Drainage basin flood early warning method and system based on regional data and storage medium

Abstract

The invention relates to the technical field of flood early warning, and discloses a drainage basin flood early warning method based on regional data, which comprises the following steps: acquiring rainfall duration of a set region, and calculating a riverbed convergence rate of the set region as an auxiliary function to be fused into a preset BP neural network so as to construct a river reach runoff risk early warning model and output a river reach runoff risk level; the sediment inflow mode and the sediment content of a set area are obtained, the potential sediment risk of the river reach is determined, and the water and sediment risk of the river reach is further determined in combination with the rainfall and the sediment content; and taking actually monitored rainfall and forecast rainfall as driving data, obtaining a predicted water level through the distributed hydrological model, and issuing early warning in the set area in combination with the river reach water and sediment risk. According to the method, the river reach runoff risk early warning model is constructed, the sediment risk level identification mode is determined in combination with the sediment inflow mode, the sediment content and the actual rainfall, the water and sediment disaster risk assessment early warning method considering the sediment effect is provided, and the effectiveness and reliability of flood disaster prevention and control are improved.

Description

Technical Field

[0001] The present invention relates to the field of flood warning technology, and in particular to a basin flood warning method, system and storage medium based on regional data. Background Art

[0002] As flood disasters become more frequent and severe worldwide, research on various flood warning methods is constantly advancing. The main purpose of establishing a river flood warning system is to provide timely and effective early warning of flood disasters, improve the flood prevention level of the entire society, and ensure the flood prevention safety of surrounding residential areas.

[0003] Traditional river flood warning methods are typically based on meteorological parameters, weather reports, flood characteristics, hydrological conditions, and river runoff, and are widely used in hydrological forecasting. These methods typically utilize centralized and distributed hydrological prediction models. These models predict rainfall, river flow, and underlying surface permeability to simulate river seepage patterns, thereby inferring the timing and duration of floods and providing early warnings.

[0004] However, traditional early warning methods ignore the role of sediment, resulting in a high rate of underreporting in existing systems. Influenced by heavy rains, floods are often accompanied by geological disasters such as landslides, bank collapses, and debris flows. Large amounts of sediment instantly enter river channels, causing localized siltation and adjustment of gully beds, sudden increases in water levels, and even causing river diversions, amplifying the catastrophic effects of floods and posing a significant threat to the lives and property of surrounding residents. Sediment sources in small mountainous watersheds exhibit diverse characteristics. Different sediment production methods directly affect the characteristics, timing, and amount of sediment entering river channels, and have different impacts on flood evolution, significantly altering the risk level, affected area, and disaster threshold of water and sand disasters.

[0005] Therefore, traditional early warning systems struggle to accurately predict and capture changing trends. Predicted water level fluctuations often differ from actual values, leading to a decrease in warning accuracy. This inability to accurately predict and warn of river flooding makes it difficult to accurately predict and warn of flood severity, and makes it prone to errors in estimating flood levels, which in turn impacts flood control decisions. Furthermore, ignoring sediment replenishment and its coupling with flooding leads to an underestimation of disaster risks in certain river sections and regions. Summary of the Invention

[0006] The present invention provides a basin flood warning method, system and storage medium based on regional data. It uses the riverbed runoff rate to construct a river section runoff risk warning model, and combines the sediment inflow mode, sediment content and actual rainfall to determine the sediment risk level. It proposes a water and sediment disaster risk assessment and warning method that takes into account the effect of sediment, thereby improving the effectiveness and reliability of flood disaster prevention and control.

[0007] The present invention provides a basin flood early warning method based on regional data, comprising:

[0008] Get the rainfall duration t in the set area p , and according to the rainfall duration t p Calculate the riverbed runoff rate of a given area using the runoff curve method;

[0009] Integrating the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and using the river section runoff risk warning model to output a river section runoff risk level;

[0010] Obtain the sediment inflow pattern and sediment content of a specified area, and determine the potential sediment risk of a river section based on the sediment inflow pattern and sediment content;

[0011] Determine the water and sediment risk of a river section based on the rainfall, sediment content, runoff risk level of the river section and the potential sediment risk of the river section in the set area;

[0012] Using actual monitored rainfall and forecast rainfall as driving data, the predicted water level is obtained through a distributed hydrological model, and an early warning is issued in the set area based on the predicted water level and water and sediment risks in the river section.

