Mountain torrent disaster risk assessment model construction method and system, early warning method, device and system

Abstract

The invention discloses a mountain torrent disaster risk assessment model construction method and system, and an early warning method, device and system, and the model construction method comprises the following steps: building a corresponding fuzzy judgment matrix based on an assessment index, and building a corresponding objective function based on the fuzzy judgment matrix; taking the objective function as a fitness function of a grey wolf optimization algorithm, using Poisson disk sampling to initialize a wolf pack position to obtain the wolf pack position, performing iterative updating on the wolf pack position based on the fitness function, and determining each evaluation weight based on a final iterative updating result; and based on a fuzzy comprehensive evaluation method, constructing a corresponding risk evaluation model by using the obtained evaluation weight, and determining a risk level corresponding to the mountain torrent disaster by the risk evaluation model based on the corresponding risk monitoring data. According to the method, the risk assessment model for mountain torrent disasters is constructed based on the fuzzy comprehensive analysis algorithm, various factors generating mountain torrents can be comprehensively analyzed, the early warning result is matched with the specific condition of a mountain area, and the false alarm rate is reduced.

Description

Technical Field

[0001] The present invention relates to the field of natural disaster early warning, in particular to a construction scheme for a risk assessment model of mountain flood disasters, and also relates to an early warning scheme for mountain flood disasters. Background Art

[0002] Mountain flood disasters are flood disasters caused by the rapid concentration of surface runoff triggered by short-term heavy rainfall or snowmelt in mountainous and hilly areas. They are characterized by strong suddenness, great destructive power, and high risk of casualties.

[0003] Traditional static risk assessment models cannot respond in real time to sudden changes in conditions such as short-term heavy rainfall and river channel blockages, and lack a quantitative description of the gradual change process of risk levels. For example, risk grading methods based on fixed thresholds (such as rainfall warning lines) are prone to misjudgment due to dynamic fluctuations in environmental parameters. Summary of the Invention

[0004] In view of the shortcoming that the prior art is prone to misjudgment in risk grading based on fixed thresholds, the present invention provides a construction scheme for a risk assessment model of mountain flood disasters implemented based on a fuzzy comprehensive analysis method, and also provides an early warning scheme for mountain flood disasters based on the risk assessment model.

[0005] To solve the above technical problems, the present invention is solved by the following technical solutions:

[0006] A method for constructing a risk assessment model of mountain flood disasters, comprising the following steps:

[0007] Configure evaluation indicators;

[0008] Based on the evaluation indicators, establish a corresponding fuzzy judgment matrix, and based on the fuzzy judgment matrix, construct a corresponding objective function, where the objective function is a constrained programming equation system for solving the evaluation weights corresponding to each evaluation indicator.

[0009] Take the objective function as the fitness function of the grey wolf optimization algorithm, initialize the positions of the wolf pack using Poisson disk sampling to obtain the positions of the wolf pack, iteratively update the positions of the wolf pack based on the fitness function, and determine each evaluation weight based on the final iteratively updated result.

[0010] Based on the fuzzy comprehensive evaluation method, use the obtained evaluation weights to construct a corresponding risk assessment model. The risk assessment model calculates the membership degrees corresponding to each evaluation grade based on the corresponding risk monitoring data, and takes the evaluation grade corresponding to the maximum membership degree as the risk grade corresponding to the mountain flood disaster, where the risk monitoring data is the monitoring data corresponding to each evaluation indicator.

[0011] As an implementable mode:

[0012] Based on the risk level of mountain flood disasters, several evaluation levels are divided from high to low. The lowest evaluation level is taken as the first category level, the highest evaluation level is taken as the second category level, and the remaining evaluation levels are taken as the third category level;

[0013] The natural breakpoint method is used to determine the breakpoints corresponding to each evaluation level;

[0014] Based on the descending trapezoidal function and the breakpoints, the membership function corresponding to the first category level is constructed;

[0015] Based on the ascending trapezoidal function and the breakpoints, the membership function corresponding to the second category level is constructed;

[0016] Based on the triangular distribution function and the breakpoints, the membership function corresponding to the third category level is constructed;

[0017] Based on each membership function and the evaluation weights, the corresponding risk assessment model is constructed.

[0018] As an implementable manner:

[0019] The evaluation indicators include slope, rainfall, and distance from the flood source.

[0020] A warning method for mountain flood disasters includes the following steps:

[0021] Taking the monitored area as the target area, monitoring the water flow velocity corresponding to the target area. When it is detected that the water flow velocity is greater than the preset velocity threshold, it is determined that a mountain flood disaster has occurred;

[0022] When it is determined that a mountain flood disaster has occurred, based on the risk monitoring data corresponding to the target area, input it into the risk assessment model constructed by any of the above methods to obtain the corresponding risk level;

[0023] Report the risk level, turn on the alarm device corresponding to the target area, and control the operation of the alarm device based on the risk level.

