Method and system for judging and predicting stable state of river regime of alluvial river based on river swing

Abstract

The invention discloses a method and system for judging and predicting the stable state of the river regime of an alluvial river based on the amplitude of river channel, and the method comprises the steps: carrying out the batch preprocessing of satellite remote sensing images of a target region, and obtaining a water binary image after noise reduction; constructing a historical river regime database based on the water body binary image; calculating a river regime relative stability index of the hydrological section based on the historical river regime database; the optimal section number of the overall river regime relative stable state of the river reach is determined by correcting and encrypting original hydrological sections; and judging the river regime of the river and predicting the future river regime of the river based on the encrypted and corrected river regime relative stability index of the hydrological section. According to the method, the problem of insufficient spatial resolution of a traditional hydrological observation section in river regime stable state judgment is solved while the river regime evolution time accumulation effect is considered, and certain technical support is provided for alluvial river flood control safety and river regulation.

Description

Technical Field

[0001] The invention relates to the technical field of water conservancy engineering, and in particular to a method and system for determining and predicting the stable state of an alluvial river regime based on river channel swing amplitude. Background Art

[0002] River regime refers to the planar form and development trend of river flow, including the location and direction of the dynamic axis of river flow, and the distribution and change trend of river bends, shorelines, sandbars, and mid-shoals. Early scholars divided alluvial rivers in nature into different river types based on different river regimes, movement characteristics, and boundary characteristics of rivers: straight and curved rivers are generally considered to be in a relatively stable state of river regime, while bifurcated and wandering rivers are generally considered to be in an unstable state of river regime. The drastic adjustment of river regime in a short period of time seriously threatens the safety of life and property of people on both sides of the river. Therefore, scientifically and efficiently evaluating and predicting the stability of alluvial river regimes, and adjusting the river plane morphology within a controllable range by controlling the inflow of water, sand, and boundary conditions have long been important issues in the study of riverbed evolution and river sediment.

[0003] At present, the traditional determination and prediction of river regime stability of alluvial rivers are mainly realized through the calculation of hydrodynamic indicators (such as riverbed stability indicators, river type discriminant indicators, etc.) and plane morphological indicators (such as mainstream swing amplitude and curvature of the river channel, etc.). The former starts from the basic laws of riverbed evolution and river type transformation, with clear physical meaning and comprehensive considerations, but involves many variables, including a series of hydrological and riverbed evolution variables such as flow, gradient, river width, water depth, median particle size of bed sand, etc., so its calculation accuracy is usually limited by the spatiotemporal resolution of the measured data obtained. The latter focuses on quantifying the changes of one or several plane geometric or topological indicators in a certain period of time to determine the stability of the river regime in a certain period of time. Its expression form is intuitive, and the data acquisition methods (such as prototype measurement, remote sensing images, etc.) are diverse, which can realize the rapid evaluation of the stability of the river regime; however, this type of method ignores the physical mechanism of river regime evolution, and the accuracy of river regime stability evaluation and prediction still needs to be improved.

[0004] At the same time, the determination and prediction methods of the stability of alluvial river regime based on plane morphological indicators (such as channel swing) still face the following challenges. On the one hand, due to the existence of hysteresis response, the evolution of river regime in a short period of time may reflect the long-term cumulative effect of its dynamics and boundary impedance. When determining and predicting the stability of river regime, it is necessary to further consider the inherent physical mechanism of river regime evolution; on the other hand, replacing the stability of river regime at the river section scale with the calculation results of indicators at the cross-section scale faces the problem of insufficient representativeness. Summary of the invention

[0005] In order to solve the above problems, the present invention proposes a method and system for determining and predicting the stable state of alluvial river flow based on river channel swing, which takes into account the cumulative effect of river flow evolution over time, while taking into account the calculation accuracy of hydrodynamic indicators and the calculation efficiency of river planar morphological indicators.

