A method for correcting a mountain torrent disaster early warning index under non-stationary conditions

Abstract

The application discloses a mountain torrent disaster early warning index correction method under non-stationary conditions and relates to the technical field of disaster early warning, which comprises the following steps: step one, collecting rainfall, river level and branch stream convergence inlet water flow movement data in real time and preprocessing; step two, calculating a river jacking model by using the preprocessed data, predicting the jacking effect, and if the jacking effect exists, proceeding to step three, and if the jacking effect does not exist, keeping a critical rainfall threshold; step three, further calculating a hydrological state model, predicting whether the branch stream breaches the bank, updating the threshold and early warning if the branch stream breaches the bank, and returning to step two if the branch stream does not breach the bank. The application collects multiple data by using ultrasonic sensors and electromagnetic current meters, builds a model to predict the jacking effect, dynamically updates the critical rainfall, avoids the adaptation failure of the traditional fixed threshold, prevents early warning lag from aggravating disasters, disassembles water flow data, differentiates and weights in different zones, improves the precision of hydrological evaluation, combines data preprocessing, and provides reliable support for mountain torrent early warning and flood control decision-making under non-stationary conditions.

Description

Technical Field

[0001] This invention relates to the field of disaster early warning technology, specifically to a method for correcting flash flood disaster early warning indicators under non-stationary conditions. Background Technology

[0002] Under non-stationary conditions, such as frequent extreme rainfall, the rise in the water level of the main stream creates a significant backwater effect on the tributaries. This hydraulic interaction greatly exacerbates the risk of backwater and levee breaches in the tributary basin. The flash floods caused by the backwater effect are characterized by their strong concealment, suddenness, and destructiveness. That is, the water flow in the upstream tributary basin is slow, but the downstream flood discharge channel is completely blocked by the rapidly rising water level of the main stream. As a result, the water in the tributary cannot flow smoothly downstream, and thus accumulates and flows back into the basin. Finally, without warning, the levee is breached, flooding downstream towns and farmland and causing catastrophic consequences.

[0003] For example, Chinese Patent Publication No. CN113934777B provides a method and system for quantifying the impact of backwater on water level changes. This includes determining a backwater backwater hydrological simulation model based on historical hydrological data of a target river section; determining water level change response characteristics and their contribution rate to water level changes in the target river section based on target hydrological data of the target river section during a target time period, according to the backwater backwater hydrological simulation model; wherein the water level change response characteristics are factors influencing water level changes in the target river section. This method can quantitatively analyze the contribution of backwater backwater to water level changes, improving the accuracy of identifying the causes of high flood levels.

[0004] During the rainy season, heavy rainfall causes water levels in both the main stream and tributaries of rivers to rise simultaneously. If the main stream exerts a backwater effect on the tributaries at this time, it will significantly alter the original hydrological dynamics, leading to a suppression of the tributaries' discharge capacity and a rapid increase in water levels. This causes the critical rainfall warning threshold set based on the absence of backwater conditions to lose its accuracy, resulting in the increased speed and intensity of flash floods as the high-flowing water breaches the banks. Summary of the Invention

[0005] Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a method for correcting flash flood disaster early warning indicators under non-stationary conditions. This method solves the problem that the critical rainfall early warning threshold set based on the absence of river backwater effect becomes ineffective, thereby exacerbating the suddenness and destructiveness of flash flood disasters.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for correcting flash flood disaster early warning indicators under non-stationary conditions, comprising the following specific steps: Step 1: Real-time collection and preprocessing of rainfall data, river water level data, and tributary confluence water flow movement data, wherein the rainfall data includes rainfall in different time periods, the river water level data includes the main stream water level and tributary water level, and the tributary confluence water flow movement data includes water flow velocity and water flow direction; Step 2: Comprehensive calculation of the preprocessed river water level data and tributary confluence water flow movement data to obtain a river backwater model, and based on the river backwater model... The process involves predicting the backwater effect between the main stream and tributaries. If a backwater effect is predicted, proceed to step three. If no backwater effect is predicted, maintain the critical rainfall threshold, which serves as the critical value for the current rainfall to cause the tributary to breach the embankment. Step three: Further calculate the preprocessed river water level data and tributary confluence flow movement data to obtain a hydrological state model. Based on the hydrological state model, predict whether the tributary will breach the embankment. If a breach is predicted, update the critical rainfall threshold based on the rainfall data, send an early warning, and end this process. If no breach is predicted, return to step two.

