Mountain torrent disaster early warning method, device, equipment and medium

Abstract

The invention discloses a mountain torrent disaster early warning method, device and equipment and a medium, and relates to the technical field of hydrological monitoring and disaster early warning. The method comprises the steps of determining theoretical confluence time according to static attributes of a target drainage basin, determining an observation window of a rainfall event according to the theoretical confluence time, and performing sliding window analysis on a rainfall process of the rainfall event by using the observation window to obtain accumulated rainfall in the observation window, according to runoff reduction analysis, the accumulated rainfall is converted into effective peak flow, the effective peak flow is compared with disaster-causing flow, and early warning is given out according to a comparison result. According to the early warning method, the observation window is associated with the flow concentration time of the drainage basin, and the peak rainfall related to the flow concentration time of the drainage basin is captured based on the accumulated rainfall counted by the observation window, so that the rainfall process is prevented from being split by statistics in a fixed time period, and the early warning effectiveness is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydrological monitoring and disaster warning, and in particular to a mountain flood disaster warning method, device, equipment and medium. BACKGROUND

[0002] Mountain flood disaster has the characteristics of strong suddenness, great destructive power and concentrated influence range, and its occurrence is closely related to the rainfall process in the basin. Timely and accurate warning is the key to reducing casualties and property losses.

[0003] In the existing mountain flood disaster warning technology, the warning monitoring data mainly depends on the cumulative rainfall statistics of fixed time periods (such as 24 hours, 3 hours, 1 hour). When the rainfall reaches the preset fixed threshold, the warning is triggered. However, there are the following shortcomings: a continuous rainstorm may span multiple statistical periods, and the fixed period statistics method will disperse the total rainfall to different period calculations, which cannot accurately reflect the concentrated impact of rainfall on the basin. In addition, the conditions such as topography, vegetation and soil of different basins are different, resulting in different response speeds to rainfall. Inappropriate fixed time periods cannot effectively identify the key rainfall period that leads to the formation of flood peaks. Different basins have complex runoff and confluence factors, and the warning indicators are diversified. It is difficult to set parameters for warning sites, and the warning accuracy is low. There may be cases where the rainfall statistics value is large but there is no disaster, or the statistics value is small but a mountain flood occurs.

[0004] Therefore, the cumulative rainfall statistics based on fixed time periods cannot guarantee effective warning. SUMMARY

[0005] The present application provides a mountain flood disaster warning method, device, equipment and medium, which can dynamically capture the key rainfall process combined with the characteristics of the basin, to solve the technical problem of insufficient warning effectiveness of the existing warning technology.

[0006] The present application is realized by the following technical solutions:

[0007] In a first aspect of the present application, a mountain flood disaster warning method is provided, comprising:

[0008] determining a theoretical confluence time according to the static attributes of the target basin;

[0009] determining an observation window of a rainfall event based on the theoretical confluence time;

[0010] performing sliding window analysis on the rainfall process of the rainfall event using the observation window to obtain the cumulative rainfall in the observation window;

[0011] converting the cumulative rainfall into effective peak-forming flow according to runoff reduction analysis;

[0012] Compare the effective peak flow with the disaster-causing flow of the analysis survey, and issue a warning according to the comparison result.

[0013] Further, the effective peak flow is calculated from the cumulative rainfall according to the runoff reduction analysis, including:

[0014] Obtain a plurality of influence factors of the target basin, and the influence factors at least include the current water level, soil moisture, air temperature and humidity;

[0015] The influence factors are normalized to obtain the normalized values of the influence factors;

[0016] According to the predetermined weight of each influence factor, the weighted calculation of the normalized influence factors is obtained to obtain the conversion coefficient;

[0017] The cumulative rainfall is converted into effective peak flow based on the conversion coefficient, wherein the effective peak flow ; The cumulative rainfall is represented by K, and the conversion coefficient is represented by K.

[0018] Further, the method further comprises:

[0019] According to the comparison result of the effective peak flow triggering the warning and the measured flow, the conversion coefficient used for runoff reduction analysis is corrected;

[0020] According to the conversion coefficient obtained by correcting N rainfall events, a sample array is generated, and the sample array is applied to the mountain remote sensing warning site of the same or similar basin.

[0021] Further, the conversion coefficient used for runoff reduction analysis is corrected according to the comparison result of the effective peak flow triggering the warning and the measured flow, including:

[0022] After triggering the warning each time, record the influence factors and their data used by this warning, the effective peak flow RX calculated and the actual monitored flood peak flow ;

[0023] Calculate the error of this time: ;

[0024] If the error exceeds the set error threshold, the conversion coefficient K is corrected, and the correction method includes adjusting the weight of each influence factor, recording the corrected conversion coefficient K as a sample.