[0013] Furthermore, the rainfall duration t in the set area is obtained. p , and according to the rainfall duration t p The steps for calculating the riverbed runoff rate of a given area using the runoff curve method include:

[0014] The runoff curve method is used to calculate the surface runoff during the flood process. The rainfall duration is t p Delay time of flood peak d The relationship expression is:

[0015] t p =5.5t d

[0016] The relationship between the peak flow and delay of surface runoff generated by precipitation in a fixed area of ​​a catchment area is expressed as follows:

[0017]

[0018] Among them, A is the unit basin catchment area, Q m is the peak flow value under standard conditions, C p is the peak flood coefficient, which is 0.85; ΔQ is the difference between the initial and final outflows, A m is the cross-sectional area of ​​the river;

[0019] According to the runoff curve method, the relationship between the peak value of the river flood peak and the peak time is obtained. The specific relationship expression is:

[0020]

[0021] Where V is the volume of the river, t z The peak time is expressed by the precipitation duration and flood peak delay in a fixed area. The specific expression is:

[0022]

[0023] The riverbed confluence rate R p The calculation formula is:

[0024]

[0025] Here, i represents the rainfall period, and i=1 represents the first time point in the rainfall period.

[0026] Furthermore, the step of integrating the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and outputting the river section runoff risk level using the river section runoff risk warning model includes:

[0027] Assume that the time series input vector of river flood is x, and satisfies x∈R d , R d Represents the d parameter set of riverbed confluence rate, then the BP neural network output layer is activated by Gaussian function, and the riverbed confluence rate is used as an auxiliary function. The specific formula is:

[0028] R i (x) = exp(a 2 ||xc i ·R p ||

[0029] Among them, R i (x) is the output value when the number of neurons is i, c i is the stability coefficient of the Gaussian function, a is the number of activations; the weighted summation of the BP neural network output layer is performed, and the specific formula is:

[0030]

[0031] Among them, h is the actual number of hidden nodes, w ij is the parameter weight from the hidden layer to the input layer, y j is the input value when the number of input nodes is j, and q is the initial number of hidden nodes;

[0032] The Hirt function is used to reduce the error. The specific formula is:

[0033] y min =σE(y j )x

[0034] Where σ is the variance of the weight vector, y min is the minimum value of the weighted sum error, E(y j ) represents the relative error function; construct the limit vector error model output o j , the specific formula is as follows:

[0035]

[0036] Among them, the value range of j is [1, m], and the subscripts j and y j The value range of the subscript j is consistent, g(x) is the activation function, b i is the deviation value when the number of neurons is i;

[0037] Taking the deviation value as the objective function, a river runoff risk warning model can be constructed. The specific formula is as follows:

[0038]

[0039] Among them, Q is the flood flow, P r is the rainfall in the river basin, Δt c Represents the interval between different time periods. The value range of x and y is x∈[0,1], y∈[1,0]. f is a nonlinear activation mapping function;

[0040] A river section warning value is calculated according to the river section runoff risk warning model, and the river section warning value is compared with a preset threshold to output a river section runoff risk level; wherein the river section runoff risk level includes level one risk, level two risk, level three risk, and level four risk.

[0041] Furthermore, the step of obtaining the sediment inflow pattern and sediment content of the set area and determining the potential sediment risk of the river section based on the sediment inflow pattern and sediment content includes:

[0042] When there are traces of landslides, bank collapses, or debris flows on the river bank, it is determined that sediment has entered the river section in the form of landslides, collapses, or debris flows, and the area within the set range upstream and downstream of the traces is defined as a potential high-risk sediment section;

[0043] When the sediment concentration reaches the set threshold, it causes riverbed siltation. It is determined that the sediment enters the river section in the form of suspended load and bed load from upstream. The river section with siltation characteristics is defined as a potential medium-risk sediment section.

[0044] After the potential high-risk sediment river sections and potential medium-risk sediment river sections are determined, the remaining river sections are determined as potential low-risk sediment river sections.

[0045] Furthermore, the step of determining the water and sediment risk of a river section based on the rainfall, sediment content, runoff risk level of the river section and potential sediment risk of the river section in the set area includes:

[0046] For river sections with potential high sediment risk, if the rainfall reaches the critical value for triggering debris flow and landslide disasters, the river section water and sediment risk level is determined to be level 4 risk; if the rainfall does not reach the critical value for triggering debris flow and landslide disasters, the river section water and sediment risk level is determined to be equal to the river section runoff risk level;

[0047] For a river section with potential medium sediment risk, if the sediment concentration reaches the preset critical value, the river section water and sediment risk level is determined to be one level higher than the river section runoff risk level; if the sediment concentration does not reach the preset critical value, the river section water and sediment risk level is determined to be equal to the river section runoff risk level;

[0048] For river sections with potential low sediment risk, the water and sediment risk level of the river section is determined to be consistent with the runoff risk level of the river section.