[0008] Furthermore, the specific method for obtaining the river backwater model is as follows: the difference between the tributary water level and the main stream water level is calculated to obtain the water level difference value; the water flow movement data at the tributary confluence is comprehensively calculated to obtain the water flow movement anomaly coefficient; the rainfall, water level difference, and water flow movement anomaly coefficient at different time periods are comprehensively calculated and standardized at the same time to obtain the river backwater model. ;in, This represents a river top support model. This indicates rainfall amounts over different time periods. Indicates the water level difference. Indicates the anomaly coefficient of water flow movement. It represents a positive real number.

[0009] Furthermore, the specific method for obtaining the anomaly coefficient of water flow movement is as follows: A two-dimensional coordinate system is established, where the horizontal axis represents the water flow direction and the vertical axis represents the water flow velocity. This coordinate system includes only the first and second quadrants. The horizontal axis of the first quadrant represents the positive direction, and the horizontal axis of the second quadrant represents the negative direction. The vertical axis represents only the magnitude of the water flow velocity, which is greater than or equal to zero. The water flow movement data at the tributary confluence is compiled into a tributary confluence water flow movement dataset. Coordinate points are established for the water flow velocity and direction of each tributary confluence water flow movement dataset. The horizontal coordinate of the coordinate point represents the water flow direction, and the water flow direction is assigned a value... or , It is a positive real number, that is Indicates positive. The positive direction is indicated by the vertical coordinate of the coordinate point, which represents the water flow velocity. A detection time period is set for the coordinate points. During the detection time period, the number of coordinate points in the first quadrant and the second quadrant are counted respectively to obtain the number of positive coordinate points and the number of negative coordinate points. The quotient of the number of negative coordinate points and the number of positive coordinate points is calculated. The number of positive coordinate points is not zero, and the abnormal coefficient of water flow movement is obtained.

[0010] Furthermore, the specific method for predicting the backwater effect of the main stream and tributaries based on the river backwater model is as follows: set a river backwater threshold, compare the output value of the river backwater model with the river backwater threshold, if the output value of the river backwater model is greater than the river backwater threshold, it indicates that a backwater effect has been formed, if the output value of the river backwater model is less than or equal to the river backwater threshold, it indicates that no backwater effect has been formed.

[0011] Furthermore, the specific method for obtaining the hydrological state model is as follows: During the coordinate point detection period, the changes in the number of positive and negative coordinate points are comprehensively calculated to obtain the abnormal trend aggravation coefficient. This abnormal trend aggravation coefficient is then compared with... The product calculation is performed to obtain the hydrological state model.

[0012] Furthermore, the specific method for obtaining the aggravation coefficient of the abnormal trend is as follows: The distribution area of ​​positive coordinate points in the first quadrant is divided into an average region and a second region. The distance from the vertical axis to the horizontal axis of the positive coordinate points in the first region is greater than that in the second region. Similarly, the distribution area of ​​negative coordinate points in the second quadrant is divided into an average region and a fourth region. The distance from the vertical axis to the horizontal axis of the positive coordinate points in the fourth region is greater than that in the third region. Since the third region is more affected by the supporting effect than the second region, the first and second regions are... The first, second, third, and fourth regions are weighted, and the number of coordinate points in each region is counted. The number of coordinate points in each region is then calculated. Each region is multiplied by its respective weight to obtain four distribution quantities. These four distribution quantities are then sorted in ascending order using a quicksort algorithm to obtain the maximum distribution quantity, which is recorded as the anomalous trend aggravation coefficient.