[0025] Further, the observation window of the rainfall event is determined based on the theoretical concentration time, including:

[0026] Determine a time range containing the theoretical concentration time;

[0027] a plurality of time nodes are taken within the time range, and a value corresponding to each of the time nodes is taken as a window length of an observation window, so as to obtain a plurality of observation windows;

[0028] The early rainfall process of the rainfall event is analyzed by using the plurality of observation windows in a sliding window manner, so as to obtain a plurality of early cumulative rainfall sequences; wherein each of the early cumulative rainfall sequences is obtained by continuous observation by using an observation window.

[0029] The correlation between each of the early cumulative rainfall sequences and an early water level rising rate is analyzed, and an observation window corresponding to an early cumulative rainfall sequence with the largest correlation is selected as a target observation window; wherein the early water level rising rate is obtained by monitoring a water level-flow relationship of the target river basin during the early rainfall process.

[0030] The subsequent rainfall process of the rainfall event is analyzed by using the target observation window in a sliding window manner, so as to obtain a cumulative rainfall amount in the target observation window.

[0031] Further, the cumulative rainfall amount obtained by analyzing the rainfall process of the rainfall event by using the observation window is represented as:

[0032]

[0033] wherein, represents a cumulative rainfall amount of the observation window at an observation time , represents a cumulative rainfall amount in a time unit before the observation time , represents a cumulative rainfall amount in a time unit before the observation time , represents a cumulative rainfall amount in a time unit before the observation time , and X represents a window length of the observation window.

[0034] Further, the static attributes include a terrain slope, a vegetation coverage, a soil type and a soil water content.

[0035] In a second aspect, the present application provides a mountain flood disaster early warning device, comprising:

[0036] a first determining module configured to determine a theoretical concentration time according to static attributes of a target river basin;

[0037] a second determining module configured to determine an observation window of a rainfall event based on the theoretical concentration time;

[0038] a first analyzing module configured to analyze a rainfall process of the rainfall event by using the observation window in a sliding window manner, so as to obtain a cumulative rainfall amount in the observation window;

[0039] a second analysis module configured to convert the accumulated rainfall into effective peak-forming flow according to runoff reduction analysis;

[0040] a warning module configured to compare the effective peak-forming flow with a disaster-causing flow obtained through analysis and investigation, and issue a warning according to a comparison result.

[0041] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the mountain torrent disaster warning method according to any one of the first aspect.

[0042] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executable on a processor to implement the mountain torrent disaster warning method according to any one of the first aspect.

[0043] Compared with the prior art, the present application has the following advantages and beneficial effects: the length of the observation window is determined based on the theoretical confluence time, so that the accumulated rainfall based on the observation window can capture the effective peak-forming flow related to the confluence time of the basin, and the complete rainfall process of the rainfall event is analyzed through the sliding observation window, avoiding the fragmentation of the rainfall process caused by the fixed period statistics. Based on the accurate statistics, the warning response time is effectively shortened based on the comparison result between the effective peak-forming flow and the disaster-causing flow obtained through statistical analysis. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present application, the following will briefly introduce the drawings needed in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be considered as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor. In the drawings:

[0045] Figure 1 is a flowchart of the mountain torrent disaster warning method;

[0046] Figure 2 is one of the flowcharts of the sliding window analysis of the rainfall process using the observation window;

[0047] Figure 3 is another flowchart of the sliding window analysis of the rainfall process using the observation window;

[0048] Figure 4 is a flowchart of dynamically correcting the warning parameters according to the monitoring data;

[0049] Figure 5 is a structural block diagram of the mountain torrent disaster warning device. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0051] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims, and accompanying drawings of this invention are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to other steps or units inherent in the device.

[0052] The terminology used in the various embodiments of the invention is for the purpose of describing particular embodiments only and is not intended to limit the various embodiments of the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of the invention pertain. The terms (such as those defined in a generally used dictionary) are to be interpreted as having the same meaning as in the context of the relevant technical field and are not to be interpreted as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of the invention.

[0053] The method proposed in this invention determines key parameters by combining the static attributes of the target watershed, dynamically constructs observation windows to capture rainfall processes, obtains effective runoff through runoff reduction analysis, and finally achieves early warning by comparing with historical thresholds. It is particularly suitable for real-time monitoring and early warning of flash floods in small watersheds and can be widely applied to disaster prevention and control work in departments such as water conservancy, emergency management, and meteorology.

[0054] Please see Figure 1 An embodiment of the present invention provides a flash flood disaster early warning method, which includes the following steps.

[0055] S110, determine the theoretical confluence time based on the static properties of the target watershed.