[0006] In order to achieve the above object, the present invention is implemented by the following technical solutions:

[0007] The method for determining and predicting the stability state of an alluvial river based on the channel swing amplitude of the present invention comprises the following operations:

[0008] The satellite remote sensing images of the target area are batch preprocessed to obtain the de-noised binary images of the water body;

[0009] Construct a historical river regime database based on binary images of water bodies;

[0010] Calculate the relative stability index of river flow at hydrological sections based on the historical river flow database;

[0011] By revising and encrypting the original hydrological sections, the optimal number of sections for quantifying the relatively stable state of the river regime of the entire river section is determined;

[0012] Based on the relative stability index of river flow of the encrypted and revised hydrological sections, the river flow is judged and the future flow of the river is predicted.

[0013] A further improvement of the present invention is that the satellite remote sensing images of the target area are batch preprocessed to obtain a de-noised binary image of the water body, and the specific operations include:

[0014] Determine the resolution Re and size of the satellite remote sensing image, and set the relevant parameter values, including the maximum size σ of the noise reduction filter max , the number of clustering groups of hydrological sections g, the number of encrypted sections Δn, the expected probability of unstable river regime P ex ;

[0015] The improved normalized difference water index of satellite remote sensing images is calculated in the ENVI model, and the water body is roughly extracted from the satellite remote sensing images with this index as the model input to obtain a binary image of the water body. The interior of the roughly extracted water body is filled with a denoising filter to obtain a binary image of the water body after denoising.

[0016] A further improvement of the present invention is that the size of the noise reduction filter is variable, the size is denoted by σ, and the value is σ=3, 5, 7, ..., σ max The filling process of the interior of the roughly extracted water body by using the denoising filter specifically includes: using the denoising filter to sequentially convolve with each background pixel in the roughly extracted water body:

[0017]

[0018] Among them, f(i,j) and g(i,j) are the pixel values ​​corresponding to the i-th row and j-th column of the water body binary image and denoising filter, respectively, σ is the size of the denoising filter, and (f*g)(i,j) is the convolution operation;

[0019] If the convolution result is (4σ-4), all background pixels in the convolution kernel are corrected to water bodies, and the size of the convolution kernel is gradually increased to σ max , and finally the denoised water body binary image is obtained.

[0020] A further improvement of the present invention is to construct a historical river regime database based on the binary image of the water body, specifically including:

[0021] The edge detection method and thinning algorithm are used to extract the left water surface line, the right water surface line and the skeleton line of the river channel from the de-noised water body binary image, and the coordinates of the pixel points belonging to different line types are recorded respectively;

[0022] The river regime-related data of different years are extracted respectively, including the river skeleton line, left water surface line, right water surface line and hydrological section line, to construct a historical database of river regime.

[0023] A further improvement of the present invention is that the relative stability index of river regime is calculated based on the historical river regime database, specifically including:

[0024] Based on Taylor series, a quantitative index for judging the relative stability of river flow is established, namely, the relative stability index of river flow RSIRR, where the relative stability index of river flow RSIRR is a linear combination of the swing amplitude s, swing speed v, swing acceleration a and water surface width b of the skeleton line of the river channel;

[0025] Calculate the swing amplitude s, swing velocity v and swing acceleration a of the channel skeleton line on each hydrological section in turn;

[0026] Cluster analysis is performed on all hydrological sections according to the three indicators of the swing amplitude s, swing velocity v and swing acceleration a of the river skeleton line, and the corresponding weight of each indicator is determined;

[0027] The relative stability index of river regime (RSIRR) is calculated on each hydrological section in turn, and the critical threshold between adjacent section categories is determined.

[0028] The further improvement of the present invention is that the optimal number of sections for quantifying the relatively stable state of the river regime of the entire river section is determined by modifying and encrypting the original hydrological sections, specifically including:

[0029] Generate a set of new hydrological sections at equal intervals to perform spatial interpolation on the river section, and calculate the mean of the relative stability index RSIRR of all new hydrological sections at each moment in turn;

[0030] Based on the mean value of the river regime relative stability index RSIRR, a sensitivity analysis was conducted on the number of new hydrological sections. Starting from the number of hydrological sections being 1, the number of section densification was gradually increased with a set step size Δn. The mean values ​​of the river regime relative stability index RSIRR of the corresponding hydrological sections were calculated in turn until the result tended to be stable. The corresponding number of hydrological sections was determined as the optimal number of densified sections for the corresponding river section. The hydrological sections of the river section were densified and corrected with the optimal number of densified sections.

