment. After the water level correction nodes are determined, the calculation can be performed segment by segment from the starting point to the fixed node 99 to complete the calculation of the first water level H1 and generate the first corrected water level H3.

[0270] There are basic and extended algorithms for single-peak flood water level correction. The corrected water level is calculated based on the water level correction type of the correction node before segmentation. For multi-peak floods, a continuous multi-segment single-peak combination processing method is added to address the characteristics of flow-water-level rope segments.

[0271] The basic requirement for water level correction node positioning is: the correction node after the segment of the previous segment is the correction node before the segment of the next segment, and the H3 of the node after the segment of the previous segment is used as the H3 of the node before the segment of the next segment.

[0272] The algorithm classifications corresponding to the water level correction types are as follows:

[0273] 1) Basic algorithms: ID50, ID51, ID52;

[0274] 2) Extended algorithms, ID 57 and ID 58;

[0275] 3) Loop algorithm: ID 53, ID 54, ID 55, ID 56;

[0276] In one embodiment, the water level correction type includes at least one of the basic type, the extension type, and the rope type.

[0277] The basic types include at least one of the zero basic type, the first basic type, and the second basic type.

[0278] The method for determining the zero-based type is as follows: Starting from the starting point or the correction node before segmentation, if the first adapted flow rate QP1 before the flow-water level relationship control node continuously increases, that is, the flow rate of the flood event... If the change trends of the first adaptive flow and the first water level are the same within the preset shortest time before the segmentation correction node, then the post-correction node of the zero basic type is located at the flow-water level relationship control node.

[0279] The correction algorithm ID50 corresponding to the zero-based type has the following water level correction calculation formula:

[0280] (16)

[0281] Since step S9 calculates the first corrected water level H3 based on the first water level H1, therefore, It is the first corrected water level after the time series number i has been corrected. It is the first corrected water level of the node before segmentation, and it is a known quantity; It is the sequence number starting from the pre-segmentation correction node, and the pre-segmentation correction node's... The value is 1, and the water level correction calculation i starts from 2; It is the first watermark of time point i. It is the first water level of the node before segmentation; when the node before segmentation is the starting point. .

[0282] The extended types include the seventh extended type and the eighth extended type.

[0283] The method for determining the first basic type and the seventh extended type is as follows: the first basic type and the seventh extended type are located by the sign change of the time-varying acceleration aQP1 of the first adaptive flow QP1 being closest to 0 or a local extremum (aQP1 < 0 locates the local minimum, aQP1 > 0 locates the local maximum). After the first basic type and the seventh extended type continue to rise after the previous correction node, the first basic type and the seventh extended type are located by the local extremum of the time-varying acceleration aQP1 of the first adaptive flow QP1.

[0284] When the pre-segmentation correction node is within the preset shortest time period, relative to the pre-segmentation correction node, the slope of the first adapted flow rate and the first water level curve at each time point is greater than that in formula (15). If the value is 1.2, then the segment is the seventh extended type; otherwise, the segment is the first basic type.

[0285] The water level correction amount for the pre-segmentation correction node of the first basic type is 0m. The correction algorithm ID51 corresponding to the first basic type is as follows: The calculation formula for the first water level H1 correction is:

[0286] (18)

[0287] Since step S9 calculates the first corrected water level H3 based on the first water level H1, therefore, It is the first corrected water level after the time series number i has been corrected. It is the first corrected water level of the node before segmentation, and it is a known quantity; It is the sequence number of the time point starting from the starting point, and the preceding node's The value is 1, and the water level correction calculation i starts from 2; It is the time-varying acceleration of the first adapted flow of the segmented correction node; It is the time-varying acceleration of the first adaptive flow of the correction node before segmentation; It refers to the duration of each segment; It is the first adapted traffic of the correction node before segmentation; It is the first adapted traffic of the segmented and corrected node; It is the first water level of the node before segmentation; These are preset parameters;

[0288] In one example, for a scenario analysis of the first basic type: the predicted flow rate Q > QP1 of the previous correction node is not much different (flow rate ratio bp < 1.5). If it is the starting point, QP1 does not exceed twice the flow rate of the starting point. QP1 before node 51 (i.e. the previous correction node) is synchronized with the water level change trend and basically conforms to the flow rate-water level function relationship.

[0289] For the seventh expansion type, analyzing the changes in QP1 and H1 before segmentation, if the upward trend of water level H1 lags behind the matching flow rate QP1, meaning the relationship between water level and flow rate is affected by river regulation and the water level change is delayed, then the segmentation is the seventh expansion type. The water level correction amount at the correction node before segmentation for the seventh expansion type is 0.00m. The correction algorithm ID57 corresponding to the seventh expansion type, and the calculation formula for the first water level H1 correction, are as follows:

[0290] (19)

[0291] Since step S9 calculates the first corrected water level H3 based on the first water level H1, therefore, It is the first corrected water level after the time series number i has been corrected. It is the first corrected water level of the node before segmentation, and it is a known quantity; It is the sequence number starting from the pre-segmentation correction node, and the pre-segmentation correction node's... The value is 1, and the water level correction calculation i starts from 2; It is the time-varying acceleration of the first adapted flow of the segmented correction node; It is the time-varying acceleration of the first adaptive flow of the correction node before segmentation; It refers to the duration of each segment; It is the first adapted traffic of the correction node before segmentation; It is the first adapted traffic of the segmented and corrected node; It is the first water level of the node before segmentation; These are preset parameters.

[0292] In one example, for a scenario analysis of the seventh extension type: the flow rate of the QP1 algorithm is constrained by the predicted flow rate at local high points. The observed or predicted flow rate at node 57 (i.e., the node before the piecewise correction of the seventh extension type) is greater than or significantly different from QP1 (flow ratio bp > 1.5). Judging from the predicted flow rate Q, the adapted flow rate QP1, and the water level process before node 57, the water level change is affected by river regulation and lags behind the flow rate. This causes the buffering effect of the change in water surface gradient to push up the calculated water level of the subsequent nodes.

[0293] The pre-correction node positioning methods for the second basic type and the eighth extended type are as follows: (1) The time-varying velocity VQP1 of the first adaptive flow QP1 continuously rises to a local maximum value and the time-varying acceleration aQP1 of the first adaptive flow QP1 is greater than 0; or (2) The time-varying acceleration aQP1 of the first adaptive flow QP1 is a local maximum value and the time-varying velocity VQP1 of the first adaptive flow QP1 is greater than 0. Wherein, the local maximum value is the first peak of VQP1 that satisfies aQP1 greater than 0 or the first peak of aQP1 that satisfies VQP1 greater than 0, starting from the pre-correction node.