[0024] As an implementable manner, the alarm device is a sound device;

[0025] When receiving the corresponding warning instruction, or identifying the warning sound of other alarm devices, control the operation of the alarm device corresponding to the target area.

[0026] As an implementable manner, the method for detecting the water flow velocity of the target area includes the following steps:

[0027] Based on the inter-frame difference method, identify the floating objects in the water flow video corresponding to the target area to obtain one or more target floating objects;

[0028] Use the optical flow method to perform cross-frame tracking on each target floating object, and calculate the speed corresponding to the target floating object based on the displacement between adjacent frames of each target floating object and the time difference between frames, so as to determine the corresponding water flow velocity.

[0029] As an implementable manner:

[0030] Preprocess the water flow video corresponding to the target area to obtain the video to be recognized;

[0031] Identify the floating objects in the video to be recognized based on the inter-frame difference method to obtain the corresponding target floating objects;

[0032] Among them, the preprocessing includes contrast adjustment and / or edge enhancement algorithms to enhance the water flow boundary and speed change characteristics.

[0033] A construction system for a risk assessment model of mountain flood disasters, which is used to execute the construction method described in any one of the above.

[0034] A warning device for mountain flood disasters, which uses the monitored area as the target area and is used to identify, evaluate and warn of mountain flood disasters in the target area, including:

[0035] An identification module, which is used to monitor the water flow velocity corresponding to the target area, and when it detects that the water flow velocity is greater than a preset velocity threshold, it determines that a mountain flood disaster has occurred;

[0036] An evaluation module, which is used to input the risk monitoring data corresponding to the target area into the risk assessment model constructed by any of the above methods when it is determined that a mountain flood disaster has occurred, so as to obtain the corresponding risk level;

[0037] A warning module, which is used to report the risk level, turn on the alarm device corresponding to the target area, and control the operation of the alarm device based on the risk level.

[0038] A warning system for mountain flood disasters, including:

[0039] A warning center;

[0040] Several alarm devices;

[0041] Several warning devices, the warning devices are signal-connected to the warning center and also signal-connected to the alarm devices in the same monitoring area;

[0042] The warning device uses the monitored area as the target area. When it monitors that a mountain flood disaster occurs in the target area, it evaluates the corresponding risk level, controls the operation of the corresponding alarm device based on the risk level, and also reports the risk level to the warning center;

[0043] The warning center is used to sequentially control one or more alarm devices in the monitoring areas adjacent to the warning devices based on the risk levels reported by the warning devices.

[0044] Due to the adoption of the above technical solutions, the present invention has remarkable technical effects:

[0045] The present invention constructs a risk assessment model for mountain flood disasters based on the fuzzy comprehensive analysis algorithm, which can comprehensively analyze various factors causing mountain floods. The warning results match the specific conditions of the mountainous areas, reducing the false alarm rate;

[0046] In the design of the objective function of the present invention, the consistency requirement is transformed into a mathematical programming problem, bypassing the requirement of constructing a fuzzy consistent matrix, and the objective function is used as the fitness function of the grey wolf algorithm. By simulating the social hierarchy and hunting strategies of grey wolves, the non-linearity and multi-peak problems in weight calculation are effectively processed, quickly converging to the global optimal solution. At the same time, the algorithm has fewer parameters and is easy to adjust, which enables flexible and accurate determination of the weights of various factors in the face of complex and changeable mountain flood risk factors, improving the accuracy and reliability of risk assessment;

[0047] The present invention also further improves the grey wolf optimization algorithm by using Poisson disk, reducing the calculation amount and optimizing the search efficiency, and can quickly obtain the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0049] Figure 1 is a schematic flowchart of a warning method for mountain flood disasters of the present invention;

[0050] Figure 2 is a schematic flowchart of a method for constructing a risk assessment model for mountain flood disasters. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The following will further elaborate on the present invention in conjunction with embodiments. The following embodiments are explanations of the present invention, and the present invention is not limited to the following embodiments.