[0031] The further improvement of the present invention is to judge the river regime and predict the future river regime based on the relative stability index of the river regime of the encrypted and corrected hydrological section, specifically including:

[0032] Based on the encrypted and corrected relative stability index of the hydrological section, the spatiotemporal distribution matrix RSIRR1 of the relative stability index of the river flow is constructed. According to the spatiotemporal distribution matrix RSIRR1 of the relative stability index of the river flow and the critical thresholds of different section categories, the judgment result of the relative stability state of the river flow at the corresponding moment is obtained.

[0033] Based on the spatiotemporal distribution matrix RSIRR1 of the relative stability index of river regime, the river regime stability probability function P(RSIRR') is calculated. The value of the river regime stability probability function P(RSIRR') is equal to the expected probability of river regime instability P ex The relative stability index of river regime RSIRR' corresponding to the time is used to obtain the prediction result of the relative stability state of river regime.

[0034] The present invention provides a system for determining and predicting the stability of an alluvial river regime based on river channel swing amplitude, the system comprising: a network interface, a memory and a processor; wherein the network interface is used for receiving and sending signals in the process of sending and receiving information with other external network elements;

[0035] A memory for storing computer programs that can be run on the processor;

[0036] The processor is used to execute the steps of the above-mentioned method for determining and predicting the stable state of the river flow of an alluvial river based on the river channel swing when running a computer program.

[0037] A computer storage medium of the present invention stores a program for determining and predicting the stable state of an alluvial river based on channel swing amplitude. When the program for determining and predicting the stable state of an alluvial river based on channel swing amplitude is executed by at least one processor, the steps of the method for determining and predicting the stable state of an alluvial river based on channel swing amplitude are implemented.

[0038] The beneficial effects of the present invention are as follows: compared with the traditional visual interpretation method based on prototype data, and the river type discriminant and riverbed stability index based on a large amount of hydrological and riverbed evolution measured data, the present invention only considers the changes in the geometric morphology of alluvial rivers, which is simpler in form and avoids the influence of the subjectivity of traditional methods. In addition, the present invention not only considers the time accumulation effect in the process of riverbed evolution, but also solves the problem of insufficient spatial resolution in evaluating the relative stability of river regime using the hydrological section method. Therefore, it has strong practicality for alluvial rivers, especially wandering rivers characterized by "frequent changes in river regime and unstable migration of the mainstream". BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a schematic diagram of a flow chart of an embodiment of the present invention;

[0040] Figure 2 Schematic diagram of a binary image of a river water body and a noise reduction filter extracted in an embodiment of the present invention;

[0041] Figure 3 Schematic diagram of the skeleton line of the river channel and the left and right waterside lines extracted in the embodiment of the present invention;

[0042] Figure 4 1 is a schematic diagram of calculating the river regime relative stability index RSIRR in an embodiment of the present invention;

[0043] Figure 5 It is a schematic diagram of the river hydrological section encryption and correction process in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0045] like Figure 1 As shown, the method for determining and predicting the stability state of an alluvial river based on the channel swing amplitude of this embodiment includes the following steps:

[0046] Step S1, batch preprocessing the satellite remote sensing images of the target area to obtain a de-noised binary image of the water body;

[0047] Step S2, construction of a historical river regime database: constructing a historical river regime database based on the water body binary image;

[0048] Step S3, calculation of the relative stability index of the cross-section river regime: calculating the relative stability index of the river regime based on the historical river regime database;

[0049] Step S4, encryption and correction of hydrological sections: generating new hydrological sections, and determining the optimal number of sections for quantifying the relatively stable state of the river regime of the entire river section by correcting and encrypting the original hydrological sections;

[0050] Step S5, determination and prediction of the relative stability of the river regime: judging the river regime and predicting the future river regime based on the encrypted and corrected relative stability index of the hydrological section.