[0049] Furthermore, the step of using actual monitored rainfall and forecast rainfall as driving data, obtaining predicted water levels through a distributed hydrological model, and issuing warnings in the set area based on the predicted water levels and river section water and sediment risks includes:

[0050] For river sections with potential high sediment risk, when the water and sediment risk level of the river section is level 4, a danger warning will be issued to residents nearby and they will be required to evacuate immediately; when the water and sediment risk level of the river section is level 1, 2, or 3, a warning will be issued to residents nearby and they will be required to prepare for evacuation;

[0051] Using actual monitored rainfall and forecast rainfall as driving data, simulated water levels and predicted water levels are obtained through distributed hydrological models;

[0052] For river sections with potential medium sediment risk, the predicted water level is adjusted based on the difference between the simulated water level obtained by actual rainfall and the measured water level as the target predicted water level;

[0053] When the water and sediment risk level of the river section rises by one level, if the target predicted water level is greater than or equal to the disaster level, a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the target predicted water level is less than the disaster level and greater than or equal to 0.8 times the disaster level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the target predicted water level is less than 0.8 times the disaster level, no warning will be issued;

[0054] For river sections with potential low risk of sediment, if the predicted water level is greater than or equal to 1.2 times the disaster water level, a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the predicted water level is less than 1.2 times the disaster water level and greater than or equal to the disaster water level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the predicted water level is less than the disaster water level, no warning will be issued.

[0055] Furthermore, the step of adjusting the predicted water level as the target predicted water level based on the difference between the simulated water level obtained by actual rainfall and the measured water level includes:

[0056] Input the actual rainfall and run the distributed hydrological model to get the simulated water level, which is denoted as H sim Input the forecast rainfall and run the distributed hydrological model to obtain the predicted water level, denoted as H pred_raw ;

[0057] Compare the simulated water level H sim and the measured water level H obs , we get the error sequence ΔH=H obs -H sim ;

[0058] According to the error sequence, the predicted water level is adjusted to the target predicted water level. The target predicted water level H pred_adj =H pred_raw +μ(ΔH); where μ(ΔH) is the historical error mean.

[0059] The present invention also provides a basin flood early warning system based on regional data, comprising:

[0060] Acquisition module, used to obtain the rainfall duration t in the set area p , and according to the rainfall duration t p Calculate the riverbed runoff rate of a given area using the runoff curve method;

[0061] A construction module is used to integrate the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and output a river section runoff risk level using the river section runoff risk warning model;

[0062] The first determination module is used to obtain the sediment inflow pattern and sediment content of a set area, and determine the potential sediment risk of the river section based on the sediment inflow pattern and sediment content;

[0063] The second determination module is used to determine the water and sediment risk of a river section based on the rainfall, sediment content, runoff risk level of the river section and the potential sediment risk of the river section in the set area;

[0064] The early warning module is used to use actual monitored rainfall and forecast rainfall as driving data, obtain predicted water levels through a distributed hydrological model, and issue early warnings in the set area based on the predicted water levels and water and sediment risks in the river section.

[0065] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0066] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0067] The beneficial effects of the present invention are:

[0068] This paper uses riverbed runoff rate combined with a BP neural network structure to construct a river section runoff risk warning model, which can accurately estimate the river section runoff risk level and improve the accuracy of warnings. At the same time, combining sediment inflow patterns, sediment content, and actual rainfall, it analyzes the impact of different sediment production methods on flood evolution characteristics and risk levels. Combined with the BP neural network-based river section runoff risk level assessment, it introduces a method for rapidly identifying sediment risk levels, and then proposes a water and sediment disaster risk assessment and warning method that takes sediment effects into account. This considers the role of sediment and improves the effectiveness and reliability of flood disaster prevention and control, providing technical support for flood prevention and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 Schematic diagram of a method flow according to an embodiment of the present invention.

[0070] Figure 2 FIG. 1 is a schematic diagram of the device structure according to an embodiment of the present invention.

[0071] Figure 3 Schematic diagram of the internal structure of a computer device according to an embodiment of the present invention.