[0294] The modified node positioning methods for the second basic type and the eighth extended type are as follows: (1) If the time from the node before segmented correction to the flow-water level relationship control node does not exceed the preset minimum time, then it is positioned as the flow-water level relationship control node, or (2) The time-varying velocity VQP1 of the first adapted flow QP1 is closest to 0 and the first adapted flow reaches a local maximum value. Among them, the local maximum value is the first peak after VQP1 is closest to 0.

[0295] The method for determining the second basic type and the eighth extended type is as follows: if the difference between the first water level of the node before segmentation correction and the first correction water level is greater than the preset correction threshold dHk, then the segmentation is the eighth extended type; otherwise, the segmentation is the second basic type.

[0296] In one example, for a scenario analysis of the second basic type: the segmented adaptive flow QP1 increases, aQP1 decreases from its maximum value before the flood peak, and the corrected node flow relative to the forecast section represents a large-scale flood, with a stable flow-level relationship. The node water level correction before correction is no more than 0.10m, and the corrected node water level correction after correction is 0.00m, including the correction for the flow-level relationship control nodes.

[0297] In one example, a scenario analysis for the eighth extension type: the main problem is that the predicted flow rate QP1 before node 58 (the node before correction for the eighth extension type) is too large after correction, especially since this section is the rope outlet section. The water level correction of the node before correction exceeds dHk. If the water level correction of the node after correction is 0.00m, the water surface gradient is obviously unreasonable.

[0298] The calculation process of the water level correction algorithm corresponding to the second basic type and the eighth extended type is as follows:

[0299] (1) Using the three consecutive time points before the node before piecewise correction and the node after piecewise correction, a total of four fitting points, determine the two calculation parameters of formula (20). and The first three points use the adapted flow rate QP1 and the first corrected water level H3. The fourth point uses the adapted flow rate QP1 and H1 of the node after segmented correction, plus the water level difference before and after the segmented correction, to determine the parameters. and Then, the first corrected water level H3 for each time point in the segment is calculated using formula (20);

[0300] (20)

[0301] (2) Calculate the first corrected water level based on the first and second points after the node before segmented correction. :

[0302] (twenty one)

[0303] Since step S9 calculates the first corrected water level based on the first water level, therefore and It is the first corrected water level of the first and second points after the correction node before segmentation;

[0304] If the first correction water level of the correction node before segmentation is corrected If the first corrected water level in the segment is not smooth, the solution is as follows:

[0305] 1) Since step S9 calculates the first corrected water level based on the first water level, the first corrected water level is calculated using the first corrected water level at three points: the node before segmented correction and the two points before the node before segmented correction. :

[0306] (twenty two)

[0307] if ,

[0308] in, and It is the first corrected water level at the point one and two points before the node before the segmented correction; It is the first corrected water level of the node before segmented correction;

[0309] 2) Calculate the first corrected water level for the four water level fitting points:

[0310] (twenty three)

[0311] in, It is the first water level of the node before segmented correction; It is a preset correction threshold; It is the first corrected water level of the node before segmented correction; and It is the first corrected water level at the point one and two points before the node before the segmented correction; It is the first corrected water level of the node after segmented correction.

[0312] (3) Using the water levels at the four water level fitting points, recalculate the two parameters of formula (20). and Then, the first corrected water level of each time point in the segment is calculated using formula (20).

[0313] Rope loop type: For multi-peak floods, in addition to the secondary flood flow and water level rope loop morphology segments of the maximum flow peak segment including nodes 98 and 99, four consecutive water level correction nodes are grouped together and a set method is constructed based on the characteristics of the flow and water level rope loop. The water level correction nodes (short codes) of a group of nodes appearing in chronological order are: rope loop front node 53, rope loop inlet node 54, rope loop apex node 55, and rope loop outlet node 56. After node 56, the flow decreases within the shortest time of hydrological analysis and there are flow fluctuation nodes. The rope loop type processing does not include the water level correction calculation after node 56.

[0314] Therefore, the rope loop types, from front to back, include the rope loop front type, rope loop inlet type, rope loop apex type, and rope loop outlet node, consisting of three consecutive segments.

[0315] The method for locating the rope loop nodes is as follows: First, determine the rope loop vertex node, where the time-varying acceleration aQP1 of the first adaptive flow rate QP1 changes from negative to positive and is closest to 0. Next, determine the rope loop inlet node, where the time-varying velocity VQP1 of the first adaptive flow rate QP1 moving forward from the rope loop vertex node changes from positive to negative and is closest to 0. Then, determine the rope loop front node, where the time-varying acceleration aQP1 of the first adaptive flow rate QP1 moving forward from the rope loop inlet node reaches a local maximum value and is greater than 0, and the time-varying acceleration aQP1 subsequently changes from positive to negative and gradually decreases, or reaches a local maximum value point of the first adaptive flow rate QP1. Finally, determine the rope loop outlet node, where the time-varying velocity VQP1 of the first adaptive flow rate QP1 moving backward from the rope loop vertex node changes from negative to positive and is closest to 0. The water level correction node after the rope loop outlet node is the point where the time-varying acceleration aQP1 of the first adaptive flow rate QP1 reaches a local maximum value point.

[0316] The corresponding water level correction algorithm for the node before the rope loop is ID53. Its segmentation is from the node before the rope loop to the rope loop inlet node. After correction, the water level rises at each point. After the node before the rope loop, VQP1 and aQP1 gradually decrease to the local maximum flow rate. The change in water surface gradient lags behind, and the water level rise will continue after the rope loop inlet node. The water level correction calculation formula is:

[0317] (twenty four)

[0318] Since step S9 calculates the first corrected water level H3 based on the first water level H1, therefore It is the time-varying acceleration of the first adaptable flow at the rope loop inlet node; It is the time-varying acceleration of the first adaptive flow at the front node of the rope loop; It refers to the duration of each segment; It is the first adaptive flow of the node before the rope loop before the segmentation correction node; It is the first compatible flow rate at the rope loop inlet node; It is the first corrected water level at the node before the rope loop; It is the first water level at the node before the rope loop; These are preset parameters.