[0051] Step S1 specifically includes:

[0052] Step S11, batch processing of satellite remote sensing images, taking Landasat5 satellite images as an example, first determine the resolution Re and size M×N of the satellite remote sensing image, where M and N are the number of pixel rows and columns of the image respectively. Then set the values ​​of relevant parameters, including: the maximum size of the denoising filter σ max , the number of clustering groups of hydrological sections g, the number of encrypted sections Δn, the expected probability of unstable river regime P ex .

[0053] Step S12, calculating the modified normalized difference water index (MNDWI) of the remote sensing image one by one in the ENVI model, and the calculation expression is: Among them, Green and MIR are the reflectances of the green band and the mid-infrared band in the satellite image, respectively. The MNDWI calculation results of the above remote sensing images are used as the input data of the model for judging and predicting the relative stability of the alluvial river regime based on the channel swing amplitude, and the water body is roughly extracted from it to obtain a water body binary image. The water body binary image is as follows Figure 2 As shown in the label 1, Figure 2 The pixel value 1 represents the water pixel in the image, and 0 represents the non-water pixel, that is, the background pixel. Figure 2 As shown in the middle label 2, river beaches, sandbanks, ships and bridges are identified as non-water bodies. A noise reduction filter with a variable size of σ is used to roughly extract the interior of the water body. The pixels identified as non-water bodies due to the influence of river beaches, sandbanks, ships and bridges are filled with water bodies, where σ = 3, 5, 7, …, σ max Specific use is as follows Figure 2 The denoising filter shown by number 3 is convolved with each background pixel in the roughly extracted water body in turn: Where f(i,j) and g(i,j) are the pixel values ​​corresponding to the i-th row and j-th column of the water body binary image and the denoising filter, respectively. If the convolution result is 4σ-4, all background pixels in the convolution kernel are corrected to water bodies, and the size of the convolution kernel is gradually increased to σ max Repeat the above steps to finally obtain the de-noised binary image of the water body. The noise area that can be eliminated is σ max ×σ max ×Re 2 m 2 .

[0054] The specific steps of step S2 include:

[0055] Step S21, respectively use the edge detection method and the Zhang-Suen fast parallel thinning algorithm to extract the left water surface line, the right water surface line and the river channel skeleton line of the water body binary image from the denoised water body binary image, and respectively record the coordinates of the pixel points belonging to different line types, where the left water surface line and the right water surface line are as follows: Figure 3 As shown in the middle mark 4, the skeleton line of the river channel is as follows Figure 3 As shown in the middle mark number 5, the river skeleton line can be used as an effective substitute for the mainstream line or geometric center line of the river.

[0056] Step S22, matrixing the pixel position coordinates of the relevant river regime historical information to construct a historical river regime database. Specifically including: river channel skeleton line Left water line Right water line Hydrological section line Where t is the year corresponding to the remote sensing image, are the X and Y coordinates of the i1th measuring point in the skeleton line of the river channel in the tth year in the satellite remote sensing image, are the X and Y coordinates of the i1th measuring point in the left water surface line in the tth year in the satellite remote sensing image, are the X and Y coordinates of the i1th measuring point in the right water surface line in the tth year in the satellite remote sensing image, are the X and Y coordinates of the i1th measuring point in the hydrological section line in the tth year in the satellite remote sensing image. Figure 4 As shown in the number 6.

[0057] The specific steps of step S3 include:

[0058] Step S31, define the dimensionless river channel swing index s t =|d t / b t |,d t is the distance between the skeleton lines of the river channel on the same hydrological observation section (hydrological section) within the adjacent time interval Δt, such as Figure 4As shown in the number 7, b t is the river width corresponding to time t (year), such as Figure 4 As shown in the number 8 in the figure. The river channel swing amplitude is regarded as a continuous function that changes with time. At time t0 (year), it is expanded based on the Taylor series and the third-order and higher-order Lagrangian residual terms are omitted to obtain the river channel swing amplitude index s t , which can be expressed as: in For the historical information of river regime with the same time interval, Δt can be taken as a constant, and the relative stability index of river regime RSIRR can be expressed as a linear combination of the swing amplitude s, swing velocity v, and swing acceleration a of the river skeleton line: in, are the swing amplitude, swing speed and swing acceleration at time t0 respectively, β1, β2 and β3 are the weight coefficients of the swing amplitude s, swing speed v and swing acceleration a respectively.