[0052] A warning method for mountain flood disasters, referring to Figure 1 , includes the following steps:

[0053] S100. Take the monitored area as the target area, monitor the water flow velocity corresponding to the target area, and when it is detected that the water flow velocity is greater than the preset velocity threshold, it is determined that a flash flood disaster has occurred;

[0054] In practical applications, several monitoring areas are pre-divided, and monitoring devices, such as cameras, are laid in the monitoring areas according to the trend of the terrain;

[0055] The monitoring device performs real-time video acquisition on the water flow in the monitored area. The obtained water flow video has characteristics such as continuity, accuracy, and high resolution, and the changes in the monitoring area are identified by recognizing the water flow video;

[0056] S110. Identify floating objects in the water flow video corresponding to the target area based on the inter-frame difference method, and obtain one or more target floating objects;

[0057] Specifically:

[0058] S111. Preprocess the water flow video corresponding to the target area to obtain a video to be recognized;

[0059] In this embodiment, the water flow video is preprocessed to improve the accuracy of subsequent video recognition.

[0060] Among them, the preprocessing methods include:

[0061] Denoising: Use denoising methods such as Gaussian filtering and median filtering to remove noise and interference in the water flow video;

[0062] Feature enhancement: Use methods such as contrast adjustment and edge enhancement algorithms to enhance key features such as water flow boundaries and speed changes;

[0063] Adjust the video frame rate: Based on actual needs, use downsampling or frame interpolation methods to adjust the video frame rate to improve video clarity and feature expression ability, and provide a more reliable data basis for subsequent water flow velocity recognition.

[0064] S112. Identify floating objects in the video to be recognized based on the inter-frame difference method, and obtain corresponding target floating objects;

[0065] The inter-frame difference method is a motion detection technology based on pixel intensity changes. Specifically, by calculating the pixel intensity differences between two or three consecutive frames, the areas with significant changes are determined through a preset difference threshold, thereby identifying moving objects;

[0066] The calculation formula for pixel intensity difference is:

[0067] D(x,y)=|I t (x,y)-I t-1 (x,y)|

[0068] Where D(x, y) is the pixel intensity of the frame difference video, and I t (x, y) represents the pixel value at the current frame coordinates (x, y), and I t-1 (x, y) represents the pixel value at the coordinates (x, y) in the previous frame.

[0069] The obtained difference video is thresholded to extract the pixel regions with significant changes to obtain the target floating objects:

[0070]

[0071] Where T is the preset difference threshold for removing pixel points with weak changes.

[0072] S120. Use the optical flow method to perform cross-frame tracking on each target floating object, and calculate the speed corresponding to the target floating object based on the displacement and inter-frame time difference between adjacent frames of each target floating object to determine the corresponding water flow velocity.

[0073] The optical flow method is a technique based on the continuity assumption of object motion in a video sequence for estimating the motion vector of an object.

[0074] S121. Calculate the displacement corresponding to the target floating object;

[0075] In this embodiment, the Lucas-Kanade algorithm is adopted. This optical flow algorithm estimates the motion of an object by minimizing the error of video brightness change. Specifically, for each tracking point, calculate its position change in the previous and next frames to obtain its motion vector;

[0076] The calculation formula for calculating the displacement of the target floating object in this embodiment is:

[0077]

[0078] Where W is the neighborhood window and (u, v) is the motion vector of the target floating object.

[0079] On this basis, the optimal solution (u, v) of the motion vector is solved by an iterative method and used as the displacement of the target floating object.

[0080] S122. Calculate the speed corresponding to the target floating object;

[0081] Based on the displacement (u, v) of the target floating object between adjacent frames and the inter-frame time difference Δt, its speed v obj can be calculated by the following formula:

[0082]

[0083] Where f is the frame rate of the video (frames per second); p is the scale factor of the pixel to the actual distance (meters per pixel).

[0084] S123. Determine the water flow velocity based on the velocities of the target floating objects;

[0085] In this embodiment, to enhance accuracy, the velocities of the target floating objects are averaged, and the obtained average velocity is used as the water flow velocity:

[0086]

[0087] In the formula, n is the number of floating objects, and v obj,i is the velocity of the i-th target floating object.

[0088] S130: Compare the obtained water flow velocity with the velocity threshold, and trigger a mountain flood warning based on the comparison result.

[0089] In this embodiment, when the water flow velocity is greater than the velocity threshold, it is determined that a mountain flood disaster has occurred, and the mountain flood warning is activated;

[0090] When the water flow velocity is less than the threshold, continue to monitor.

[0091] In this embodiment, by collecting the historical videos of the target area during a mountain flood outbreak, the corresponding water flow velocity is calculated based on the historical videos, and the minimum value of the water flow velocities corresponding to each historical video is used as the velocity threshold for the current monitoring area.

[0092] S200. When it is determined that a mountain flood disaster has occurred, input the risk monitoring data corresponding to the target area into a preset risk assessment model to obtain the corresponding risk level;

[0093] The risk assessment model calculates the membership degrees corresponding to each evaluation level based on the corresponding risk monitoring data, and takes the evaluation level corresponding to the maximum membership degree as the risk level corresponding to the mountain flood disaster, where the risk monitoring data is the monitoring data corresponding to each evaluation index.