[0056] Confluence time refers to the time required for rainwater to flow from the farthest point in the basin to the outlet section (early warning monitoring point), and it is closely related to the basin's static properties and rainfall process. During flood formation, peak flow is often dominated by short-duration heavy rainfall, rather than the total rainfall over a fixed period. Therefore, the portion of rainfall that contributes the most to the flood peak (i.e., peak rainfall) is closely related to the confluence time of the basin.

[0057] The core basic parameter of determining the observation window based on the confluence time determines the time range that can accurately capture the disaster-causing rainfall. For a basin with sufficient historical data, the actual confluence time can be deduced by analyzing the relationship between the rainfall start time and the flood peak arrival time in historical flood events of the basin. For more common small basins without data, the confluence time can be estimated based on the static properties of the basin using empirical formulas.

[0058] Among the static properties of the target basin, the available data include terrain slope, vegetation coverage, soil type, and soil moisture content, etc. The terrain slope can be extracted from public digital elevation model (DEM) data, and the average slope of the basin or the slope of the key river channel can be calculated through geographic information system (GIS) software. The vegetation coverage can be obtained based on satellite remote sensing images (such as NDVI data), or the vegetation survey data published by the local ecological environment department can be referenced. The soil type refers to the national soil survey database to determine the texture (such as sandy soil, clay soil) and permeability characteristics of the dominant soil in the basin, and sensors can also be deployed on site for monitoring. The soil moisture content can be monitored in real time by soil moisture sensors deployed in the basin, or the initial moisture content can be estimated based on the impact of previous rainfall.

[0059] Specifically, the empirical formula method for calculating the confluence time can use the Horton formula or the SCS curve number method, which is suitable for areas lacking detailed topographic and rainfall data. Among them, the Horton formula is as follows:

[0060]

[0061] wherein, is the confluence time (hours), is the slope flow length (kilometers), is the slope gradient (dimensionless, the ratio of elevation difference to horizontal distance), is the confluence time constant, , is a regional empirical parameter obtained by rating adjacent similar basins or consulting regional hydrological manuals.

[0062] Other empirical formulas can also be used, such as the simplified form: wherein, is the slope flow length, is the confluence time constant, is an empirical index, usually between 0.5 and 0.8.

[0063] The SCS curve number method is mainly used to estimate runoff, which determines a curve number CN based on soil moisture conditions, soil type, and previous rainfall, and then calculates effective rainfall and runoff. The confluence time can be indirectly estimated based on the CN value and the characteristics of the basin.

[0064] It should be understood that the calculation of the theoretical concentration time in this step can be performed by known methods, and an estimation method suitable for the target basin can be found, and the present application does not limit this.

[0065] In S120, an observation window of the rainfall event is determined based on the theoretical concentration time.

[0066] The observation window is a time interval for dynamically statistically accumulating rainfall, and the length of the window is associated with the theoretical concentration time to ensure that the statistical rainfall process matches the basin concentration characteristics and avoids the limitations of a fixed period.

[0067] In one specific implementation, the length of the observation window is taken as the theoretical concentration time . This method directly takes the theoretical concentration time as the length of the observation window, relies on the selection of a more reasonable calculation method for the theoretical concentration time, and relies on the reliability of the theoretical concentration time to accurately capture the peak rainfall.

[0068] In one specific implementation, the theoretical concentration time is corrected in combination with the current rainfall intensity, and the corrected is taken as the length of the observation window. When the real-time rainfall intensity is greater than a certain threshold (for example, the historical average rainfall intensity), it indicates that the current rainfall is strong, and the actual concentration time may be shortened, so the length of the observation window is set to be less than the calculated theoretical concentration time, such as 0.8 ; the real-time rainfall intensity is less than a certain threshold (for example, 0.8 of the historical average rainfall intensity), which is set to be greater than the calculated theoretical concentration time, such as 1.2 .

[0069] In one specific implementation, the theoretical concentration time is corrected in combination with the antecedent rainfall, and the corrected is taken as the length of the observation window. When the antecedent rainfall is large (the soil is relatively wet), the runoff is fast, and the concentration time is shortened, so the window length can be appropriately shortened, and vice versa.

[0070] In one specific implementation, the theoretical concentration time is corrected in combination with the current rainfall intensity and the antecedent rainfall, and the corrected is taken as the length of the observation window.

[0071] The catchment concentration time is not a fixed value, but changes dynamically with rainfall intensity and soil moisture. During heavy rainfall, surface runoff is quickly generated, and the concentration time is shortened. When the previous rainfall is high (soil saturation), there is less infiltration of rainwater, and the proportion of surface runoff is high, which accelerates the concentration speed. Dynamic correction of window length can synchronize the observation window with the actual concentration process, avoid the problem of lagging observation due to fixed window length, or missing measurement due to short window length in the case of soil drought in the early stage, and realize dynamic matching between the early warning monitoring process and the real-time state of the basin.