[0072] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0073] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0074] like Figure 1 As shown, the present invention provides a basin flood early warning method based on regional data, comprising:

[0075] S1. Get the rainfall duration t in the set area p , and according to the rainfall duration t pThe runoff curve method is used to calculate the riverbed flow rate of a set area.

[0076] Riverbed confluence refers to the concentration process of the river inside the riverbed. The riverbed confluence rate is the amount of water passing through a certain section of the river per unit time, which is affected by factors such as rainfall, landform, vegetation cover, and soil type. In hydrology, riverbed confluence rate is one of the important indicators for evaluating water resource utilization, flood prediction, and hydrological model establishment. By calculating the confluence rate, we can understand the surface and underground runoff conditions of the river and provide a data basis for flood warning. The surface runoff during the flood process is calculated using the runoff curve method (SCS, Soil Conservation Service), and the rainfall duration t p Delay time with flood peak d The relationship expression is:

[0077] t p =5.5t d

[0078] In general, the relationship between the peak flow and the delay of surface runoff generated by precipitation in a fixed area of ​​a catchment area is expressed as:

[0079]

[0080] Among them, A is the unit basin catchment area, Q m is the peak flow value under standard conditions, C p is the flood peak coefficient, which is 0.85; ΔQ is the difference between the initial and final outflows, A m is the cross-sectional area of ​​the river;

[0081] According to the runoff curve method, the relationship between the peak value of the river flood peak and the peak time is obtained. The specific relationship expression is:

[0082]

[0083] Where V is the volume of the river, t z The peak time is expressed by the precipitation duration and flood peak delay in a fixed area. The specific expression is:

[0084]

[0085] The riverbed confluence rate R p The calculation formula is:

[0086]

[0087] Here, i represents the rainfall period, and i=1 represents the first time point in the rainfall period.

[0088] According to the above formula, the riverbed runoff rate of the river is obtained, and the runoff rate is used as an auxiliary function to construct a neural network hydrological model.

[0089] S2. Integrate the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and use the river section runoff risk warning model to output the river section runoff risk level.

[0090] Data-driven models use data to construct corresponding models and, through machine learning, continuously reduce the error between actual and predicted values. Therefore, they offer significant advantages for solving river flood forecasting problems. The BP neural network model, as one type of data-driven model, is widely used in flood warning. After using data-driven models to predict river floods, the BP neural network can be used to construct a hydrological warning model.

[0091] Assume that the time series input vector of river flood is x, and satisfies x∈R d , R d Represents the d parameter set of riverbed confluence rate, then the BP neural network output layer is activated by Gaussian function, and the riverbed confluence rate is used as an auxiliary function. The specific formula is:

[0092] R i (c) = exp(a 2 ||xc i ·R p ||)

[0093] Among them, R i (x) is the output value when the number of neurons is i, c i is the stability coefficient of the Gaussian function, a is the number of activations; the weighted summation of the BP neural network output layer is performed, and the specific formula is:

[0094]

[0095] Among them, h is the actual number of hidden nodes, w ij is the parameter weight from the hidden layer to the input layer, y j is the input value when the number of input nodes is j, and q is the initial number of hidden nodes;

[0096] The Hirt function is used to reduce the error. The specific formula is:

[0097] y min =σE(y j )x

[0098] Where σ is the variance of the weight vector, y min is the minimum value of the weighted sum error, E(y j) represents the relative error function; construct the limit vector error model output o j , the specific formula is as follows:

[0099]

[0100] Among them, the value range of j is [1, m], and the subscripts j and y j The value range of the subscript j is consistent, g(x) is the activation function, b i is the deviation value when the number of neurons is i;

[0101] Taking the deviation value as the objective function, a river runoff risk warning model can be constructed. The specific formula is as follows:

[0102]

[0103] Among them, Q is the flood flow, P r is the rainfall in the river basin, Δt c Represents the interval between different time periods. The value range of x and y is x∈[0,1], y∈[1,0]. f is a nonlinear activation mapping function;

[0104] A river section warning value is calculated based on the river section runoff risk warning model, and the river section warning value is compared with a preset threshold value. The preset threshold value is divided into four level intervals, each level interval corresponds to a risk level, so as to output the river section runoff risk level; wherein, the river section runoff risk level includes level one risk, level two risk, level three risk, and level four risk.

[0105] S3. Obtain the sediment inflow pattern and sediment content of the set area, and determine the potential sediment risk of the river section based on the sediment inflow pattern and sediment content.