[0319] The corresponding water level correction algorithm ID54 for the rope inlet type, after the first adapted flow rate QP1 at the rope inlet node reaches a local maximum before segmented correction, begins to decrease, and aQP1 decreases. However, the change in the additional water surface gradient within the segment causes the water level calculated based on the flow-level function to need correction. The water level correction calculation formula is as follows:

[0320] (25)

[0321] Since step S9 calculates the first corrected water level H3 based on the first water level H1, therefore It is the time-varying acceleration of the first adapted flow at the vertex node of the loop; It is the first adaptive flow of the loop vertex node; It is the first corrected water level at the rope loop inlet node; It is the first water level at the rope loop inlet node; It refers to the duration of each segment; These are preset parameters.

[0322] The corresponding water level correction algorithm ID55 for the rope loop apex type indicates that the first adapted flow rate QP1 within the rope loop continues to decrease, aQP1 changes from negative to positive, and the water level gradient begins to reverse. The water level correction calculation formula is as follows:

[0323] (26)

[0324] in, It is the time-varying acceleration of the first adaptable flow rate at the rope loop outlet node; It is the first compatible flow rate at the rope loop outlet node; It is the first corrected water level at the top node of the rope loop; It is the first water level at the top node of the rope loop; It refers to the duration of each segment; These are preset parameters.

[0325] Step S9, through the above-mentioned determination of water level correction nodes, segments and corresponding algorithms, starts from the starting point and completes the first water level H1 correction calculation segment by segment from front to back to obtain the first corrected water level H3.

[0326] S10, Calculate the first corrected water level between the flow-water level relationship control node and the calculation end node: Calculate the first corrected water level H3 using the first adapted flow rate QP1 between the flow-water level relationship control node and the calculation end node.

[0327] In one embodiment, since the calculation termination node is a convex point on the flow-water level curve, the water level stagnates significantly between the flow-water level control node and the calculation termination node. The water level calculation between the flow-water level control node and the calculation termination node uses a linear function. The first corrected water level H3 is directly calculated using the first adapted flow rate QP1 between the flow-water level control node and the calculation termination node. The calculation formula is as follows:

[0328] (28)

[0329] Since step S10 calculates the first corrected water level H3 based on the first adapted flow rate QP1, therefore It is the first correction water level of time series number i. The calculation of i starts from 2; It is the sequence number of the time point starting from the flow-water level relationship control node, and the flow-water level relationship control node... =1; These are preset parameters; It is the first adapted traffic for time sequence number i; It is the first adapted flow rate of the flow-water level relationship control node; It is the first corrected water level of the flow-water-level relationship control node.

[0330] S11, Calculate the second water level between the starting point and the flow-water level relationship control node: Calculate the second water level H2 between the starting point and the flow-water level relationship control node based on the second adapted flow rate QP2.

[0331] Step S11 uses the same calculation method as step S8 to calculate the second water level between the starting point and the flow-water level relationship control node.

[0332] Unlike step S8, the flow data used in step S11, the first adapted flow rate QP1, is changed to the second adapted flow rate QP2. Therefore, the second water level H3 of the flow-water level relationship control node is calculated. 99 Parameters of formula (14) , , , and This is different from the first adapted flow rate QP1 and needs to be calibrated and verified based on historical flood data. Calculate water level H3. 99 Then, the second water level H2 corresponding to the second adapted flow rate QP2 between the starting point and the flow-water level relationship control node can be calculated using formula (15).

[0333] S12, the second water level between the starting point and the flow-water level relationship control node is corrected to generate the second corrected water level: from the starting point to the flow-water level relationship control node, the calculation scheme of H1 to calculate H3 in step S9 is adopted, including the same water level correction node, the same segmentation, the same water level correction type and the corresponding correction algorithm. From the starting point to the flow-water level relationship control node, the second water level H2 is corrected point by point from front to back to calculate the second corrected water level H4.

[0334] Unlike step S9, step S12 uses flow-related data QP2, VQP2, and aQP2, with the water level being the second water level H2, and the calculated result being the second corrected water level H4.

[0335] S13, Calculate the second corrected water level between the flow-water level relationship control node and the calculation end node: Calculate the second corrected water level H4 using the second adapted flow rate QP2 between the flow-water level relationship control node and the calculation end node.

[0336] Step S13 uses the same calculation scheme as step S10.

[0337] Unlike step S10, step S13 uses the second adapted flow rate QP2 and the second water level H2. The parameters of the calculation formula (28) are different, and the water level after H2 correction is the second corrected water level H4.

[0338] S14, calculate the water surface gradient corresponding to the first and second adaptive flow rates of the wavefront: calculate the time series values ​​of the two water surface gradient processes based on the first adaptive flow rate QP1 and the corresponding first corrected water level H3, the second adaptive flow rate QP2 and the corresponding second corrected water level H4.

[0339] Verification data from forecast hydrological stations includes tables showing the relationship between cross-sectional water level and flow area, and between cross-sectional water level and hydraulic radius. Furthermore, based on the graded roughness of the cross-section, a relationship between cross-sectional water level or flow rate and roughness can be established. Therefore, based on the compiled hydrological data from forecast cross-sections, and using the first adapted flow rate QP1 and the first corrected water level H2, the second adapted flow rate QP2 and the second corrected water level H4, the time series values ​​of the two wavefront water surface gradient processes can be calculated using the Manning formula. The Manning formula is:

[0340] (29)

[0341] in, This refers to the corresponding traffic, which is either the first or second adapted traffic. It is the cross-sectional roughness; It is the cross-sectional water flow area, which is obtained from the relationship table between cross-sectional water level and water flow area based on the first corrected water level H2 or the second corrected water level H4. It is the hydraulic radius of the cross section, which is obtained from the table of the relationship between cross section water level and hydraulic radius based on the first corrected water level H2 or the second corrected water level H4. It is the water surface gradient of the corresponding flow rate and the corresponding corrected water level.