[0094] The construction method of the risk assessment model includes the following steps:

[0095] S210. Configure the evaluation indexes;

[0096] In this embodiment, the evaluation indexes include slope, rainfall, and distance from the flood source.

[0097] Those skilled in the art can configure the evaluation indexes according to actual needs, and this embodiment does not limit them in detail.

[0098] S220. Establish a corresponding fuzzy judgment matrix based on the evaluation indicators, and construct a corresponding objective function based on the fuzzy judgment matrix;

[0099] The objective function is a constrained programming equation system, which is used to solve the evaluation weights corresponding to each evaluation indicator;

[0100] S221. Establish a fuzzy judgment matrix;

[0101] Let experts evaluate the importance of each evaluation indicator, and construct a fuzzy judgment matrix A=(r ij ) n×n , where n represents the number of evaluation indicators, and r ij represents the importance scale of the i-th evaluation indicator relative to the j-th evaluation indicator;

[0102] In this embodiment, 8 experts are invited to use the 0.1-0.9 scale to judge 3 indicators and construct a fuzzy judgment matrix.

[0103] The 0.1-0.9 scale is shown in Table 1

[0104] Table 1 0.1-0.9 scale method

[0105]

[0106]

[0107] The obtained fuzzy judgment matrix A is:

[0108]

[0109] S222. Construct an objective function according to the requirements of the fuzzy consistent matrix.

[0110] In the traditional Analytic Hierarchy Process (AHP), the calculation of weights depends on constructing a fuzzy complementary matrix and converting it into a fuzzy consistent matrix. However, this preprocessing process often reduces the accuracy of data and increases the complexity of data processing;

[0111] Aiming at the above defects, in this embodiment, the consistency requirement is transformed into a mathematical programming problem, bypassing the requirement of constructing a fuzzy consistent matrix.

[0112] Under the weight constraint condition, the relationship between the weight w and the importance scale r ij of two indicators is equivalent to solving the following constrained programming problem. In this embodiment, the objective function is:

[0113]

[0114] In the formula, i = 1, 2,..., n, ω iis the weight of evaluation index i; n is the number of evaluation indexes; r ij is an element of the fuzzy judgment matrix A; b is a measurement unit for the difference in importance between two elements, and generally takes a value of In this embodiment, n = 3; b = 1.

[0115] S230. Solve the objective function based on the grey wolf optimization algorithm to obtain the evaluation weights corresponding to each evaluation index.

[0116] The grey wolf optimization algorithm can effectively handle the non-linearity and multi-peak problems in weight calculation by simulating the social hierarchy and hunting strategies of grey wolves, quickly converge to the global optimal solution, and at the same time, the algorithm has fewer parameters and is easy to adjust. This enables it to flexibly and accurately determine the weights of various factors in the face of complex and changeable mountain flood risk factors, improving the accuracy and reliability of risk assessment.

[0117] In this embodiment, the objective function is used as the fitness function of the grey wolf optimization algorithm, and Poisson disk sampling is used to initialize the positions of the wolf pack to obtain the positions of the wolf pack. The positions of the wolf pack are iteratively updated based on the fitness function, and each evaluation weight is determined based on the final iterative update result;

[0118] The specific steps are as follows:

[0119] S231. Data configuration:

[0120] The previously constructed objective function is used as the fitness function.

[0121] Configure parameters such as the population size N, the maximum number of iterations T, and the convergence factor a; in this embodiment, the population size N is taken as 50, the maximum number of iterations T is taken as 100, and the convergence factor a is set to 2.

[0122] S232. Use Poisson disk sampling to initialize the positions of the wolf pack to obtain the positions of the wolf pack;

[0123] In the existing grey wolf optimization algorithm, the population samples are usually generated by random initialization. This method is simple and easy to implement, but due to the way of generating the initial population by random initialization, it cannot guarantee good population diversity. This means that the initial population may be too concentrated in some areas and sparse in other areas, resulting in limited search ability of the algorithm.

[0124] To address the above problems, Poisson disk sampling is added to generate population samples in this embodiment;

[0125] Poisson disk sampling is a method for generating points with a uniform distribution, where the minimum distance between these points has a determined radius.

[0126] The steps for generating population samples using Poisson disk sampling are as follows:

[0127] Parameter configuration: Predetermine the boundaries of the sampling area and the radius radius of the disk. According to the objective function, in this embodiment, the boundaries are taken as (0, 1), and the radius radius of the disk is 0.26.

[0128] Create a grid: Create a grid based on the disk radius and the boundaries of the sampling area. The grid includes a number of grid cells.