[0072] The concentration characteristics of different basins (such as small mountainous basins and plain basins) or different seasons (rainy season and dry season) in the same basin are significantly different. The dynamic correction mechanism does not need to set window length parameters for different scenarios separately. Through real-time data automatic adaptation, the method can be applied to multiple types of basins and multiple climate conditions, reducing the cost of manual parameter adjustment.

[0073] As the early warning monitoring time delay, the historical data of the target basin is supplemented, and the theoretical concentration time can be corrected based on the historical data , and the corrected is applied to the real-time early warning process.

[0074]

[0075] Sliding window analysis refers to moving the observation window along the rainfall time axis by each time period (e.g., moving every 5 minutes), and real-time statistics of the cumulative rainfall in the window, so as to dynamically capture the rainfall intensity changes in different time periods, and accurately capture the peak rainfall.

[0076] Specifically, the cumulative rainfall is represented by the following formula:

[0077]

[0078] In the formula, represents the cumulative rainfall in the observation window at the observation time , which is the core variable of the analysis, representing the rainfall in the previous time unit at any time point in the rainfall process of a rainfall event. is the cumulative rainfall at time point , and is the cumulative rainfall in the previous time unit at time point , and X is the window length of the observation window .

[0079] ​​For a small watershed, X can take 1-3 hours, for a large watershed, X can take 6-12 hours or even longer.

[0080] In a complete rainfall event, The maximum value of is the peak rainfall of the storm, which is more representative of the concentrated impact on the watershed than any fixed period of statistical value.

[0081] S140, according to the runoff reduction analysis, the cumulative rainfall is converted into effective peak flow.

[0082] Rainfall is not all converted into surface runoff, part of which is lost through evapotranspiration, plant interception, infiltration, filling of surface depressions, etc. The core of runoff reduction analysis is to determine a reduction coefficient , which reduces the peak rainfall (this parameter in the real-time warning scenario is the cumulative rainfall observed in the sliding analysis process using the observation window ).

[0083] The reduction coefficient integrates various factors affecting runoff, and its value is between 0 and 1. There are many methods in the prior art to determine . Including but not limited to fixed empirical coefficient method, antecedent rainfall index method and hydrological model inversion method, etc.

[0084] Among them, the fixed empirical coefficient method is an easy-to-operate method, which assigns a fixed K value to the entire watershed by consulting the hydrological manual or based on local experience through the static properties of the watershed such as soil type, land use (such as vegetation coverage, urbanization degree). For example, for forest areas with strong soil permeability, 0.2~0.4 can be taken; for mountainous areas with saturated soil or high proportion of impervious area in cities, 0.7~0.9 can be taken. Reduction is the reduction of the current rainfall, and the reduction of the window cumulative rainfall over time.

[0085] The antecedent rainfall index method indirectly reflects the dry and wet state of the watershed, as well as the bottom water and other conditions by calculating the antecedent influence factor , and determines the value through the pre-established relationship curve or calculation table. Generally, the larger the value, the more humid the soil, the stronger the runoff capacity of the watershed, the higher the value.

[0086] For a watershed with long-term historical data, a hydrological model can be calibrated, and by inputting the of the historical storm and the corresponding actual runoff , the integrated equivalent of the storm can be calculated. Values.

[0087] Preferably, the present application determines the conversion coefficient K by the following method:

[0088] S210, obtaining a plurality of influence factors of the target basin, wherein the influence factors at least include the current water level, soil moisture, air temperature and humidity;

[0089] S220, normalizing each influence factor to obtain the normalized value of each influence factor;

[0090] S230, according to the weight of each influence factor, weighted calculation of each normalized influence factor to obtain the conversion coefficient.

[0091] For example, the normalized values of the current water level, soil moisture, air temperature and humidity are respectively denoted as C, S, T and H, and their weights are respectively denoted as k1, k2, k3 and k4. The conversion coefficient K = C×k1 + S×k2 + T×k3 + H×k4 is calculated.

[0092] Among them, the current water level reflects the base flow of the river, and the higher the water level, the more saturated the soil, and the larger the runoff coefficient K. The ratio of the current water level to the warning water level is used for normalization; soil moisture, i.e. soil water content, is usually expressed by volume water content or relative humidity. The higher the soil moisture, the lower the infiltration capacity, and the larger the runoff coefficient K. The ratio of the current soil water content to the field water capacity is used for normalization; air temperature affects evaporation, and the higher the air temperature, the greater the evaporation, and the smaller the runoff coefficient K. The normalized value is obtained by dividing the standard air temperature value and taking the inverse; the larger the air humidity, the smaller the evaporation, and the larger the runoff coefficient K. The relative humidity is used as the normalized value.