[0342] Example 1:

[0343] A national-level hydrological station responsible for forecasting sudden floods in the middle and lower reaches of small and medium-sized rivers was selected, and station verification data were collected from this station. The premise for a stable flow-water level relationship at the forecast section is that the upstream and downstream river channels are relatively stable compared to the smaller forecast section morphology. Analysis of the underwater topography of the test section and historical cross-sectional measurement data revealed that the forecast section underwent dredging and improvement in 2017, resulting in significant changes in cross-sectional morphology before and after construction. However, the cross-section at the forecast station remained stable after construction. Therefore, from 2018, the year after the completion of river improvement construction, to 2023, when the hydrological data was fully entered into the database, all historical flood data were collected. After processing, 16 flood data were available for water level forecasting scheme preparation. Using the pre-flood cross-sectional measurement data from 2023, relevant cross-sectional maps were compiled as follows: Figures 2 to 5 Collect cross-sectional flow forecasting schemes and forecasting operation application software. Based on the collected rainfall, water level, and flow data, complete the flow recap forecast for each flood event. Combine the observed water level and flow at the forecast cross-section, calculate the wavefront water level for each flood event according to this method.

[0344] Adopting such Figure 1 The method of steps S1 to S14 of the present invention shown calculates the wavefront water level based on the forecast flow rate for the forecast cross section in the middle and lower reaches of small and medium-sized rivers.

[0345] In step S7, formula (12) and They are -0.2590 and 0.6431 respectively; in step S8, formula (14) , , , and The values ​​are 0.000034, -0.015379, 3.871, 0.089088, and 0.594143, respectively; in step S11, formula (14) , , , and The values ​​are 0.000056, -0.023403, 4.573, 0.087849, and 0.611642, respectively; in steps S8 and S11, the preset flow boundary control value is 400m. 3 / s; In step S10, formula (28) It is 0.00365; in step S13, formula (28) It is 0.00515.

[0346] Due to the numerous influencing factors and high uncertainty in forecasting the flow of sudden floods, the "Specification for Hydrological Information Forecasting" (GB22482-2008) stipulates: "6.3.3 Forecasting of sudden floods can adopt a forecasting scheme based on a combination of meteorological and hydrological methods, and use it to estimate the flow (water level)." "6.3.4 Forecasting of sudden floods does not require accuracy assessment." The example used all historical flood data available for the forecasting hydrological station, totaling 16 historical floods. Table 1 is a statistical table of peak flow forecasting data for these 16 historical floods.

[0347]

[0348] In this embodiment, the measured maximum peak flow of the two historically largest floods was 1026.4 m³ / h. 3 / s and 1031.5m 3 / s. The statistical analysis of the accuracy of flow forecast calculations for 16 historical floods is as follows: the average absolute value of the absolute error in the peak flow forecast is 81.9 m³ / s. 3 / s, with an average relative error of 18.4%, an accuracy rate of 81.6%, a coefficient of determination (DC) of 0.39, and a pass rate of 56.3%. This does not consider floods that are unlikely to cause disasters, i.e., floods with a maximum peak flow rate of less than 300.0 m³ / s. 3 For events with a flow rate of / s, the pass rate for flood peak flow forecasts has significantly increased to 88.9%. Considering the technical requirements of flood disaster risk forecasting and early warning for hydrological operations, the flow forecasting scheme developed by the forecasting stations in this embodiment is not only practical but also reliable and technologically advanced.

[0349] Table 2 presents the statistical data of 16 historical flood level forecasts. The accuracy of the 16 historical flood level forecast calculations is statistically analyzed as follows: Looking at the coefficient of determination (DC) for the water level forecast process, the DC values ​​for H3 and H4 increased from 0.7 in the original forecast flow calculation scheme to 0.9, significantly improving the consistency between the forecast and measured water level processes. Regarding the predicted highest water level, the average absolute error decreased from 0.20m to 0.16m and 0.14m, the average relative error decreased from 6.1% to 4.8% and 4.4%, and the pass rate increased from 68.8% to 75.0% and 87.5%. Therefore, the accuracy and reliability of the water level forecast using this method are significantly higher than the flood forecast scheme in the example, effectively reducing the false alarm rate of flood risk warnings in the middle and lower reaches of small and medium-sized rivers. Its widespread application will yield significant socio-economic benefits.

[0350]

[0351] In this embodiment, in specific implementation, after step S14, the following may also be included:

[0352] S15, Determine the water level forecast calculation results: Cross-sectional forecast water levels are one of the most important decision support data for issuing flood warning signals. A comprehensive display of time-series data such as basin rainfall, cross-sectional forecast flow, adapted flow, water level, and water surface gradient is beneficial for digital consultation on hydrological forecasts and flood defense decision analysis. The most important application scenario for water level forecast calculation results is analyzing the time required to reach various levels of flood level warning signals, and forming water level forecast warning recommendations.

[0353] According to the management system for the release of forecast results by hydrological agencies, forecasters need to determine the expected time for reaching each level of flood level warning after consultation within the relevant hydrological agency, based on the natural geographical characteristics, hydrological characteristics, hydrological and meteorological evolution patterns, historical flood processes and disaster situations of the forecast basin, and the water level forecast calculation results. After the results are signed by the supervisor, the forecasters submit the water level forecast calculation results.

[0354] S16, Rolling Forecasting: The longer the time between the flood forecast calculation results and the starting point, the greater the uncertainty of the calculated results for the forecast items, such as flow and water level. While computer calculations for water level forecasts take only seconds, forecast calculation schemes often require forecasters to continuously optimize and adjust them based on hydrological observation data and their personal experience. In actual operations, establishing an intelligent rolling forecasting mechanism allows the application system to automatically adjust the starting point and perform calculations after receiving hydrological observation data, once the forecaster has prepared the calculation scheme, thus meeting the requirements for rolling flood forecasting operations.

[0355] S17, End Water Level Forecast Operation: End the wavefront water level forecast operation for this flood according to the preset rules, and output the wavefront water level forecast calculation results.

[0356] The preset rule is: if the time between the last forecast operation for a flood event and the occurrence of the forecasted maximum flood peak flow does not exceed the flow duration at the forecast section, the flood front water level forecast operation for that flood event can be terminated in principle, and this result can be used as the final flood front water level forecast calculation result.

[0357] The preferred embodiments of the present invention have been described in detail above. However, it should be understood that after reading the above teachings, those skilled in the art can make various alterations or modifications to the present invention. These equivalent forms also fall within the scope defined by the appended claims.