[0129] Obtain the target sampling points: When initially selecting sampling points, randomly generate a point within the sampling area as a valid sampling point.

[0130] Generate candidate sampling points: Select a valid sampling point as the target sampling point and generate a number of candidate sampling points near the target sampling point.

[0131] Check the boundaries: Ensure that the newly generated candidate sampling points are within the boundaries of the sampling area.

[0132] Calculate the grid index: Calculate the index position of the candidate sampling points in the grid.

[0133] Check the distance: Check the distance between the candidate sampling points and the existing valid sampling points.

[0134] Update the grid: Take the candidate sampling points whose distances from all valid sampling points are greater than or equal to the disk radius as new sampling points, add them to the list of valid sampling points, and update the grid.

[0135] Redetermine the target sampling point and generate new candidate sampling points near the target sampling point. Repeat the above steps until the required number of sampling points is reached or other stopping conditions are met. Finally, return all the generated sampling points, which will be used as the initial positions of the wolf pack.

[0136] In the wolf pack optimization algorithm, initializing the positions of the wolf pack using Poisson disk sampling has the following advantages:

[0137] Uniform distribution: Poisson disk sampling can ensure that the initial positions of the wolf pack are evenly distributed throughout the search space, which helps the algorithm explore the solution space from multiple directions and increases the possibility of finding the global optimal solution.

[0138] Avoid aggregation: Since the minimum distance between points is guaranteed during sampling, it is possible to avoid the initial positions of the wolf pack being too concentrated, reducing the risk of the algorithm falling into a local optimal solution.

[0139] Reduce the computational amount: Compared with random initialization, Poisson disk sampling can reduce the computational amount during the subsequent update of the positions of the wolf pack because the initial positions are already relatively dispersed, reducing the direct competition between the wolf packs.

[0140] Improve search efficiency: The evenly distributed initial positions help the algorithm cover the entire search space faster, improving search efficiency.

[0141] Parameters controllable: By adjusting the sampling radius, the degree of dispersion of the initial positions of the wolf pack can be controlled, providing additional flexibility for algorithm adjustment.

[0142] S233. Iteratively update the positions of the wolf pack based on the fitness function;

[0143] Each iterative update performs the following steps:

[0144] ①. Fitness calculation: Determine three leading wolves based on fitness.

[0145] Calculate the fitness value of each wolf according to the objective function, and save the position information of the wolf with the optimal fitness value in the population Save the position information of the wolf with the second-best fitness value in the population as Save the position information of the gray wolf with the third-best fitness value in the population as

[0146] ②. Parameter update:

[0147] Update the convergence factor a, as well as the first random coefficient and the second random coefficient The first random coefficient and the second random coefficient The initial value is a random number.

[0148] The expression is:

[0149]

[0150] In the formula, a is the convergence factor, and the convergence factor a will linearly decrease from 2 to 0 as the number of iterations increases;

[0151] and are both random variables, and both follow a uniform distribution between [0, 1];

[0152] t is the current iteration number;

[0153] T is the maximum number of iterations.

[0154] ③. Position update.

[0155]

[0156] The position of the gray wolf after the (t + 1)-th iteration is:

[0157]

[0158] In the formula:

[0159] is the position of the alpha wolf; is the position of the beta wolf; is the position of the delta wolf;

[0160] is the second position of the target wolf; is the third position of the target wolf;

[0161] represents the distance between the alpha wolf and the target wolf; represents the distance between the beta wolf and the target wolf; represents the distance between the delta wolf and the target wolf;

[0162] represents the distance that the target wolf needs to move towards the alpha wolf; represents the distance that the target wolf needs to move towards the beta wolf; represents the distance that the target wolf needs to move towards the delta wolf;

[0163] and are both the first random coefficients, which are calculated from the convergence factor a corresponding to the current iteration step and the random variable corresponding to the corresponding leading wolf ;

[0164] and are both the second random parameters, which are generated based on the random variable corresponding to the corresponding leading wolf ;

[0165] ④. Judgment of iteration completion:

[0166] Judge whether the maximum iteration number T is reached. If it is satisfied, the algorithm stops and returns the value as the finally obtained optimal solution. Otherwise, based on the obtained iterate and update again.

[0167] The obtained optimal solution reflects each evaluation weight.

[0168] S240. Construct a membership function;

[0169] S241. Determine the evaluation levels;

[0170] In this embodiment, based on the risk degree of mountain flood disasters, several evaluation levels are divided from high to low. The lowest evaluation level is taken as the first type of level, the highest evaluation level is taken as the second type of level, and the remaining evaluation levels are taken as the third type of level; and the natural breakpoint method is used to determine the segmentation points corresponding to each evaluation level;

[0171] As a specific case, the comprehensive evaluation indicators are divided into 4 levels: V = {low risk, general risk, relatively high risk, major risk}, and the evaluation sub-goal set U = {slope, rainfall, distance from flood source} is established.