[0093] The above influence factors can be collected in real time by field sensors, and are not limited to the above four types, but can also include other influence factors or other combinations.

[0094] Finally, the cumulative rainfall is converted into effective peak-forming flow based on the conversion coefficient K, and the effective peak-forming flow calculation formula is:

[0095] ;

[0096] In the formula, The cumulative rainfall is denoted by K, and the conversion coefficient is denoted by K.

[0097] RX is a corrected effective rainfall that is closer to the actual runoff. It eliminates non-runoff parts such as evaporation and infiltration, so it can more accurately reflect the flood potential of a single rainstorm event.

[0098] S150, comparing the effective peak-forming flow with the disaster-causing flow analyzed and investigated, and issuing a warning according to the comparison result.

[0099] The disaster-causing flow can be obtained by statistical analysis of the minimum effective runoff corresponding to the mountain flood disaster in the history of the target basin, and is the core basis for determining whether to trigger the warning. Specifically, the historical disaster-causing runoff threshold method can be as follows: collecting historical mountain flood disaster event records of the target basin, extracting rainfall data and flood water level, disaster impact range and other basin response data corresponding to each event, using the runoff reduction method of step S140, calculating the effective runoff when each event occurs, and taking the minimum value of the effective runoff as the historical disaster-causing runoff threshold. If the historical data of the target basin is insufficient, the data of similar basins can be referred to and adjusted in combination with local static properties.

[0100] The method of issuing a warning according to the comparison result is realized by warning level division and triggering mechanism. For example, when the effective runoff is less than the historical disaster-causing runoff threshold, the basin has no mountain flood risk, and no warning is issued; when , the basin has potential mountain flood risk, and a blue warning is issued to prompt relevant departments to strengthen monitoring; when , the basin has a high probability of mountain flood, and a yellow warning is issued to notify the affected area of the people to prepare for transfer; when , the basin will soon have a serious mountain flood, and a red warning is issued to immediately organize the affected area of the people to transfer urgently.

[0101] The warning information is published through multiple channels such as short message, emergency broadcast, hydrological warning APP, social media, etc. The information content needs to include the warning level, the predicted impact range, the duration and the response suggestion (such as transfer route, safe place location).

[0102] Further, step S120 preferably uses the following method to determine the observation window of the rainfall event: first, determine a time range containing the theoretical concentration time , such as , ; then take multiple time nodes in the time range, such as every , and take the value corresponding to each time node as the window length of the observation window, thereby obtaining multiple observation windows.

[0103] In this embodiment, multiple observation windows are used to simultaneously perform sliding window analysis on the rainfall process of the rainfall event, and the cumulative rainfall observed by each observation window is obtained respectively. All the cumulative rainfall obtained at the current time is converted into effective runoff, and the multiple effective runoffs are compared with the historical disaster-causing runoff threshold respectively, and a warning is issued according to the comparison result. When the comparison result of any effective runoff and the historical disaster-causing runoff threshold triggers the preset warning condition, a warning information is issued. ​

[0104] Further, referring to the flowchart shown in FIG. 1, when multiple observation windows are used for analysis, the method of using observation windows to perform sliding window analysis on the rainfall process of a rainfall event in step S130 can also be the following steps: Figure 2 Further, referring to the flowchart shown in FIG. 1, when multiple observation windows are used for analysis, the method of using observation windows to perform sliding window analysis on the rainfall process of a rainfall event in step S130 can also be the following steps:

[0105] S310, using multiple observation windows to perform sliding window analysis on the early rainfall process of a rainfall event respectively, to obtain early cumulative rainfall sequences; wherein each early cumulative rainfall sequence corresponds to an observation window obtained by continuous observation of the observation window within a preset time length. S310, using multiple observation windows to perform sliding window analysis on the early rainfall process of a rainfall event respectively, to obtain early cumulative rainfall sequences; wherein each early cumulative rainfall sequence corresponds to an observation window obtained by continuous observation of the observation window within a preset time length.

[0106] S320, analyze the correlation between each early cumulative rainfall sequence and the early water level rising rate, and select the observation window corresponding to the early cumulative rainfall sequence with the largest correlation as the target observation window. ; wherein the early water level rising rate is obtained by monitoring the water level of the target basin during the early rainfall process.

[0107] S330, using the target observation window to perform sliding window analysis on the subsequent rainfall process of the rainfall event, to obtain the cumulative rainfall within the target observation window.

[0108] To calculate the correlation between the early cumulative rainfall sequence and the early water level rising rate, the water level rising rate can be first processed as data with the same time dimension as the early cumulative rainfall sequence, and then normalized, and then the similarity between the two is calculated, and the similarity of different sequences is compared to select the best one.