Claims

1. A method for calculating wavefront water level in forecasting sudden floods, characterized in that, The method for calculating wavefront water levels in the forecast of sudden floods includes: S1, real-time acquisition of hydrological observation and forecast data of flood events, determination of fixed nodes for wavefront water level forecast calculation, the hydrological observation and forecast data including observed water level, observed flow rate and forecast flow rate, the fixed nodes from front to back include starting point, flow rate-water level relationship control node and calculation end point; S2, determine whether the initial forecast flow needs to be corrected. If no correction is needed, proceed to step S4; otherwise, proceed to step S3. S3, determine the initial traffic correction control node, including the control node before initial traffic correction and the control node after initial traffic correction; S4, starting from the starting point to the flow-water level relationship control node, several flow conversion control nodes on the forecast flow are located step by step from front to back. Two adjacent flow conversion control nodes form a segment, and the forecast flow is divided into one or several segments. The node type of the control node before each segment corresponds to a preset node type algorithm for converting the forecast flow to the appropriate flow. The forecast flow before the flow-water level relationship control node is converted into the appropriate flow through each node type algorithm. S5, convert the predicted flow rate between the flow-level relationship control node and the calculation end node into an adapted flow rate; S6, the adapted traffic is filtered and then converted into the first adapted traffic; S7, convert the first adapted traffic into the second adapted traffic; S8, calculate the first water level between the starting point and the flow-water level relationship control node based on the first adaptive flow rate; S9, starting from the starting point to the flow-water-level relationship control node, several water level correction nodes on the first water level are located step by step from front to back. Two adjacent water level correction nodes form a segment, thus dividing the first water level into one or several segments. The water level correction type of the correction node before each segment corresponds to a preset correction algorithm for correcting the first water level, forming a first water level correction scheme; starting from the starting point to the flow-water-level relationship control node, the corresponding first corrected water level is obtained through the first water level correction calculation of each segment; S10, using the first adaptive flow rate between the control node and the calculation end node based on the flow-water-level relationship, calculate the first corrected water level; S11, calculate the second water level between the starting point and the flow-water level relationship control node based on the second adapted flow rate; S12, from the starting point to the flow-water-level relationship control node, the first water level correction scheme compiled in step S9 is adopted, including all water level correction nodes, segments, water level correction types and correction algorithms of segments, and from the starting point to the flow-water-level relationship control node, the corresponding second corrected water level is obtained by calculating the second water level correction of each segment. S13, using the second adaptive flow rate between the control node and the calculation end node based on the flow-water-level relationship, calculate the second corrected water level; S14. Calculate the water surface gradient process based on the first adapted flow rate and the corresponding first corrected water level, the second adapted flow rate and the corresponding second corrected water level.

2. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1, characterized in that, Step S1 includes: S101, if the predicted flow rate Q reaches the preset minimum control flow rate earlier than the observed flow rate, then the time point when the predicted flow rate Q reaches the minimum control flow rate is taken as the starting point; otherwise, the time point when the observed flow rate is not less than the minimum control flow rate and the predicted flow rate Q reaches 50% of the minimum control flow rate is taken as the starting point. S102, after the starting point, the maximum value of the predicted flow rate Q is taken as the calculation end point; S103, starting from the end node of the calculation, the time point when the time-varying acceleration aQ of the predicted flow rate Q first turns from negative to positive and the value is closest to 0 is taken as the flow rate-water level relationship control node. When the time between the end node of the calculation and the flow rate-water level relationship control node is less than the preset minimum time for cross-sectional hydrological analysis, the flow rate-water level relationship control node is located according to the preset minimum time between the end node of the calculation and the flow rate-water level relationship control node. The formula for calculating the time-varying velocity VQ of flow rate Q is: (1) The formula for calculating the time-varying acceleration aQ of flow rate Q is: (2) in, It is time-series flow Data sequence number starting from 1; It is the time difference between two adjacent data points; The value of i is calculated starting from 2. The value of i is calculated starting from 3.

3. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1, characterized in that, In step S2, if the observed water level and the predicted flow rate do not change in the same trend, or if the observed flow rate and the predicted flow rate do not change in the same trend, then the predicted flow rate needs to be corrected; otherwise, the predicted flow rate does not need to be corrected. In step S3, the control node before the initial flow correction uses the following positioning method: During the period from the control node before the initial flow correction to the starting point, the changes and trends of the observed water level and observed flow are consistent; during the period from the control node before the initial flow correction to the starting point, the time-varying velocity VQ and time-varying acceleration aQ of the predicted flow Q do not change. The time between the control node and the starting point before the initial flow correction shall not be less than the preset minimum time. Preferably, starting from the starting point, the time point at which the sign of either the time-varying velocity VQ or the time-varying acceleration aQ of the predicted flow Q changes is selected as the control node before the initial flow correction. In step S3, the control node bit after the initial flow correction is located at the last of the following three timing points: When the predicted flow rate at the starting point is less than the preset minimum control flow rate, the first point where the predicted flow rate is closest to rising to the minimum control flow rate is taken as the control node after the initial flow rate correction. After the starting point, the time-varying velocity VQ and time-varying acceleration aQ of the predicted flow rate Q remain unchanged. The time point at which either the time-varying velocity VQ or the time-varying acceleration aQ of the predicted flow rate Q first shows a positive-to-negative value transition point is used as the control node after the initial flow rate correction. After the initial flow correction, the time between the control node and the starting point is not less than the preset minimum time.

4. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1, characterized in that, In step S4, the type of the flow conversion control node includes at least one of the following: steady type, fluctuating type, and rising type; The stable advancement type includes at least one of a first stable advancement type and a second stable advancement type; The method for determining the steady-progress type is as follows: starting from the starting point or the control node before the segment, the time point where the time-varying velocity VQ of the predicted flow Q changes from positive to negative and is closest to 0 is taken as the subsequent control node of the steady-progress type; when the segment duration of the steady-progress type exceeds the cross-sectional flow history and the predicted flow and observed water level and observed flow change trends within the segment are consistent, the steady-progress type is the first steady-progress type; when the segment duration of the steady-progress type exceeds 1.5 times the cross-sectional flow history and the observed water level and observed flow change trends within the segment are consistent, but the time-varying velocity VQ of the predicted flow Q is greater than a preset deviation value, the steady-progress type is the second steady-progress type. The node type algorithm corresponding to the first stable type has the following adapted traffic calculation formula: (4) in, It is the timing point number starting from the control node before segmentation, and the control node before segmentation... =1, The time series index i is calculated starting from 2; It is the adapted flow of time series number i, and the adapted flow of the starting point is the observed flow; It is the time-varying velocity of the predicted flow rate at time point i; The node type algorithm corresponding to the second stable type has the following adapted traffic calculation formula: (5) in, It is the timing point number starting from the control node before segmentation, and the control node before segmentation... =1; It is the predicted flow rate of the last time series point within the segment. It is the predicted flow rate of the first time series point within the segment; It is the adapted traffic for time sequence number i. The time series index i is calculated starting from 2; It is the time-varying velocity of the predicted flow rate at time point i; These are preset parameters; The fluctuation type includes at least one of a first fluctuation type and a second fluctuation type; The method for determining the first type of fluctuation is as follows: starting from the control node before segmentation, the time-varying velocity VQ of the forecast flow Q changes from positive to negative and then back to positive, the time-varying acceleration aQ of the forecast flow changes from negative to positive and then back to negative, the forecast flow first rises and then falls and then rises again, and the absolute value of the flow difference in the falling segment is less than the flow difference between the two rising segments. The post-control node of the first fluctuation type is located at the time point when the time-varying velocity VQ of the predicted flow Q falls to the point where the value is closest to 0. The node type algorithm corresponding to the first fluctuation type is: Using the time series index within a segment as the independent variable, the adaptive flow rate at each time series point within the segment is calculated using an exponential function. The calculation formula is as follows: (6) in, The sequence number within the segment starts from 1, which is the sequence number of the control node before the segmentation. It is the traffic adapted for i6 timing point sequence numbers. The calculated time series index i6 starts from 2; and These are the two parameters of the exponential function; Formula (4) is used to calculate the adaptive flow at the first two time points, the middle time point, and the last two time points within a segment. The adaptive flow at these five points is used to determine the two parameters of the adaptive flow exponential function for segmented calculation. and Then, calculate the adaptive flow corresponding to each time point within the segment according to formula (6); The method for determining the second type of fluctuation is as follows: starting from the control node before the segment, the time-varying velocity VQ value of the predicted flow rate Q first decreases and turns negative, then gradually increases and turns positive, and the time-varying acceleration aQ of the predicted flow rate Q gradually increases and turns positive; preferably, if the control node after the next segment is the flow-water level relationship control node, then this segment is determined to be the second type of fluctuation. The second type of fluctuation's post-control node is located at the last of the following three reference time series points: the time series point where the forecast flow rises after a decrease to the point where it first exceeds the control node before the segment; the time series point where the forecast flow rises after a decrease to the minimum value within the segment, located dTpf hours before the time-varying velocity VQ of the forecast flow Q turns positive, where dTpf is the hydrological characteristic value of the forecast section; and the time series point where the time-varying acceleration aQ of the forecast flow Q increases to a local maximum value that is positive. The node type algorithm corresponding to the second fluctuation type has the following adapted traffic calculation formula: (7) in, The sequence number within the segment starts from the sequence number of the control node before the segmentation, which is 1. It's about adapting to traffic volume. The calculation of i starts from 2; It is a forecast of the flow rate; It is the time-varying velocity of the forecast flow; It is the minimum forecast flow within the segment; It is the maximum predicted flow rate within the segment; and These are preset parameters; The upward movement type includes at least one of the following: first upward movement type, second upward movement type, third upward movement type, fourth upward movement type, and ninth upward movement type; The method for determining the first, third, and ninth types of water level rise is as follows: starting from the control node before segmentation, the forecast flow rate continuously increases and the water surface gradient continues to increase. The method for locating the post-control node for the first, third, and ninth rise types is as follows: if the time between the pre-segment control node and the calculation end node does not exceed twice the preset minimum time, then the node is located at the calculation end node; otherwise, the node is located at the time point where the predicted flow rate Q drops to its closest point to 0 after the maximum value of the time-varying velocity VQ. The segmentation time should exceed the cross-sectional flow duration; if it does not exceed this duration, other node types are selected. The further determination method for the first, third, and ninth rise types is as follows: First, use formula (4) to calculate the initial value of the adapted flow of the control node after segmentation and the initial value of the adapted flow of the control node before segmentation. If the relative error of the initial value of the adapted flow of the control node before segmentation relative to the forecast flow exceeds 5%, then the segmentation is the ninth rise type; if the above conditions are not met, but the relative error of the initial value of the adapted flow of the control node after segmentation relative to the forecast flow does not exceed 5%, then the segmentation is the third rise type; the segmentation that does not meet the above conditions is the first rise type. The node type algorithms corresponding to the first, third, and ninth rise types, and the adapted traffic calculation process are as follows: First, determine 6 fitting points for the adaptive flow calculation. The method for determining these points is as follows: the control node before segmentation, a time series point after the control node before segmentation, and 4 consecutive time series points starting from the m-th point forward from the control node after segmentation. The method for calculating m is as follows: (8) in, The sequence number of the last point of the timing data within the segment, starting from the timing point number of the control node before the segmentation, is 1. Calculate the time-varying velocity adjustment coefficient for different node types. : (9) in, It is the predicted flow of the control nodes before segmentation; It is the predicted flow of the control nodes after segmentation; It is the adaptation traffic of the control node before the segmentation control point; It is the adaptation traffic of the control node before segmentation; according to Calculate the fitted flow at the fitting point for the 6 fitted flows within the segment: (10) The adaptive flow of the fitting point is calculated based on the time sequence number of the fitting point from the control node before segmentation and the formula (10). The adaptive flow of the fitting point is calculated based on the fitting point of ... The method for determining the second type of rise is as follows: the control node before the segment is the outlet of the flow-water level curve loop, or the control node after the fluctuation type, which predicts that the flow will continue to rise. The method for locating the post-control node of the second type of rise is as follows: if the time between the pre-control node and the calculation end node does not exceed twice the preset minimum time, then it is located at the calculation end node; otherwise, it is located at the time series point where the value of the time-varying velocity VQ of the forecast flow Q first appears after the pre-control node and then drops to the sign of the value closest to 0. The node type algorithm corresponding to the second type of rise is: the flow difference of the adapted flow within the segment is equal to the flow difference of the predicted flow, the adapted flow of the control node before the segment is equal to the adapted flow of the control node after the segment of the previous segment, and the adapted flow is calculated using formula (4). The fourth type of rise is a forecast initial flow correction algorithm, the front control node of the fourth type of rise is the starting point, and the back control node of the fourth type of rise is the back control node of the initial flow correction. The node type algorithm corresponding to the fourth type of rise is as follows: According to step S3, the control node before the initial flow correction and the control node after the initial flow correction are determined. The observed flow from the control node before the initial flow correction to the starting point is used. An autoregressive moving average model is used to construct a cubic polynomial and extrapolate to calculate the adapted flow from the starting point to the segmented control node. And / or, step S5 includes: using formula (4) to calculate the adaptive flow rate between the flow-level relationship control node and the calculation end node.

5. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1, characterized in that, Step S6 includes: using a 3-point numerical filtering algorithm to filter the adaptive flow between the starting point and the end point of the calculation and convert it into the first adaptive flow; Preferably, the calculation formula for the 3-point numerical filtering algorithm is as follows: (11) in, It is the first adaptive flow at the starting point; It is the adaptive traffic at the starting point; It is the sequence number of the time series starting from the starting point, and the starting point =1; It is the first adapted traffic for time sequence number i; It is the adapted traffic for time sequence number i; It is the first adapted traffic for the calculation of the end node; It calculates the adaptation traffic for the end node.

6. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1 or 5, characterized in that, In step S7, the calculation formula for converting the first adapted traffic into the second adapted traffic is as follows: (12) in, and These are the preset parameters; It is the sequence number of the timing point within the segment, starting from the timing point number of the control node before the segmentation, which is 1; It is the first adapted traffic for time sequence number i; It is the first adapted traffic for the calculation of the end node; It is the first adaptive flow at the starting point; It is the second adaptive flow at the starting point; It is the second adapted traffic for time sequence number i. The calculation of i starts from 2; It is the time-varying velocity of the first adapted flow of time sequence number i.

7. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1, characterized in that, Step S8 includes: Based on the preset flow boundary control value Determine the flood magnitude of the flood event, classifying it into two levels: small to medium flow and large flow. Calculate the first water level of the flow-water level control node based on the flood magnitude. (14) Among them, QP1 99 The first adaptive flow rate for the flow-level relationship control node, H1 99 The first water level at the control node for the flow-water-level relationship; , , , and These are the preset parameters; Calculate the first water level between the starting point and the flow-water-level relationship control node: (15) in, It is the sequence number of the time series starting from the starting point, and the starting point =1; It is the first watermark of time point i. The calculation of i starts from 2; It is the first adapted traffic for time sequence number i; It is the observed flow rate at the starting point; It is the observed water level at the starting point; Step S11 uses the same calculation method as step S8 to calculate the second water level between the starting point and the flow-water level relationship control node. When calculating the second water level in step S11, the first adaptive flow rate is replaced with the second adaptive flow rate, and the first water level is replaced with the second water level. The parameters used in formula (14) are different from those in step S8.