[0106] Different ways of sediment inflow have different effects on water level changes. When sediment enters the river channel instantly and in large quantities in the form of debris flows, landslides, etc., it is easy to cause siltation of the ditch bed near the inflow point and a sudden increase in water level; when sediment enters from the upstream of the target river section in the form of suspended load and bed load, the sediment content reaches a certain threshold and causes the riverbed to rise and the water level to rise at the characteristic section (such as "from narrow to wide" and "from steep to gentle"). In view of this characteristic, without calculating the sediment production in the basin, a rapid river section potential sediment risk assessment method is established, and then combined with the river section runoff risk assessment results in step S2, a comprehensive risk assessment result considering the impact of sediment is obtained and an early warning is issued.

[0107] The potential sediment risk of a river section should be determined in advance before a disaster occurs, with the main indicators being the sediment inflow pattern and sediment content. Based on the analysis of the characteristics of the impact of sediment on flash floods, the specific indicators include:

[0108] S301. When there are traces of landslide, bank collapse, or debris flow on the river bank, it is determined that sediment enters the river section in the form of landslide, collapse, or debris flow, and the area within a set range upstream and downstream of the trace is defined as a potential high-risk sediment section.

[0109] When sediment enters a river channel through landslides, collapses, and debris flows, it often causes localized siltation in the gully bed and a sudden increase in water levels. River sections with such risks are defined as potentially high-risk for sedimentation. Landslides, collapses, and debris flows often exhibit recurring characteristics. Therefore, when evidence of landslides, bank collapses, or debris flows is observed along a river bank, the river within a certain range (e.g., 500 meters) upstream and downstream of the evidence is considered high-risk for sedimentation.

[0110] S302. When the sediment concentration reaches the set threshold, it causes riverbed siltation. It is determined that the sediment comes from upstream and enters the river section in the form of suspended sediment and bed load. The river section with siltation characteristics is defined as a potential medium-risk sediment river section.

[0111] Sediment can also enter target river sections via upstream sediment transport as suspended load and bedload. When sediment concentration reaches a certain threshold, bedload movement is intensified, leading to riverbed siltation in sections where the river changes from narrow to wide or from steep to gentle. Such river sections are defined as potentially medium-risk sections. Hydrodynamic simulations can be performed on the entire river section. Under the same flow boundary conditions, different sediment concentrations can be designed as sediment boundary conditions to simulate the erosion and deposition of the riverbed under the influence of sediment. Sections exhibiting significant siltation characteristics are considered potentially medium-risk sections.

[0112] S303. After the potential high-risk sediment river sections and the potential medium-risk sediment river sections are determined, the remaining river sections are determined as potential low-risk sediment river sections.

[0113] A river section with a potential low sediment risk is one where the water level will not rise significantly due to an increase in sediment concentration. This section is characterized by a straight river channel and riverbed erosion due to the movement of suspended and bedload. After identifying high- and medium-risk river sections, the remaining sections are considered to have a potential low sediment risk.

[0114] S4. Determine the water and sediment risk of a river section based on the rainfall, sediment content, runoff risk level of the river section and the potential sediment risk of the river section in the set area.

[0115] (1) For river sections with potential high sediment risk

[0116] If the rainfall reaches the critical value that triggers debris flow and landslide disasters, causing a large amount of sediment to enter the river channel, regardless of the runoff risk level determination result, the river section risk level will be the highest level (Level 4 risk), that is, the water and sediment risk level of the river section will be determined to be Level 4 risk;

[0117] If the rainfall does not reach the critical value for causing debris flow and landslide disasters, the water and sediment risk level of the river section is determined to be equal to the river section runoff risk level, that is, the risk is determined by the river section runoff risk level.

[0118] (2) For river sections with potential medium-risk sediment

[0119] If the sediment concentration reaches a preset critical value, at which the sediment will cause the river section water level to rise, the risk level will increase. The river section's runoff risk level will be increased by one level, meaning the river section's water and sediment risk level is determined to be one level higher than its runoff risk level.

[0120] If the sediment content does not reach the preset critical value, the river section water and sediment risk level is determined to be equal to the river section runoff risk level, that is, the risk is determined by the river section runoff risk level.

[0121] (3) For river sections with potential low sediment risk, sediment does not affect the risk assessment, and the water and sediment risk level of the river section is determined to be consistent with the runoff risk level of the river section.

[0122] The critical rainfall value that triggers debris flows, landslides and other disasters in potential high-risk sediment river sections and the critical sediment content value in potential medium-risk sediment river sections are obtained based on the specific river channel characteristics through simulation and calculation of the water-sediment dynamics model considering the water-sediment coupling effect.