[0109] The early stage of rainfall (such as 1-3 hours after the start of rainfall) usually has low rainfall intensity and low disaster risk, which can be used as a key period for determining the target observation window, and this stage does not affect the early warning. During the early rainfall process, the basin state is relatively stable, and the correlation between water level response and rainfall is clearer, and it is not easy to be disturbed by complex factors, so the target observation window can be selected based on the data of this stage, which can select a window that adapts to the characteristics of the basin before the disaster risk increases.

[0110] Further, referring to the flowchart shown in FIG. 1, when multiple observation windows are used for analysis, the method of using observation windows to perform sliding window analysis on the rainfall process of a rainfall event in step S130 can also be the following steps: Figure 3 Further, referring to the flowchart shown in FIG. 1, when multiple observation windows are used for analysis, the method of using observation windows to perform sliding window analysis on the rainfall process of a rainfall event in step S130 can also be the following steps:

[0111] S410, using multiple observation windows to perform sliding window analysis on the rainfall process of a rainfall event respectively, to obtain time period cumulative rainfall sequences; wherein each time period cumulative rainfall sequence corresponds to an observation window obtained by continuous observation of the observation window within a preset time length. ​​

[0112] S420, analyze the correlation between each period cumulative rainfall sequence and period water level rising rate, and select the observation window corresponding to the period cumulative rainfall sequence with the largest correlation as the target observation window as the target observation window , and obtain the cumulative rainfall in the target observation window;

[0113] The period cumulative rainfall sequence participating in the correlation calculation and the period water level rising rate are obtained in the same period, and are continuously monitored in the same preset time length. The period water level rising rate is obtained by monitoring the water level change of the target basin. The preset time length is longer than the maximum window length of the observation window , and shorter than the total observation time of the rainfall event.

[0114] In a complex rainfall scenario, a single window is difficult to adapt continuously, and the real-time election mechanism can cope with sudden situations through multi-window parallel monitoring and dynamic switching. The present embodiment utilizes multiple observation windows to participate in the whole monitoring process, and can obtain multiple observation window observed cumulative rainfall sequences. The preset time length is longer than the maximum window length of the observation window , and shorter than the total observation time of the rainfall event. In the sliding window analysis process, the cumulative rainfall observed by each observation window according to the preset time length is extracted to form the period cumulative rainfall sequence of the period, and the period cumulative rainfall sequence is calculated with the period water level rising rate in the same period. The observation window corresponding to the period cumulative rainfall sequence with the largest correlation in the current period is selected as the target observation window , that is, the target observation window selected in each period may be different.

[0115] The real-time election mechanism can be to recalculate the correlation and update the target observation window every interval preset time length (such as 30 minutes), which can quickly respond to the change of the basin state. For example, when the rainfall intensity increases from 5mm / h to 20mm / h, the window length can be reduced in the next election period to capture the peak rainfall formed by heavy rainfall in time and avoid the risk of early warning lag.

[0116] In the process of analyzing the cumulative rainfall for runoff reduction to obtain the effective runoff for early warning judgment, the cumulative rainfall adopts the cumulative rainfall observed by the target observation window.

[0117] ​The embodiment method is more suitable for various dynamic changes in the rainfall process. For example, in a rainfall event, the rainfall intensity can present various forms such as weak-strong-weak-continuous heavy rainfall, resulting in the change of the catchment concentration characteristics over time. In the early stage of rainfall, the intensity is low, the actual concentration time is long, and a long observation window is needed; in the middle stage of rainfall, the intensity suddenly increases, the concentration time is shortened, and the window length needs to be shortened; in the late stage of rainfall, the intensity decreases, and the concentration time is prolonged again. By selecting the target window in real time within the preset time length, it can be ensured that the window length of each period matches the current concentration characteristics, and compared with the fixed method after selecting the window early, it can better adapt to the dramatic fluctuations of rainfall intensity.

[0118] Further, after completing a warning process for a target catchment, a step of feedback correction of the warning parameters is further included, specifically: correcting the conversion coefficient used for runoff reduction analysis according to the comparison result of the effective peak flow triggering the warning and the measured flow; generating a sample array according to the conversion coefficients obtained by correcting N rainfall events. Referring to FIG. 6, the specific correction method can include the following steps: Figure 4

[0119] S510, after triggering the warning each time, recording the influence factors and their data used in this warning, the effective peak flow RX calculated and the actual monitored flood peak flow ;

[0120] S520, calculating the error of this time: ;

[0121] S530, if the error exceeds the set error threshold, the conversion coefficient K is corrected. The correction method can be to adjust the weight of each influence factor, or to adjust the K value according to the error, and then record the corrected conversion coefficient K and store it as a sample.