[0172] Meanwhile, the natural breakpoint method is adopted to select the segmentation points for each evaluation level. As shown in Table 2 below:

[0173] Table 2

[0174]

[0175]

[0176] S242. Based on the descending half-trapezoidal function and the segmentation points, construct the membership function corresponding to the first type of level; the corresponding membership function is:

[0177]

[0178] Where:

[0179] X 1 and X 2 are the segmentation points corresponding to the evaluation indicators;

[0180] a is the monitoring data corresponding to the evaluation indicator.

[0181] S243. Based on the ascending half-trapezoidal function and the segmentation points, construct the membership function corresponding to the second type of level; the corresponding membership function is:

[0182]

[0183] Where:

[0184] X 3 and X 4 are the segmentation points corresponding to the evaluation indicators;

[0185] a is the monitoring data corresponding to the evaluation indicator.

[0186] S244. Based on the triangular distribution function and the segmentation points, construct the membership function corresponding to the third type of level; the corresponding membership function is:

[0187]

[0188] Where:

[0189] X j is the j-th segmentation point corresponding to the evaluation indicator;

[0190] a is the monitoring data corresponding to the evaluation indicator.

[0191] S250. Based on the fuzzy comprehensive evaluation method, a corresponding risk assessment model is constructed by using the obtained evaluation weights and the constructed membership function;

[0192] The risk assessment model includes a weight matrix and a membership function. The input of the risk assessment model is the risk monitoring data, that is, the monitoring data corresponding to each evaluation index, and the output is the membership degree corresponding to each evaluation level. The evaluation level corresponding to the maximum membership degree is used as the risk level.

[0193] When performing risk assessment, based on the risk monitoring data corresponding to the target area, the preset risk assessment model is input. At this time, based on the risk monitoring data and the membership function, the relative membership degree r of the i-th evaluation index at the j-th evaluation level is calculated i j, and a fuzzy matrix R is obtained;

[0194]

[0195] The weight matrix W and the fuzzy matrix R are combined and calculated to obtain a combined fuzzy matrix B, that is:

[0196] B = W × R = (b 1 , b 2 , b 3 , b 4 );

[0197] Among them, b j represents the membership degree corresponding to the j-th evaluation level.

[0198] Compare the membership degrees b corresponding to each evaluation level in the combined fuzzy matrix B j , and according to the principle of maximum membership degree, select the evaluation level where the maximum membership degree is located as the corresponding risk level.

[0199] S300. Conduct risk early warning based on the mountain flood level;

[0200] In this embodiment, the risk level is reported, the alarm device corresponding to the target area is activated, and the alarm device is controlled to work based on the risk level.

[0201] S310. Information reporting:

[0202] In this embodiment, but based on the water flow velocity, it is judged that a mountain flood disaster occurs, and after its risk is evaluated, the corresponding risk level is reported to the external early warning center, and the early warning center is timely notified to do a good job in prevention and control to reduce the loss of mountain floods; in practical applications, alarm information can also be sent to nearby villages and communities to ensure that the rescue force quickly assembles and makes preparations; through the establishment of a linkage mechanism, the rapid and accurate transmission of information is ensured, and the impact of mountain floods on people's lives is reduced.

[0203] Affected by environmental factors. Under complex terrain and weather conditions, signals may be interfered with, resulting in unstable data transmission, which affects the real-time performance and accuracy of flash flood monitoring, and further reduces the accuracy of early warning effects;

[0204] To address the above deficiencies, the signal transmission method is optimized. The specific transmission steps are as follows:

[0205] Monitor the mobile network communication status in the flash flood area;

[0206] If the mobile network is normal, transmit the signal through the mobile network;

[0207] If the mobile network is abnormal or not covered, use satellite communication for transmission.

[0208] S320. Control the alarm device to work:

[0209] In this embodiment, according to the evaluated risk level, play the flash flood warning sound through a loudspeaker, so as to notify the masses of the flash flood in time, remind them to transfer and take shelter, thereby protecting the masses and reducing flash flood losses; among them, the flash flood warning sound can be a siren sound, or it can also be a pre-recorded voice audio, etc.

[0210] In practical applications, the early warning center can issue early warning instructions based on the reported risk level and the corresponding monitoring area, so as to sequentially activate one or more alarm devices adjacent to the monitoring area, thereby transmitting the flash flood information to the potentially dangerous watershed in sequence to achieve effective downward transmission of information.