8. The method for calculating wavefront water level for forecasting sudden floods as described in claim 7, characterized in that, In step S9, the water level correction type includes at least one of the basic type, the extension type, and the rope type; The basic type includes at least one of the zero basic type, the first basic type, and the second basic type; The extension types include the seventh extension type and the eighth extension type; The method for determining the zeroth basic type is as follows: starting from the starting point or the pre-segment correction node, if the first adapted flow rate before the flow-water level relationship control node increases continuously, and the first adapted flow rate and the first water level change trend are the same within the preset shortest time before the pre-segment correction node, then the post-correction node of the zeroth basic type is located at the flow-water level relationship control node. The correction algorithm corresponding to the zeroth basic type, the water level correction calculation formula is as follows: (16) in, It is the first corrected water level after the time series number i has been corrected. It is the first correction water level of the correction node before segmentation; It is the sequence number starting from the pre-segmentation correction node, and the pre-segmentation correction node's... The value is 1, and the water level correction calculation i starts from 2; It is the first watermark of time point i. It is the first water level of the pre-segmentation correction node; when the pre-segmentation correction node is the starting point. ; The method for determining the first basic type and the seventh extended type is as follows: the first basic type and the seventh extended type are located by the sign change of the time-varying acceleration aQP1 of the first adaptive flow QP1 being closest to 0 or a local extreme value; after the first basic type and the seventh extended type continue to rise after the previous correction node, the first basic type and the seventh extended type are located by the local extreme value of the time-varying acceleration aQP1 of the first adaptive flow QP1; when the segmented previous correction node is within the preset shortest time, relative to the segmented previous correction node, the slope of the first adaptive flow and the first water level curve at each time point is greater than that in formula (15). When the value is 1.2 times, the segment is the seventh extended type; otherwise, the segment is the first basic type. The correction algorithm corresponding to the first basic type has the following water level correction calculation formula: (18) in, It is the first corrected water level after the time series number i has been corrected. It is the first correction water level of the correction node before segmentation; It is the sequence number of the time series starting from the starting point, and the previous correction node. The value is 1, and the water level correction calculation i starts from 2; It is the time-varying acceleration of the first adapted flow of the segmented correction node; It is the time-varying acceleration of the first adaptive flow of the correction node before segmentation; It refers to the duration of each segment; It is the first adapted traffic of the correction node before segmentation; It is the first adapted traffic of the segmented and corrected node; It is the first water level of the correction node before segmentation; These are preset parameters; The correction algorithm corresponding to the seventh extension type has the following water level correction calculation formula: (19) in, It is the first corrected water level after the time series number i has been corrected. It is the first correction water level of the correction node before segmentation; It is the sequence number starting from the pre-segmentation correction node, and the pre-segmentation correction node's... The value is 1, and the water level correction calculation i starts from 2; It is the time-varying acceleration of the first adapted flow of the segmented correction node; It is the time-varying acceleration of the first adaptive flow of the correction node before segmentation; It refers to the duration of each segment; It is the first adapted traffic of the correction node before segmentation; It is the first adapted traffic of the segmented and corrected node; It is the first water level of the correction node before segmentation; These are preset parameters; The pre-correction node positioning method for the second basic type and the eighth extended type is as follows: the time-varying velocity VQP1 of the first adaptive flow QP1 continuously increases to a local maximum value and the time-varying acceleration aQP1 of the first adaptive flow QP1 is greater than 0, or the time-varying acceleration aQP1 of the first adaptive flow QP1 is a local maximum value and the time-varying velocity VQP1 of the first adaptive flow QP1 is greater than 0. The modified node positioning method for the second basic type and the eighth extended type is as follows: if the time from the node before segmented correction to the flow-level relationship control node does not exceed the preset minimum time, then it is positioned as the flow-level relationship control node, or the time-varying velocity VQP1 of the first adapted flow QP1 is closest to 0 and the first adapted flow reaches a local maximum value for positioning. The method for determining the second basic type and the eighth extended type is as follows: if the difference between the first water level of the node before segmentation correction and the first correction water level is greater than the preset correction threshold, then the segmentation is the eighth extended type; otherwise, the segmentation is the second basic type. The calculation process of the water level correction algorithm corresponding to the second basic type and the eighth extended type is as follows: The two calculation parameters of formula (20) are determined by using three consecutive time points before the node before piecewise correction and the node after piecewise correction, for a total of four fitting points. and The first three points use the first adapted flow rate and the first corrected water level. The fourth point uses the first adapted flow rate and the first water level of the node after segmented correction, plus the water level difference before and after the segmented correction, to determine the parameters. and Then, the first corrected water level of each time point in the segment is calculated using formula (20); (20) Calculated based on the first corrected water level at the first and second points after the node before segmented correction. : (21) in, and It is the first corrected water level of the first and second points after the correction node before segmentation; If the first correction water level of the correction node before segmentation is corrected If the first corrected water level in the segment is not smooth at the preceding node, the solution is as follows: Using the first corrected water level at three points—the node before the piecewise correction and the two points before the node before the piecewise correction—the following calculations are performed. : (22) if , ; in, and It is the first corrected water level at the point one and two points before the node before the segmented correction; It is the first corrected water level of the node before segmented correction; Calculate the first corrected water level for the four water level fitting points: (23) in, It is the first water level of the node before segmented correction; It is a preset correction threshold; It is the first corrected water level of the node before segmented correction; and It is the first corrected water level at the point one and two points before the node before the segmented correction; It is the first corrected water level of the node after segmented correction; Using the water levels at four fitting points, the two parameters of formula (20) are recalculated. and Then, the first corrected water level of each time point in the segment is calculated using formula (20); The types of rope loops, from front to back, include rope loop front type, rope loop inlet type, rope loop apex type, and rope loop outlet node; The method for locating the rope loop nodes is as follows: First, determine the rope loop vertex node, where the time-varying acceleration aQP1 of the first adaptive flow rate QP1 changes from negative to positive and is closest to 0; then, determine the rope loop inlet node, where the time-varying velocity VQP1 of the first adaptive flow rate QP1 moving forward from the rope loop vertex node changes from positive to negative and is closest to 0; determine the rope loop front node, where the time-varying acceleration aQP1 of the first adaptive flow rate QP1 moving forward from the rope loop inlet node reaches a local maximum value and is greater than 0, and the time-varying acceleration aQP1 subsequently changes from positive to negative and gradually decreases, or reaches a local maximum value point of the first adaptive flow rate QP1; finally, determine the rope loop outlet node, where the time-varying velocity VQP1 of the first adaptive flow rate QP1 moving backward from the rope loop vertex node changes from negative to positive and is closest to 0. The segmentation of the rope loop type is from the rope loop inlet node to the rope loop entrance node. The correction algorithm corresponding to the rope loop type, and the water level correction calculation formula are as follows: (24) in, It is the time-varying acceleration of the first adaptable flow at the rope loop inlet node; It is the time-varying acceleration of the first adaptive flow at the front node of the rope loop; It refers to the duration of each segment; It is the first adaptive flow rate of the node before the rope loop before the segmentation correction node; It is the first compatible flow rate at the rope loop inlet node; It is the first corrected water level at the node before the rope loop; It is the first water level at the node before the rope loop; These are preset parameters; The segmentation of the rope loop inlet type is from the rope loop inlet node to the rope loop vertex node. The correction algorithm corresponding to the rope loop inlet type, and the water level correction calculation formula are as follows: (25) in, It is the time-varying acceleration of the first adapted flow at the vertex node of the loop; It is the first adaptive flow of the loop vertex node; It is the first corrected water level at the rope loop inlet node; It is the first water level at the rope loop inlet node; It refers to the duration of each segment; These are preset parameters; The segmentation of the rope loop vertex type is from the rope loop vertex node to the rope loop exit node. The correction algorithm corresponding to the rope loop vertex type, and the water level correction calculation formula are as follows: (26) in, It is the time-varying acceleration of the first adaptable flow rate at the rope loop outlet node; It is the first compatible flow rate at the rope loop outlet node; It is the first corrected water level at the top node of the rope loop; It is the first water level at the top node of the rope loop; It refers to the duration of each segment; These are preset parameters; Step S12 uses the same correction node positioning method, water level correction type, and correction algorithm as step S9. When calculating the second corrected water level in step S12, the first adapted flow rate is replaced with the second adapted flow rate, the time-varying velocity of the first adapted flow rate is replaced with the time-varying velocity of the second adapted flow rate, the time-varying acceleration of the first adapted flow rate is replaced with the time-varying acceleration of the second adapted flow rate, the first water level is replaced with the second water level, and the first corrected water level of the calculated result is replaced with the second corrected water level.

9. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1, characterized in that, In step S10, the formula for calculating the first corrected water level is: (28) in, It is the first corrected water level after the time series number i has been corrected. The calculation of i starts from 2; It is the sequence number of the time point starting from the flow-water level relationship control node, and the flow-water level relationship control node... =1; These are preset parameters; It is the first adapted traffic for time sequence number i; It is the first adaptable flow rate of the flow-water level relationship control node; It is the first corrected water level of the flow-water-level relationship control node; Step S13 uses the same calculation formula as step S10. When calculating the second corrected water level in step S13, the first adapted flow rate is replaced with the second adapted flow rate, and the first water level is replaced with the second water level. The parameters used in formula (28) are different from those in step S10. The first corrected water level of the calculation result is replaced with the second corrected water level.

10. The method for calculating wavefront water level for forecasting sudden floods as described in claim 1, characterized in that, Step S14 includes: Based on the first adapted flow rate and the corresponding first corrected water level, and the second adapted flow rate and the corresponding second corrected water level at each time point, the time series values ​​of the two wavefront water level gradient processes are obtained by back-calculating using the Manning formula. The Manning formula is as follows: (29) in, This refers to the corresponding traffic, which is either the first or second adapted traffic. It is the cross-sectional roughness; It is the cross-sectional water flow area, which is obtained from the relationship table between cross-sectional water level and water flow area based on the first or second corrected water level. It is the hydraulic radius of the cross section, which is obtained from the relationship table between the cross section water level and the hydraulic radius based on the first or second corrected water level. It is the water surface gradient of the corresponding flow rate and the corresponding corrected water level.

Images