[0059] Step S32: Extract and record the coordinates of the intersections of each hydrological section line and the river skeleton line at all times one by one And the coordinates of the intersection with the left water surface line and the right water surface line in The image rows and columns represent the pixel points of the k-th hydrological section and the skeleton line of the river at the t-th time (year), Respectively represent the image row and column numbers of the pixel points of the k-th hydrological section and the left water surface line at the t-th time (year), Respectively represent the image row and column numbers of pixel points of the k-th hydrological section and the right water surface line at the t-th time (year).

[0060] Calculate the value of the swing amplitude s of the river skeleton line at different times on each hydrological section one by one:

[0061]

[0062] Among them, s k,t It is the swing amplitude of the river skeleton line on the kth section at the tth time (year).

[0063] According to the calculation method of step S31, the values ​​of the channel skeleton line swing speed v and the swing acceleration a at different times on each hydrological section are calculated one by one.

[0064] Step S33: Calculate the multi-year average of the channel swing amplitude s, swing velocity v and swing acceleration a on each hydrological section and The multi-year average and As the three types of indicators for cluster analysis, the above hydrological sections (assuming the total number of sections is N) are clustered according to the three types of indicators and finally divided into g groups. After determining the hydrological sections in each cluster group, calculate and The ratio of the inter-group variation to the intra-group variation of the three indicators, that is, the F statistic: Between-group sum of squares Within-group sum of squares where n l is the number of hydrological sections in the lth category, and are the means of the variables in the lth category and all categories, respectively, and x lk It is the variable value of the kth hydrological section in the lth category. Since the larger the F statistic, the greater the dispersion of the index between categories, but the smaller the variation within the category, the F statistic of each cluster index is used as the weight of the undetermined coefficients β1, β2, and β3 in the river regime relative stability index RSIRR: F1:F2:F3=β1:β2:β3 to construct the mathematical expression of the river regime relative stability index RSIRR: RSIRR=β1s+β2v+β3a.

[0065] Step S34: Substitute the three types of indicators into the mathematical expression of the river relative stability index RSIRR, and calculate the average values ​​of all sections over the years in turn, and obtain the maximum value of the river relative stability index RSIRR in each category of the cluster analysis results. max,l With the minimum RSIRR min,l , where l = 1, 2, ..., g, the threshold of the relative stability index RSIRR of two adjacent categories is determined as RSIRR T,l =(RSIRR max,i +RSIRR min,l ) / 2, it can be seen that the larger the critical threshold, the more unstable the state of the hydrological section belonging to the corresponding group.

[0066] The specific steps of encrypting and correcting the hydrological section in step S4 include:

[0067] Step S41: Generate a new set of equally spaced hydrological sections to perform spatial interpolation on the river section, where the newly generated hydrological sections are as follows: Figure 5 As shown in the middle label 9. The number of hydrological sections in this group is n, the initial value of n is n0 = 1, the length is equal and slightly longer than the maximum water surface width of the river channel, and each newly generated hydrological section is evenly divided by the skeleton line of the river channel in the current year t and is perpendicular to the local direction of the skeleton line of the river channel. Repeat steps S32 and S34 to calculate the relative stability index RSIRR of the river regime in year t on each newly generated hydrological section k,t , the RSIRR of all newly generated hydrological sections at that momentk,t The values ​​are averaged to quantify the relative stability of the overall river regime: Where k represents the kth hydrological section in a set of newly generated hydrological sections.

[0068] Step S42: Based on the calculation result of the relatively stable state of the overall river regime, a sensitivity analysis of the number of densified sections is carried out, and n is gradually increased from 1 with a certain step length Δn of the number of densified sections, and the corresponding Wait until the result tends to be stable, and the corresponding number of sections n can be determined as the optimal number of encrypted sections for this river section.