[0123] S5. Using actual monitored rainfall and forecast rainfall as driving data, a distributed hydrological model is used to obtain predicted water levels, and an early warning is issued in the set area based on the predicted water levels and water and sediment risks in the river section.

[0124] (1) For river sections with potential high sediment risk

[0125] When the water and sediment risk level of the river section is level 4, a danger warning will be issued to residents around the river section and they will be evacuated immediately;

[0126] When the water and sediment risk level of the river section is level one, two or three, a warning will be issued to residents living near the river section to prepare for evacuation;

[0127] (2) For river sections with potential medium-risk sediment

[0128] Using actual monitored rainfall and forecast rainfall as driving data, the distributed hydrological model is used to obtain simulated and predicted water levels. The predicted water level is adjusted based on the difference between the simulated water level obtained from actual rainfall and the measured water level as the target predicted water level. Specifically, the following steps are involved:

[0129] 1> Input actual rainfall and run the distributed hydrological model to obtain the simulated water level, which is recorded as H sim Input the forecast rainfall and run the distributed hydrological model to obtain the predicted water level, denoted as H pred_raw;

[0130] 2> Compare the simulated water level H sim and the measured water level H obs , we get the error sequence ΔH=H obs -H sim ;

[0131] 3> According to the error sequence, the predicted water level is adjusted to the target predicted water level. The target predicted water level H pred_adj =H pred_raw +μ(ΔH); where μ(ΔH) is the historical error mean.

[0132] When the water and sediment risk level of the river section rises by one level, if the target predicted water level is greater than or equal to the disaster water level (the preset water level, or the water level obtained based on historical data analysis), a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the target predicted water level is less than the disaster water level and greater than or equal to 0.8 times the disaster water level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the target predicted water level is less than 0.8 times the disaster water level, no warning will be issued;

[0133] When the water and sediment risk level of the river section is equal to the runoff risk level of the river section, an early warning shall be issued in accordance with the early warning rules for potential high-risk sediment river sections.

[0134] (3) For river sections with potential low sediment risk

[0135] If the predicted water level is greater than or equal to 1.2 times the disaster water level, a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the predicted water level is less than 1.2 times the disaster water level and greater than or equal to the disaster water level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the predicted water level is less than the disaster water level, no warning will be issued.

[0136] like Figure 2 As shown, the present invention also provides a basin flood early warning system based on regional data, comprising:

[0137] Acquisition module 1, used to obtain the rainfall duration t in the set area p , and according to the rainfall duration t p Calculate the riverbed runoff rate of a given area using the runoff curve method;

[0138] Construction module 2 is used to integrate the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and output the river section runoff risk level using the river section runoff risk warning model;

[0139] The first determination module 3 is used to obtain the sediment inflow pattern and sediment content of a set area, and determine the potential sediment risk of the river section based on the sediment inflow pattern and sediment content;

[0140] The second determination module 4 is used to determine the water and sediment risk of a river section based on the rainfall, sediment content, runoff risk level of the river section and the potential sediment risk of the river section in the set area;

[0141] The early warning module 5 is used to use the actual monitored rainfall and forecast rainfall as driving data, obtain the predicted water level through the distributed hydrological model, and issue an early warning in the set area based on the predicted water level and the water and sediment risk of the river section.

[0142] In one embodiment, the acquisition module 1 includes:

[0143] The runoff curve method is used to calculate the surface runoff during the flood process. The rainfall duration is t p Delay time with flood peak d The relationship expression is:

[0144] t p =5.5t d

[0145] The relationship between the peak flow and delay of surface runoff generated by precipitation in a fixed area of ​​a catchment area is expressed as follows:

[0146]

[0147] Among them, A is the unit basin catchment area, Q m is the peak flow value under standard conditions, C p is the flood peak coefficient, which is 0.85; ΔQ is the difference between the initial and final outflows, A m is the cross-sectional area of ​​the river;

[0148] According to the runoff curve method, the relationship between the peak value of the river flood peak and the peak time is obtained. The specific relationship expression is:

[0149]

[0150] Where V is the volume of the river, t z The peak time is expressed by the precipitation duration and flood peak delay in a fixed area. The specific expression is:

[0151]

[0152] The riverbed confluence rate R p The calculation formula is:

[0153]

[0154] Here, i represents the rainfall period, and i=1 represents the first time point in the rainfall period.