[0122] S540, by collecting a plurality of such samples, a sample array is formed. The sample array is fitted according to (multi-factor, corrected K value), and the fitted curve can be applied to the initial setting of similar catchments, or used to obtain the K value by table lookup or interpolation in the subsequent warning of the same catchment.

[0123] The sample array will be applied to the mountain remote warning station of the same or similar catchment. According to the formation form of the sample array, the following two specific application forms can be constituted, that is:

[0124] Method one: according to the sample array, the weight of each influence factor is recalculated, and the conversion coefficient K is obtained according to the weight of each influence factor and the normalized value of each influence factor;

[0125] Method two: according to the values of the current plurality of influence factors, the most similar sample is found (or interpolated) from the sample array, and the K value of the sample is used as the attenuation conversion coefficient K.​

[0126] Finally, by modularizing the early warning method and applying it to mountainous remote early warning sites, intelligent and accurate early warning is realized.

[0127] Embodiments of the present application also provide a mountain flood disaster early warning device, as shown in the figure, the device comprises the following functional modules. Figure 5

[0128] The first determination module 100 is configured to determine a theoretical concentration time according to the static properties of the target basin.

[0129] The second determination module 200 is configured to determine an observation window of a rainfall event based on the theoretical concentration time.

[0130] The first analysis module 300 is configured to perform sliding window analysis on the rainfall process of the rainfall event using the observation window to obtain the cumulative rainfall in the observation window.

[0131] The second analysis module 400 is configured to convert the cumulative rainfall into effective peak-forming flow according to runoff reduction analysis.

[0132] The early warning module 500 is configured to compare the effective peak-forming flow with the disaster-causing flow obtained through analysis and investigation, and issue an early warning according to the comparison result.

[0133] The first determination module 100 can communicate with a hydrological monitoring station to obtain the static property information of the target basin from the hydrological monitoring station, or be connected to an external device through a data transmission interface and transmit relevant data through the external device.

[0134] After the first determination module 100 obtains the static property information of the target basin, the theoretical concentration time is calculated and sent to the second determination module 200.

[0135] After the second determination module 200 obtains the theoretical concentration time, the window length data related to the theoretical concentration time is determined according to the internal preset algorithm as the window length of the observation window, and the data is stored. The second determination module 200 can also update the stored window length data according to the internal preset algorithm.

[0136] Further, when the observation window is set to multiple, the election method of the target observation window can also be executed in the first analysis module 300 (see steps S310-S330, S410-S420 in the foregoing embodiments).

[0137] After the first analysis module 300 completes the sliding analysis and target window selection, the second analysis module 400 performs runoff reduction to obtain the effective runoff, and the early warning module 500 finally triggers the early warning.

[0138] ​The implementation process of the scheme does not require manual intervention, the early warning method is modularized and deployed in remote mountainous basins, all-weather monitoring can be realized, and the timeliness of early warning is greatly improved.

[0139] Further, the device further comprises a third analysis module for dynamically correcting based on the early warning result data after each early warning is completed, to form a sample array (see the foregoing steps S510-S530).

[0140] The modules are connected through wired (such as RS485) or wireless (such as LoRa) communication, adapt to complex communication environments in the wild, ensure real-time transmission of key data, avoid calculation deviation of window length caused by data delay, and ensure real-time calculation of effective runoff.

[0141] Modular design enables the device to be flexibly configured according to actual needs, for example, the rainfall intensity of a plain basin changes gently, the dynamic correction function of the second determination module 200 can be simplified, and the equipment cost is reduced; the rainfall burst characteristics of a mountainous small basin are significant, the real-time election window function of the first analysis module 300 can be enhanced, and the response capability is improved. At the same time, when a single module fails (such as damage of the sensor of the first analysis module), only the module needs to be replaced, and the whole equipment does not need to be replaced, thereby reducing the maintenance cost.

[0142] Embodiments of the present application also provide an electronic device, which includes a processor and a memory. The number of processors can be one or more. The memory, as a kind of computer readable storage medium, can be used to store software programs, computer executable programs and modules. The processor executes various functions of the electronic device and data processing by running the software programs, instructions and modules stored in the memory, so as to realize the mountain torrent disaster early warning method of any one of the above-mentioned embodiments of the present application.

[0143] The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; the data storage area can store data created according to the use of the terminal and the like. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some examples, the memory can further include a memory remotely arranged with respect to the processor, which can be connected to the electronic device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0144] Embodiments of the present application also provide a computer readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the mountain torrent disaster early warning method of any one of the embodiments of the present application is realized.

[0145] The computer storage medium of the embodiments of the present application can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples (non-exhaustive list) of the computer readable storage medium include: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus.

[0146] The computer readable signal medium can include a data signal propagating in a baseband or as part of a carrier wave propagating through a transmission medium, in which the computer readable program code is embodied. Such a propagating data signal can take many forms, including but not limited to electro-magnetic, optical, or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate or transport program for use by or in connection with an instruction execution system, apparatus, or device.