[0013] Furthermore, the specific methods for obtaining the first and second regions are as follows: The distance between the ordinate of each positive coordinate point in the first quadrant and the horizontal axis is calculated to obtain a set of positive distances. This set is then sorted in ascending order using a quicksort algorithm to obtain the maximum positive distance. The maximum positive distance is averaged and divided into two segments. The segment furthest from the horizontal axis is designated as the first segment, and the segment closest to the horizontal axis is designated as the second segment. The length of the first segment is then averaged with the horizontal axis. Performing a product calculation yields the first region, and the length of the second line segment is... Perform product calculations to obtain the second region.

[0014] Furthermore, the specific steps for assigning weights to the first region, the second region, the third region, and the fourth region are as follows: The fourth region is assigned a value... Assign a value to the third region Assign a value to the second region Assign a value to the first region ,in It is a real number greater than one.

[0015] Furthermore, the specific method for predicting whether a tributary will breach a bank based on the hydrological state model is as follows: a bank breach threshold is set, and the output value of the hydrological state model is compared with the bank breach threshold. If the output value of the hydrological state model is greater than the bank breach threshold, it indicates that the bank has breached; if the output value of the hydrological state model is less than or equal to the bank breach threshold, it indicates that the bank has not breached.

[0016] Beneficial effects

[0017] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0018] 1. Through multi-dimensional key data collection and professional model construction, accurate prediction of the backwater effect of rivers is achieved. The critical rainfall threshold can be dynamically updated based on the backwater effect, effectively avoiding the shortcomings of existing technologies that fail to provide early warnings due to neglecting the backwater effect. Ultrasonic water level sensors near the confluence of main and tributary streams accurately collect water level data, combined with the flow velocity and direction data at the tributary confluence obtained by a bidirectional electromagnetic current meter. After preprocessing, the water level difference and flow movement anomaly coefficient are calculated, and then rainfall is incorporated to construct a river backwater model. This allows for scientific judgment of whether a backwater effect has formed under non-stationary conditions. When a backwater effect is predicted, the risk of tributary breach is further assessed through a hydrological state model. If a breach is possible, the critical rainfall threshold is immediately updated based on the current rainfall, avoiding the situation where fixed thresholds under traditional conditions without backwater cannot adapt to the backwater effect of the main stream, leading to tributary congestion and preventing the exacerbation of the suddenness and destructiveness of flash floods due to delayed early warnings. If no backwater effect has formed, the original threshold is maintained, balancing the flexibility and accuracy of early warnings.

[0019] 2. By refining the breakdown, regional division, and differentiated weighting of water flow data, the accuracy of hydrological status assessment has been significantly improved, providing more reliable technical support for flash flood early warning. In the water flow status analysis stage, a two-dimensional coordinate system containing only the first and second quadrants is established, transforming the water flow data at tributary confluences into coordinate points. The anomaly coefficient of water flow movement is calculated by statistically analyzing the number of positive and negative coordinate points, intuitively reflecting whether there is an abnormal trend of backflow from the main stream into the tributaries. At the same time, the restricted areas are further subdivided, and weights are differentiated according to the magnitude of the backwater effect in each area. Combined with rapid sorting to obtain the aggravation coefficient of abnormal trends, this design ensures that the weight size is accurately matched with the actual degree of impact. Coupled with the instrument characteristics and preprocessing in the data acquisition stage, the quality of the model input data and the scientific nature of the calculation logic are ensured, significantly improving the accuracy of tributary breach prediction and providing more effective data support for flood control decision-making.

[0020] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0021] Figure 1 This invention provides a flowchart for a method to correct early warning indicators for flash floods under non-stationary conditions.

[0022] Figure 2 This invention provides a comparison diagram of water levels at normal tributary cross sections and water levels at tributary cross sections affected by the backwater effect. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0025] Example 1:

[0026] like Figure 1 As shown in the figure, this invention provides a method for correcting flash flood disaster early warning indicators under non-stationary conditions, including the following specific steps:

[0027] Step 1: Real-time rainfall data is retrieved via the API interface of local weather stations. This unstructured rainfall data undergoes text parsing to extract structured data regarding time and rainfall amount for the region. Therefore, the rainfall data includes rainfall amounts for different time periods. Real-time river water level data is collected using ultrasonic water level sensors. This data is then cleaned to remove redundant values ​​and improve data quality. The river water level data includes both main stream and tributary water levels. Ultrasonic water level sensors are deployed near the tributary and main stream sections, close to their confluence. The sensors emit ultrasonic waves perpendicularly towards the water surface and receive the echoes reflected from the water. The distance from the sensor to the water surface is calculated based on the sound wave propagation time difference, and then converted into... The system accurately acquires two types of water level data simultaneously to provide a basis for analyzing the correlation between the water volume of the main stream and tributaries. Real-time data on the flow movement at the tributary confluence is collected using a bidirectional electromagnetic current meter. This data is also cleaned to remove redundant values ​​and improve its quality. The data includes both flow velocity and direction. Based on Faraday's law of electromagnetic induction, the bidirectional electromagnetic current meter measures the amplitude and polarity of the induced electromotive force generated by the fluid cutting magnetic field lines to calculate the magnitude and direction of the water flow velocity. Since the bidirectional electromagnetic current meter detects only one dimension of flow direction (positive and negative), multiple bidirectional electromagnetic current meters are placed near the tributary cross-section, with the normal flow direction of the tributary into the main stream being positive.

[0028] Step 2: Perform comprehensive calculations on the cleaned river water level data and tributary confluence water flow movement data to obtain the river backwater model. Based on the river backwater model, predict the backwater effect between the main stream and tributaries. If the backwater effect is predicted to occur, proceed to Step 3. If the backwater effect is predicted not to occur, maintain the critical rainfall threshold, which serves as the critical value for the current rainfall to cause the tributary to breach the bank.

[0029] Step 3: Further calculations are performed on the cleaned river water level data and tributary confluence water flow movement data to obtain a hydrological state model. Based on the hydrological state model, it is predicted whether the tributary will breach the embankment. If the embankment breach is predicted, the critical rainfall threshold is updated based on the rainfall data, i.e., updated to the current rainfall. This means that if the subsequent rainfall is greater than or equal to the current rainfall, it will cause the water flow to breach the embankment. This avoids the critical rainfall warning threshold set based on the absence of backwater conditions losing its accuracy. Warning prompts are also sent to the staff's equipment so that they can prepare for flood prevention and avoid delaying rescue time, which could exacerbate the suddenness and destructiveness of flash floods due to the backwater breaching the embankment. This process ends. If the embankment breach is not predicted, return to Step 2.

[0030] Example 2 differs from Example 1 in that:

[0031] The specific method for obtaining the river top support model is as follows:

[0032] The difference between the tributary water level and the main stream water level is calculated to obtain the water level difference value. Under normal circumstances, the water level of the tributary is higher than that of the main stream, which allows the water of the tributary to flow smoothly into the main stream. That is, the water level difference value is greater than zero under normal circumstances. The water flow movement data at the tributary confluence is comprehensively calculated to obtain the water flow movement anomaly coefficient, which is used to reflect whether the water flow of the tributary is abnormal. The rainfall, water level difference, and water flow movement anomaly coefficient of different time periods are comprehensively calculated and standardized to eliminate the dimensional differences of rainfall, water level difference, and water flow movement anomaly coefficient of different time periods. Furthermore, the different orders of magnitude of rainfall, water level difference, and water flow movement anomaly coefficient of different time periods are transformed into a unified numerical range to obtain the river backwater model.

[0033] ;

[0034] in, This represents a river top support model. This represents the rainfall amount over different time periods. When the rainfall amount over a different time period is zero, If we consider the total rainfall as one, we can ignore the impact of precipitation. However, when the rainfall in different time periods is greater than zero, A value greater than zero indicates that rainfall affects the river backwater model. This represents the water level difference. Under normal circumstances, the water level of a tributary is higher than that of the main stream, meaning the water level difference is greater than zero. It has a relatively small impact on the river backwater model, but when the tributary water level is equal to or less than the main stream water level, i.e., the water level difference is less than or equal to zero, then... This has a significant impact on the river top support model. This represents the anomaly coefficient of water flow movement. The more anomalous the water flow movement, the greater its impact on the river backwater model. Representing positive real numbers avoids the situation where the river backwater model is meaningless when the anomaly coefficient of water flow movement is zero.