[0069] The specific operations of determining and predicting the relatively stable state of the river regime in step S5 include:

[0070] Step S51: after encrypting and correcting the hydrological section of the river section with the optimal value n of the encrypted section, construct a river regime relative stability index matrix RSIRR1={RSIRR k,t} n×(T-3) , where the element RSIRR k,t represents the RSIRR value of the kth encrypted section at time t, T is the number of all moments, calculates the mean of the row vectors in the spatiotemporal distribution matrix RSIRR1 of the relative stability index of the river regime to obtain the spatial distribution of the multi-year stable state of the river regime, calculates the mean of the column vectors in the spatiotemporal distribution matrix RSIRR1 of the relative stability index of the river regime to obtain the time distribution of the overall stable state of the river regime, and calculates the upper and lower threshold points RSIRR of different section categories determined in step S34 T,l 、RSIRR T,l+1 , determine the section category to which each encrypted hydrological section belongs at time t, and then output the determination result of the relatively stable state of the river regime at time t (year).

[0071] Step S52: In the spatiotemporal distribution matrix RSIRR1 of the relative stability index of river regime, detect all hydrological sections whose relative stability index RSIRR value at time t-1 (year) is greater than a certain set value RSIRR', and calculate the proportion of the detected hydrological sections belonging to the unstable section category at the next time t n' represents the number of sections whose RSIRR value is greater than the set value RSIRR' at time t-1, and m' represents the number of hydrological sections detected that belong to the unstable section category at the next time t. Let RSIRR' gradually increase from 0, calculate the proportion P corresponding to different RSIRR' values ​​in turn, and finally obtain the river stability probability function P(RSIRR'). Solving this function value is equal to the expected probability of river instability P ex When: P(RSIRR')=P ex, the corresponding RSIRR', the river section micro-elements represented by all the encrypted hydrological sections whose RSIRR values ​​at the t moment are less than or equal to the RSIRR' are output as the prediction results of the future stable river sections, and the river section micro-elements represented by all the encrypted hydrological sections whose river regime relative stability index RSIRR values ​​at the moment are greater than the RSIRR' are output as the prediction results of the future unstable river sections.

[0072] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have the same meaning as in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless defined similarly here.

[0073] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for determining and predicting the stability of alluvial river regime based on river channel swing amplitude, characterized by: The following operations are included: The satellite remote sensing images of the target area are batch preprocessed to obtain the de-noised binary images of the water body; Construct a historical river regime database based on binary images of water bodies; Calculate the relative stability index of river flow at hydrological sections based on the historical river flow database; By revising and encrypting the original hydrological sections, the optimal number of sections for quantifying the relatively stable state of the river regime of the entire river section is determined; Based on the relative stability index of river flow of the encrypted and revised hydrological sections, the river flow is judged and the future flow of the river is predicted.

2. The method for determining and predicting the stability of an alluvial river regime based on river channel swing according to claim 1 is characterized in that: The batch preprocessing of the satellite remote sensing images of the target area specifically includes: Determine the resolution Re and size of the satellite remote sensing image, and set the relevant parameter values, including the maximum size σ of the noise reduction filter max , the number of clustering groups of hydrological sections g, the number of encrypted sections Δn, the expected probability of unstable river regime P ex ; The improved normalized difference water index of satellite remote sensing images is calculated, and the improved normalized difference water index of satellite remote sensing images is used to roughly extract water bodies from satellite remote sensing images to obtain water body binary images. The interior of the roughly extracted water bodies is filled with a denoising filter to obtain a denoised water body binary image.

3. The method for determining and predicting the stability of an alluvial river regime based on river channel swing according to claim 2 is characterized in that: The size of the noise reduction filter is variable, and the size is denoted by σ, and the value is σ=3, 5, 7, ..., σ max The filling process of the interior of the roughly extracted water body by using the denoising filter specifically includes: using the denoising filter to sequentially convolve with each background pixel in the roughly extracted water body: Among them, f(i,j) and g(i,j) are the pixel values ​​corresponding to the i-th row and j-th column of the water body binary image and denoising filter, respectively, σ is the size of the denoising filter, and (f*g)(i,j) is the convolution operation; If the convolution result is: 4σ-4, all background pixels in the convolution kernel are corrected to water bodies, and the size of the convolution kernel is gradually increased to σ max , and finally the denoised water body binary image is obtained.