[0155] In one embodiment, building block 2 includes:

[0156] Assume that the time series input vector of river flood is x, and satisfies x∈R d , R d Represents the d parameter set of riverbed confluence rate, then the BP neural network output layer is activated by Gaussian function, and the riverbed confluence rate is used as an auxiliary function. The specific formula is:

[0157] R i (x) = exp(a 2 ||xc i ·R p ||)

[0158] Among them, R i (x) is the output value when the number of neurons is i, c i is the stability coefficient of the Gaussian function, a is the number of activations; the weighted summation of the BP neural network output layer is performed, and the specific formula is:

[0159]

[0160] Among them, h is the actual number of hidden nodes, w ij is the parameter weight from the hidden layer to the input layer, y j is the input value when the number of input nodes is j, and q is the initial number of hidden nodes;

[0161] The Hirt function is used to reduce the error. The specific formula is:

[0162] y min =σE(y j )x

[0163] Where σ is the variance of the weight vector, y min is the minimum value of the weighted sum error, E(y j ) represents the relative error function; construct the limit vector error model output o j , the specific formula is as follows:

[0164]

[0165] Among them, the value range of j is [1, m], and the subscripts j and y j The value range of the subscript j is consistent, g(x) is the activation function, b i is the deviation value when the number of neurons is i;

[0166] Taking the deviation value as the objective function, a river runoff risk warning model can be constructed. The specific formula is as follows:

[0167]

[0168] Among them, Q is the flood flow, P r is the rainfall in the river basin, Δt c Represents the interval between different time periods. The value range of x and y is x∈[0,1], y∈[1,0]. f is a nonlinear activation mapping function;

[0169] A river section warning value is calculated according to the river section runoff risk warning model, and the river section warning value is compared with a preset threshold to output a river section runoff risk level; wherein the river section runoff risk level includes level one risk, level two risk, level three risk, and level four risk.

[0170] In one embodiment, the first determining module 3 includes:

[0171] The first definition unit is used to determine if there are traces of landslides, bank collapses, or debris flows on the river bank, and to define the river section within a set range upstream and downstream of the traces as a potential high-risk sediment section;

[0172] The second definition unit is used to determine whether the sediment concentration reaches a set threshold, causing riverbed siltation. It determines whether the sediment enters the river section in the form of suspended load and bed load from upstream, and defines the river section with siltation characteristics as a potential medium-risk sediment section.

[0173] The determination unit is used to determine the remaining river sections as potential low-risk sediment river sections after the potential high-risk sediment river sections and the potential medium-risk sediment river sections are determined.

[0174] In one embodiment, the second determining module 4 includes:

[0175] For river sections with potential high sediment risk, if the rainfall reaches the critical value for triggering debris flow and landslide disasters, the river section water and sediment risk level is determined to be level 4 risk; if the rainfall does not reach the critical value for triggering debris flow and landslide disasters, the river section water and sediment risk level is determined to be equal to the river section runoff risk level;

[0176] For a river section with potential medium sediment risk, if the sediment concentration reaches the preset critical value, the river section water and sediment risk level is determined to be one level higher than the river section runoff risk level; if the sediment concentration does not reach the preset critical value, the river section water and sediment risk level is determined to be equal to the river section runoff risk level;

[0177] For river sections with potential low sediment risk, the water and sediment risk level of the river section is determined to be consistent with the runoff risk level of the river section.

[0178] In one embodiment, the early warning module 5 includes:

[0179] For river sections with potential high sediment risk, when the water and sediment risk level of the river section is level 4, a danger warning will be issued to residents nearby and they will be required to evacuate immediately; when the water and sediment risk level of the river section is level 1, 2, or 3, a warning will be issued to residents nearby and they will be required to prepare for evacuation;

[0180] Using actual monitored rainfall and forecast rainfall as driving data, simulated water levels and predicted water levels are obtained through distributed hydrological models;

[0181] For river sections with potential medium sediment risk, the predicted water level is adjusted based on the difference between the simulated water level obtained by actual rainfall and the measured water level as the target predicted water level;

[0182] When the water and sediment risk level of the river section rises by one level, if the target predicted water level is greater than or equal to the disaster level, a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the target predicted water level is less than the disaster level and greater than or equal to 0.8 times the disaster level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the target predicted water level is less than 0.8 times the disaster level, no warning will be issued;

[0183] For river sections with potential low risk of sediment, if the predicted water level is greater than or equal to 1.2 times the disaster water level, a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the predicted water level is less than 1.2 times the disaster water level and greater than or equal to the disaster water level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the predicted water level is less than the disaster water level, no warning will be issued.