[0035] The specific method for obtaining the anomaly coefficient of water flow movement is as follows:

[0036] A two-dimensional coordinate system is established, where the horizontal axis represents the water flow direction and the vertical axis represents the water flow velocity. This coordinate system only includes the first and second quadrants, meaning the water flow direction has positive and negative values. The horizontal axis in the first quadrant represents the positive direction, i.e., the normal flow direction of the tributary. The horizontal axis in the second quadrant represents the negative direction, i.e., the direction opposite to the normal flow direction of the tributary caused by the backflow of some water after the tributary is blocked by the main stream or the backflow of some water from the main stream into the tributary. The vertical axis represents only the magnitude of the water flow velocity, which is greater than or equal to zero. Since multiple bidirectional electromagnetic current meters are arranged near the tributary cross-section, the water flow movement data at the tributary confluence constitutes a tributary confluence water flow movement dataset. Coordinate points are established for the water flow velocity and direction of each tributary confluence water flow movement dataset, where the horizontal coordinate of the coordinate point represents the water flow direction, and the water flow direction is assigned a value. or , It is a positive real number, that is Indicates positive. The vertical coordinate of the coordinate point represents the water flow velocity. A detection time period is set, and the number of coordinate points in the first and second quadrants is counted within this period to obtain the number of positive and negative coordinate points. The quotient of the number of negative and positive coordinate points is calculated, and the number of positive coordinate points is not zero. This yields the water flow anomaly coefficient. ;in, Indicates the anomaly coefficient of water flow movement. Indicates the number of positive coordinate points. This indicates the number of negative coordinate points; that is, under normal circumstances, the number of positive coordinate points is greater than the number of negative coordinate points, making the anomaly coefficient of water flow movement less than one. However, after the backwater effect, the number of positive coordinate points is less than or equal to the number of negative coordinate points, making the anomaly coefficient of water flow movement greater than or equal to one.

[0037] The specific method for predicting the backwater effect between the main stream and tributaries based on the river backwater model is as follows:

[0038] A river backwater threshold was set based on historical experiments. The output value of the river backwater model was compared with the river backwater threshold. If the output value of the river backwater model was greater than the river backwater threshold, it indicated that a backwater effect had occurred. Figure 2 As shown, the main stream, due to its high water level and slow flow velocity, obstructs the inflow of tributaries and causes backflow, preventing tributaries from flowing normally into the main stream, thus causing accumulation and backflow. If the output value of the river backflow model is less than or equal to the river backflow threshold, it indicates that no backflow effect has occurred.

[0039] The specific methods for obtaining the hydrological state model are as follows:

[0040] During the coordinate point detection period, the changes in the number of positive and negative coordinate points are comprehensively calculated to obtain the abnormal trend aggravation coefficient. This abnormal trend aggravation coefficient is then compared with... By performing product calculations, the hydrological state model is obtained. ;in, Representing the hydrological state model, Indicates the coefficient of aggravation of abnormal trends. This indicates the difference in water levels.

[0041] The specific method for obtaining the aberration exacerbation coefficient is as follows:

[0042] Dividing the positive coordinate point distribution area in the first quadrant into two equal parts, we obtain the first and second regions. In the first region, the distance from the vertical axis to the horizontal axis of the positive coordinate points is greater than that in the second region. This means the first region is less affected by the backflow effect than the second region. Similarly, dividing the negative coordinate point distribution area in the second quadrant into two equal parts, we obtain the third and fourth regions. In the fourth region, the distance from the vertical axis to the horizontal axis of the positive coordinate points is greater than that in the third region. This means the fourth region is more affected by the backflow effect than the third region. In other words, the backflow of water from the main stream into the tributaries increases the probability of backflow in the tributaries, while the third region is less affected by the backflow effect. The impact is greater than that of the second region. Therefore, weights are assigned to the first, second, third, and fourth regions, and the number of coordinate points in each region is counted. The number of coordinate points in the first, second, third, and fourth regions is then multiplied by their respective weights to obtain four distribution quantities. These four distribution quantities are then sorted in ascending order using a quicksort algorithm to obtain the maximum distribution quantity, which is recorded as the anomalous trend aggravation coefficient.