4. The method for determining and predicting the stability of an alluvial river regime based on river channel swing according to claim 1 is characterized in that: The method of constructing a historical river regime database based on a binary water image specifically includes: extracting a left water surface line, a right water surface line and a river channel skeleton line from the binary water image after noise reduction using an edge detection method and a thinning algorithm, and recording the coordinates of pixel points belonging to different line types respectively; The river regime-related data of different years are extracted respectively, including the river skeleton line, left water surface line, right water surface line and hydrological section line, to construct a historical database of river regime.

5. The method for determining and predicting the stability of an alluvial river regime based on river channel swing according to claim 1 is characterized in that: The relative stability index of river flow of the hydrological section calculated based on the historical river flow database specifically includes: Based on Taylor series, a quantitative index for determining the relative stability of river flow is established, namely, the relative stability index of river flow in hydrological sections RSIRR, where the relative stability index of river flow in hydrological sections RSIRR is a linear combination of the swing amplitude s, swing velocity v, swing acceleration a and water surface width b of the skeleton line of the river channel; Calculate the swing amplitude s, swing velocity v and swing acceleration a of the channel skeleton line on each hydrological section in turn; Cluster analysis is performed on all hydrological sections according to the three indicators of the swing amplitude s, swing velocity v and swing acceleration a of the river skeleton line, and the corresponding weight of each indicator is determined; The relative stability index of river regime (RSIRR) of each hydrological section is calculated in turn, and the critical threshold between adjacent section categories is determined.

6. The method for determining and predicting the stability of an alluvial river regime based on river channel swing according to claim 5, characterized in that: By revising and encrypting the original hydrological sections, the optimal number of sections for quantifying the relatively stable state of the river regime of the entire river section is determined, including: Generate a set of new hydrological sections at equal intervals to perform spatial interpolation on the river section, and calculate the mean of the relative stability index RSIRR of all new hydrological sections at each moment in turn; Based on the mean value of the river regime relative stability index RSIRR, a sensitivity analysis was conducted on the number of new hydrological sections. Starting from the number of hydrological sections being 1, the number of section densification was gradually increased with a set step size Δn. The mean values ​​of the river regime relative stability index RSIRR of the corresponding hydrological sections were calculated in turn until the result tended to be stable. The corresponding number of hydrological sections was determined as the optimal number of densified sections for the corresponding river section. The hydrological sections of the river section were densified and corrected with the optimal number of densified sections.

7. The method for determining and predicting the stability of an alluvial river regime based on river channel swing according to claim 6, characterized in that: The judgment of the river regime and prediction of the future river regime based on the relative stability index of the river regime of the encrypted and revised hydrological section specifically include: Based on the encrypted and corrected relative stability index of the hydrological section, the spatiotemporal distribution matrix RSIRR1 of the relative stability index of the river flow is constructed. According to the spatiotemporal distribution matrix RSIRR1 of the relative stability index of the river flow and the critical thresholds of different section categories, the judgment result of the relative stability state of the river flow at the corresponding moment is obtained. Based on the spatiotemporal distribution matrix RSIRR1 of the relative stability index of river regime, the river regime stability probability function P(RSIRR') is calculated. The value of the river regime stability probability function P(RSIRR') is equal to the expected probability of river regime instability P ex The relative stability index of river regime RSIRR' corresponding to the time is used to obtain the prediction result of the relative stability state of river regime.

8. A system for determining and predicting the stability of alluvial river flow based on river channel swing, characterized by: The system comprises: a network interface, a memory and a processor; wherein the network interface is used to receive and send signals during the process of sending and receiving information with other external network elements; The memory is used to store a computer program that can be run on the processor; The processor is used to execute the steps of the method for determining and predicting the stability state of an alluvial river based on river channel swing amplitude as described in any one of claims 1 to 7 when running the computer program.

9. A computer storage medium, characterized in that: The computer storage medium stores a program for determining and predicting the stable state of an alluvial river based on channel swing. When the program for determining and predicting the stable state of an alluvial river based on channel swing is executed by at least one processor, the steps of the method for determining and predicting the stable state of an alluvial river based on channel swing as described in any one of claims 1 to 7 are implemented.

Images