[0211] Furthermore, since the flash flood may affect the issuance of instructions, to avoid missed reports, in this embodiment, the method of controlling the alarm device to work, in addition to monitoring the flash flood disaster itself and evaluating the corresponding risk level, and receiving the early warning instructions issued by the early warning center, also collects audio data for identification and analysis. When the warning sound of other alarm devices is recognized, the alarm device corresponding to the target area is controlled to work.

[0212] In this embodiment, it is possible to identify whether there is a flash flood disaster and give an early warning through three channels, and the three channels are set in parallel and can be activated when any one signal is received, thereby reducing the possibility of missed reports and false alarms and improving the accuracy of monitoring and early warning.

[0213] This specification also proposes a construction system for a risk assessment model of flash flood disasters, which is used to execute the model construction method disclosed in the above embodiment.

[0214] This specification also proposes a flash flood warning device, which takes the monitored area as the target area and is used to identify, evaluate and give an early warning of flash flood disasters in the target area, including:

[0215] An identification module, configured to monitor the water flow velocity corresponding to the target area, and when it detects that the water flow velocity is greater than a preset velocity threshold, determine that a flash flood disaster has occurred;

[0216] An evaluation module, configured to, when it is determined that a flash flood disaster has occurred, input the risk monitoring data corresponding to the target area into a constructed risk assessment model to obtain a corresponding risk level;

[0217] An early warning module, configured to report the risk level, activate the alarm device corresponding to the target area, and control the operation of the alarm device based on the risk level.

[0218] Furthermore, the early warning module is further configured to:

[0219] Receive an early warning instruction issued by an external early warning center, and control the operation of the alarm device corresponding to the target area based on the early warning instruction.

[0220] Furthermore, the identification module includes:

[0221] A video identification unit, configured to monitor the water flow velocity and perform threshold judgment based on the water flow video of the target area;

[0222] An audio identification module, configured to monitor early warning sounds based on the ambient audio of the target area;

[0223] The early warning module is further configured to, after the audio identification module identifies an early warning sound, control the operation of the alarm device corresponding to the target area based on the identification result of the early warning sound.

[0224] This specification also proposes a flash flood disaster early warning system, including:

[0225] An early warning center;

[0226] A plurality of alarm devices;

[0227] A plurality of monitoring devices:

[0228] A plurality of early warning devices, the early warning devices are signal-connected to the early warning center, and are also signal-connected to the alarm devices and monitoring devices in the same monitoring area;

[0229] The early warning device takes the monitoring area where it is located as the target area. When it detects that a flash flood disaster has occurred in the target area, it evaluates the corresponding risk level, controls the operation of the corresponding alarm device based on the risk level, and also reports the risk level to the early warning center;

[0230] The warning center is used to sequentially control one or more alarm devices in the monitoring area adjacent to the warning device based on the risk level reported by the warning device.

[0231] In the actual application process, the warning device, the monitoring device, and the alarm device can form an electronic fence for the corresponding monitoring area.

[0232] For the device embodiments, since they are basically similar to the method embodiments, they are described relatively simply. For the relevant parts, refer to the partial description of the method embodiments.

[0233] Each embodiment in this specification is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. For the same or similar parts among the embodiments, reference can be made to each other.

[0234] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a device, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0235] The present invention is described with reference to the flowcharts and / or block diagrams of methods, terminal devices (systems), and computer program products according to the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0236] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0237] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device, so that a series of operation steps are executed on the computer or other programmable terminal device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable terminal device provide steps for implementing the functions specified in one process or multiple processes and / or blocks. Figure 1 one process or multiple processes and / or blocks Figure 1 steps for implementing the functions specified in one block or multiple blocks.

[0238] It should be noted that:

[0239] The "one embodiment" or "embodiment" mentioned in the specification means that the specific features, structures or characteristics described in connection with the embodiment are included in at least one embodiment of the present invention. Therefore, the phrases "one embodiment" or "embodiment" that appear throughout the specification do not necessarily all refer to the same embodiment.

[0240] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be construed as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0241] In addition, it should be noted that the specific embodiments described in this specification may have different shapes, names of components, etc. Any equivalent or simple changes made according to the structure, features and principles described in the inventive concept of the present invention are included in the protection scope of the present invention. Those skilled in the art to which the present invention pertains can make various modifications, supplements or use similar methods of substitution to the specific embodiments described, as long as they do not deviate from the structure of the present invention or exceed the scope defined by this claim book, they should fall within the protection scope of the present invention.