[0147] The embodiments of the present application also provide a computer program product, which, when running on a computer, causes the computer to execute the mountain flood disaster early warning method of any one of the above embodiments of the present application.

[0148] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A flash flood early warning method, characterized in that, The method comprises the following steps: determining a theoretical concentration time according to the static properties of a target basin; determining an observation window of a rainfall event based on the theoretical concentration time; performing a sliding window analysis on the rainfall process of the rainfall event by using the observation window to obtain the cumulative rainfall in the observation window; converting the cumulative rainfall into effective peak-forming flow according to runoff reduction analysis; comparing the effective peak-forming flow with a disaster-causing flow obtained through analysis and investigation, and issuing a warning according to the comparison result.

2. The flash flood warning method of claim 1, wherein, The step of converting the cumulative rainfall into effective peak-forming flow according to runoff reduction analysis comprises the following steps: obtaining a plurality of influence factors of the target basin, wherein the influence factors at least include current water level, soil moisture, air temperature and humidity; performing normalization processing on each influence factor to obtain a normalized value of each influence factor; weighting each normalized influence factor according to a pre-determined weight of each influence factor to obtain a conversion coefficient; converting the cumulative rainfall into effective peak-forming flow based on the conversion coefficient, wherein effective peak-forming flow RX = δP × K; δP represents the cumulative rainfall, and K represents the conversion coefficient.

3. The flash flood warning method of claim 2, wherein, The method further comprises the following steps: correcting the conversion coefficient used for runoff reduction analysis according to the comparison result of the effective peak-forming flow triggering the warning and the measured flow; generating a sample array by using the conversion coefficient corrected according to N rainfall events, wherein the sample array is applied to a mountain telemetry warning site in the same or similar basin.

4. The flash flood warning method of claim 3, wherein, The step of correcting the conversion coefficient used for runoff reduction analysis according to the comparison result of the effective peak-forming flow triggering the warning and the measured flow comprises the following steps: recording, after each time a warning is triggered, the influence factors and their data used in the warning, the effective peak-forming flow RX calculated and the actual monitored flood peak flow R_measured; calculating the error of this time: error = |RX - R_measured| / R_measured; if the error exceeds a set error threshold, correcting the conversion coefficient K, and the correction method comprises adjusting the weight of each influence factor and recording the corrected conversion coefficient K as a sample.

5. The flash flood warning method of claim 1, wherein, The step of determining an observation window of a rainfall event based on the theoretical concentration time comprises the following steps: determining a time range containing the theoretical concentration time; taking a plurality of time nodes in the time range, taking the value corresponding to each time node as the window length of an observation window to obtain a plurality of observation windows; performing a sliding window analysis on the early rainfall process of a rainfall event by using a plurality of observation windows to obtain a sequence of early cumulative rainfall; wherein each sequence of early cumulative rainfall is obtained by continuous observation with an observation window; analyzing the correlation between each sequence of early cumulative rainfall and an early water level rising rate to select an observation window corresponding to the sequence of early cumulative rainfall with the largest correlation as a target observation window; wherein the early water level rising rate is obtained by monitoring the water level of the target basin during the early rainfall process. The target observation window is used for sliding window analysis on a subsequent rainfall process of the rainfall event, to obtain accumulated rainfall in the target observation window.

6. The flash flood warning method of claim 5, wherein, The observation window is used for sliding window analysis on a rainfall process of a rainfall event, and the obtained accumulated rainfall is represented as: wherein, represents the cumulative rainfall at the observation time of the observation window, represents the cumulative rainfall at the observation time of the observation window, represents the cumulative rainfall at the observation time of the observation window, represents the cumulative rainfall at the observation time of the observation window, and X represents the window length of the observation window.

7. The flash flood warning method of claim 1, wherein, The static attributes include terrain slope, vegetation coverage, soil type, and soil moisture content.

8. A flash flood warning device, characterized by, The method comprises the following steps: The first determining module is configured to determine a theoretical concentration time according to static attributes of the target watershed; The second determining module is configured to determine an observation window of a rainfall event based on the theoretical concentration time; The first analysis module is configured to use the observation window to perform sliding window analysis on a rainfall process of a rainfall event, to obtain accumulated rainfall in the observation window; The second analysis module is configured to convert the accumulated rainfall into effective peak-forming flow through runoff reduction analysis; The warning module is configured to compare the effective peak-forming flow with a disaster-causing flow obtained through analysis and investigation, and issue a warning according to a comparison result.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the mountain torrent disaster warning method in any one of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the mountain torrent disaster warning method in any one of claims 1-7.

Images