[0043] The specific methods for obtaining the first and second regions are as follows:

[0044] Calculate the distance between the ordinate of each positive coordinate point in the first quadrant and the x-axis to obtain the positive distance set. ;in, Represents the positive distance set, This represents the distance between the ordinate of each positive coordinate point and the x-axis. Let the y-coordinate of the positive coordinate point be used. The positive distance set is sorted in ascending order using the quicksort algorithm to obtain the maximum positive distance. The maximum positive distance is then averaged and divided into two segments: the segment furthest from the horizontal axis is designated as the first segment, and the segment closest to the horizontal axis is designated as the second segment. The length of the first segment is then compared with... Performing a product calculation yields the first region, and the length of the second line segment is... Perform product calculations to obtain the second region.

[0045] The specific steps for assigning weights to the first, second, third, and fourth regions are as follows:

[0046] Due to the influence of the support effect, the regions are arranged in descending order as the fourth region, the third region, the second region, and the first region. Therefore, the fourth region is assigned a value of Assign a value to the third region Assign a value to the second region Assign a value to the first region ,in The fourth region is the most affected because it is not only negative in direction, but also has a high reverse flow velocity, which means that the probability of the main stream flowing back into the tributary increases.

[0047] The specific method for predicting whether a tributary will breach a bank based on a hydrological state model is as follows:

[0048] A threshold for levee breach is set through historical experiments. The output value of the hydrological state model is compared with the threshold. If the output value of the hydrological state model is greater than the threshold, it indicates that the levee has breached. If the output value of the hydrological state model is less than or equal to the threshold, it indicates that the levee has not breached.

[0049] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A flash flood early warning index correction method under non-stationary conditions, characterized by: The method comprises the following specific steps: Step one: collecting rainfall data, river level data and tributary confluence entrance water flow movement data in real time and preprocessing, wherein the rainfall data comprises rainfall in different time periods, the river level data comprises main stream level and tributary level, and the tributary confluence entrance water flow movement data comprises water flow speed and water flow direction; Step two: comprehensively calculating the preprocessed river level data and tributary confluence entrance water flow movement data to obtain a river jacking model, predicting the jacking effect of the main stream and the tributary according to the river jacking model, if the jacking effect is predicted, executing step three, if the jacking effect is not predicted, keeping a critical rainfall threshold value, wherein the critical rainfall threshold value is a critical value of the current rainfall causing the tributary to breach the bank; The specific acquisition method of the river jacking model is as follows: The water level difference value is obtained by calculating the difference between the tributary level and the main stream level, the water flow movement anomaly coefficient is obtained by comprehensively calculating the tributary confluence entrance water flow movement data, and the river jacking model is obtained by comprehensively calculating and standardizing the rainfall in different time periods, the water level difference value and the water flow movement anomaly coefficient; ； wherein, represents a river backwater model, represents rainfall amount at different time periods, represents water level difference value, represents water flow movement abnormality coefficient, represents a positive real number; The specific acquisition method of the water flow movement anomaly coefficient is as follows: A two-dimensional coordinate system is established, wherein the horizontal axis of the coordinate system is the water flow direction, the vertical axis of the coordinate system is the water flow velocity, and the coordinate system only includes the first quadrant and the second quadrant, the horizontal axis of the first quadrant represents a positive direction, the horizontal axis of the second quadrant represents a negative direction, the vertical axis is only the size of the water flow velocity, and is greater than or equal to zero. The tributary inlet water flow movement data is grouped into a tributary inlet water flow movement data set, and a coordinate point is established for the water flow velocity and the water flow direction of each tributary inlet water flow movement data, wherein the horizontal coordinate of the coordinate point is the water flow direction, and the water flow direction is assigned a value of or , is a positive real number, that is represents a positive direction, represents a negative direction, the vertical coordinate of the coordinate point is the water flow velocity, a coordinate point detection time period is set, the number of coordinate points in the first quadrant and the second quadrant is counted respectively in the coordinate point detection time period, and the number of positive direction coordinate points and the number of negative direction coordinate points are obtained. The number of negative direction coordinate points and the number of positive direction coordinate points are calculated by division, and the number of positive direction coordinate points is not zero, to obtain a water flow movement anomaly coefficient. Step three: further calculating the preprocessed river level data and tributary confluence entrance water flow movement data to obtain a hydrological state model, predicting whether the tributary breaches the bank according to the hydrological state model, if the bank is breached, updating the critical rainfall threshold value according to the rainfall data and sending an early warning, ending the process, if the bank is not breached, returning to step two.