Claims

1. A method for constructing a risk assessment model for flash flood disasters, characterized in that: The following steps are involved: Configure evaluation indicators; A corresponding fuzzy judgment matrix is ​​established based on the evaluation index, and a corresponding objective function is constructed based on the fuzzy judgment matrix, wherein the objective function is a set of constraint programming equations for solving the evaluation weights corresponding to each evaluation index; The objective function is used as the fitness function of the gray wolf optimization algorithm, the wolf pack positions are initialized using Poisson disk sampling to obtain the wolf pack positions, the wolf pack positions are iteratively updated based on the fitness function, and each evaluation weight is determined based on the final iterative update result; Based on the fuzzy comprehensive evaluation method, the obtained evaluation weights are used to construct a corresponding risk assessment model. The risk assessment model calculates the membership corresponding to each evaluation level based on the corresponding risk monitoring data, and takes the evaluation level corresponding to the maximum membership as the risk level corresponding to the flash flood disaster, wherein the risk monitoring data is the monitoring data corresponding to each evaluation indicator.

2. The method for constructing a risk assessment model for flash flood disasters according to claim 1, characterized in that: Based on the risk level of flash flood disasters, it is divided into several evaluation levels from high to low, with the lowest evaluation level as the first level, the highest evaluation level as the second level, and the remaining evaluation levels as the third level; The natural breakpoint method is used to determine the split points corresponding to each evaluation level; constructing a membership function corresponding to the first class level based on the descending semi-trapezoidal function and the segmentation point; constructing a membership function corresponding to the second class level based on the ascending semi-trapezoidal function and the split point; Based on the triangular distribution function and the segmentation point, constructing a membership function corresponding to the third class level; A corresponding risk assessment model is constructed based on each membership function and the evaluation weights.

3. The method for constructing a risk assessment model for flash flood disasters according to claim 1 or 2, characterized in that: The evaluation indicators include slope, rainfall and distance from flood sources.

4. A flash flood disaster early warning method, characterized in that: The following steps are involved: The monitoring area is taken as the target area, and the water flow velocity corresponding to the target area is monitored. When it is detected that the water flow velocity is greater than a preset flow velocity threshold, it is determined that a flash flood disaster occurs; When it is determined that a flash flood disaster occurs, the risk assessment model constructed by any of the methods of claims 1 to 3 is input based on the risk monitoring data corresponding to the target area to obtain a corresponding risk level; The risk level is reported, and the alarm device corresponding to the target area is turned on, and the operation of the alarm device is controlled based on the risk level.

5. The flash flood disaster early warning method according to claim 4, characterized in that: The alarm device is an audible device; When the corresponding warning command is received, or the warning sound of other alarm equipment is recognized, the alarm equipment corresponding to the target area is controlled to work.

6. The flash flood disaster early warning method according to claim 4 or 5, characterized in that: The method for detecting the water flow velocity in a target area comprises the following steps: Identify floating objects in the water flow video corresponding to the target area based on the inter-frame difference method to obtain one or more target floating objects; The optical flow method is used to track each target floating object across frames, and the speed corresponding to the target floating object is calculated based on the displacement of each target floating object between adjacent frames and the time difference between frames to determine the corresponding water flow velocity.

7. The flash flood disaster early warning method according to claim 6, characterized in that: Preprocess the water flow video corresponding to the target area to obtain the video to be identified; Identify floating objects in the video to be identified based on an inter-frame difference method to obtain corresponding target floating objects; The preprocessing includes contrast adjustment and / or edge enhancement algorithm to enhance the water flow boundary and velocity change characteristics.

8. A system for constructing a risk assessment model for flash flood disasters, characterized in that: Used to execute the construction method described in any one of claims 1 to 3.

9. A flash flood disaster warning device, which uses the monitoring area as the target area and is used to identify, evaluate and warn flash flood disasters in the target area, characterized in that: include: An identification module is used to monitor the water flow velocity corresponding to the target area, and when it is detected that the water flow velocity is greater than a preset flow velocity threshold, it is determined that a flash flood disaster occurs; An assessment module, for inputting the risk monitoring data corresponding to the target area into a risk assessment model constructed by any method of claims 1 to 3 to obtain a corresponding risk level when it is determined that a flash flood disaster occurs; The early warning module is used to report the risk level, activate the alarm device corresponding to the target area, and control the operation of the alarm device based on the risk level.

10. A flash flood disaster early warning system, characterized in that: include: Early Warning Center; Several alarm devices; A plurality of early warning devices as claimed in claim 9, wherein the early warning devices are connected to the early warning center signal and are also connected to the alarm device signal in the same monitoring area; The early warning device takes the monitoring area as the target area, and when a flash flood disaster is detected in the target area, evaluates the corresponding risk level, controls the operation of the corresponding alarm device based on the risk level, and also reports the risk level to the early warning center; The early warning center is used to sequentially control the operation of one or more alarm devices in the monitoring area adjacent to the early warning device based on the risk level reported by the early warning device.

Images