2. The method according to claim 1, wherein the method is characterized in that: The specific method for predicting the jacking effect of the main stream and the tributary according to the river jacking model is as follows: Setting a river jacking threshold value, comparing the output value of the river jacking model with the river jacking threshold value, if the output value of the river jacking model is greater than the river jacking threshold value, it indicates that the jacking effect is formed, if the output value of the river jacking model is less than or equal to the river jacking threshold value, it indicates that the jacking effect is not formed.

3. The method according to claim 2, wherein the method is characterized in that: The specific acquisition method of the hydrological state model is as follows: During the coordinate point detection period, the changes in the number of positive and negative coordinate points are comprehensively calculated to obtain the abnormal trend aggravation coefficient. This abnormal trend aggravation coefficient is then compared with... The product calculation is performed to obtain the hydrological state model.

4. The method according to claim 3, characterized in that: The specific acquisition method of the anomaly trend aggravation coefficient is as follows: The positive coordinate point distribution area of the first quadrant is averagely divided to obtain a first area and a second area, wherein the distance from the vertical axis to the horizontal axis of the positive coordinate point in the first area is greater than that of the positive coordinate point in the second area; and the negative coordinate point distribution area of the second quadrant is averagely divided to obtain a third area and a fourth area, wherein the distance from the vertical axis to the horizontal axis of the positive coordinate point in the fourth area is greater than that of the positive coordinate point in the third area; the third area is affected by the toppling effect more than the second area, therefore, the first area, the second area, the third area and the fourth area are weighted, and the coordinate point numbers of the first area, the second area, the third area and the fourth area are counted to obtain the coordinate point number of the first area, the coordinate point number of the second area, the coordinate point number of the third area and the coordinate point number of the fourth area; the coordinate point number of the first area, the coordinate point number of the second area, the coordinate point number of the third area and the coordinate point number of the fourth area are respectively multiplied by the respective weights to obtain four distribution numbers; the four distribution numbers are sorted in ascending order by a quick sort algorithm to obtain a maximum distribution number, and the maximum distribution number is recorded as an abnormal trend intensification coefficient.

5. The flash flood warning index correction method under non-stationary conditions according to claim 4, characterized in that: The specific acquisition method of the first area and the second area is as follows: The vertical coordinate of each positive coordinate point in the first quadrant is calculated with the horizontal axis to obtain a positive distance set. The positive distance set is sorted in ascending order by a quick sorting algorithm to obtain a maximum positive distance. The maximum positive distance is averaged to divide into two line segments. The line segment far from the horizontal axis is recorded as a first line segment, and the line segment close to the horizontal axis is recorded as a second line segment. The length of the first line segment is multiplied by to obtain a first area. The length of the second line segment is multiplied by to obtain a second area.

6. The flash flood warning index correction method under non-stationary conditions according to claim 5, characterized in that: The specific steps of weighting the first area, the second area, the third area and the fourth area are as follows: the fourth region is assigned the value the third region is assigned the value the second region is assigned the value the first region is assigned the value wherein is a real number greater than one.

7. The method according to claim 6, wherein the method is characterized by: The specific method of predicting whether the tributary breaches the bank according to the hydrological state model is as follows: A breach threshold of the bank is set, and the output value of the hydrological state model is compared with the breach threshold of the bank; if the output value of the hydrological state model is greater than the breach threshold of the bank, it indicates that the bank is breached; if the output value of the hydrological state model is less than or equal to the breach threshold of the bank, it indicates that the bank is not breached.

Images