[0184] In one embodiment, adjusting the predicted water level to a target predicted water level based on the difference between the simulated water level obtained by actual rainfall and the measured water level includes:

[0185] Input the actual rainfall and run the distributed hydrological model to get the simulated water level, which is denoted as H sim Input the forecast rainfall and run the distributed hydrological model to obtain the predicted water level, denoted as H pred_raw ;

[0186] Compare the simulated water level H sim and the measured water level H obs , we get the error sequence ΔH=H obs -H sim ;

[0187] According to the error sequence, the predicted water level is adjusted to the target predicted water level. The target predicted water level H pred_adj =H pred_raw +μ(ΔH); where μ(ΔH) is the historical error mean.

[0188] The above modules and units are used to execute the corresponding steps in the above basin flood warning method based on regional data. The specific implementation method thereof is described in the above method embodiment and will not be repeated here.

[0189] like Figure 3 As shown, the present invention also provides a computer device, which can be a server, and its internal structure can be as follows Figure 3 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store all data required for the process of the basin flood warning method based on regional data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the basin flood warning method based on regional data is implemented.

[0190] Those skilled in the art will understand that Figure 3 The structure shown in is merely a block diagram of a portion of the structure related to the present application solution and does not constitute a limitation on the computer device to which the present application solution is applied.

[0191] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, any one of the above-mentioned basin flood warning methods based on regional data is implemented.

[0192] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0193] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, apparatus, article, or method comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, apparatus, article, or method. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, apparatus, article, or method comprising the element.

[0194] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A basin flood early warning method based on regional data, characterized in that: include: Get the rainfall duration t in the set area p , and according to the rainfall duration t p Calculate the riverbed runoff rate of a given area using the runoff curve method; Integrating the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and using the river section runoff risk warning model to output a river section runoff risk level; Obtain the sediment inflow pattern and sediment content of a specified area, and determine the potential sediment risk of a river section based on the sediment inflow pattern and sediment content; Determine the water and sediment risk of a river section based on the rainfall, sediment content, runoff risk level of the river section and the potential sediment risk of the river section in the set area; Using actual monitored rainfall and forecast rainfall as driving data, the predicted water level is obtained through a distributed hydrological model, and an early warning is issued in the set area based on the predicted water level and water and sediment risks in the river section.

2. The basin flood early warning method based on regional data according to claim 1, characterized in that: The rainfall duration t of the set area is obtained p , and according to the rainfall duration t p The steps for calculating the riverbed runoff rate of a given area using the runoff curve method include: The runoff curve method is used to calculate the surface runoff during the flood process. The rainfall duration is t p Delay time with flood peak d The relationship expression is: t p =5.5t d The relationship between the peak flow and delay of surface runoff generated by precipitation in a fixed area of ​​a catchment area is expressed as follows: Among them, A is the unit basin catchment area, Q m is the peak flow value under standard conditions, C p is the peak flood coefficient, which is 0.85; ΔQ is the difference between the initial and final outflows, A m is the cross-sectional area of ​​the river; According to the runoff curve method, the relationship between the peak value of the river flood peak and the peak time is obtained. The specific relationship expression is: Where V is the volume of the river, t z The peak time is expressed by the precipitation duration and flood peak delay in a fixed area. The specific expression is: The riverbed confluence rate R p The calculation formula is: Here, i represents the rainfall period, and i=1 represents the first time point in the rainfall period.

3. The basin flood early warning method based on regional data according to claim 2, characterized in that: The steps of integrating the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and outputting a river section runoff risk level using the river section runoff risk warning model include: Assume that the time series input vector of river flood is x, and satisfies x∈R d , R d Represents the d parameter set of riverbed confluence rate, then the BP neural network output layer is activated by Gaussian function, and the riverbed confluence rate is used as an auxiliary function. The specific formula is: R i (x)=exp(a 2 ||x-c i ·R p || Among them, R i (x) is the output value when the number of neurons is i, c i is the stability coefficient of the Gaussian function, a is the number of activations; the weighted summation of the BP neural network output layer is performed, and the specific formula is: Among them, h is the actual number of hidden nodes, w ij is the parameter weight from the hidden layer to the input layer, y j is the input value when the number of input nodes is j, and q is the initial number of hidden nodes; The Hirt function is used to reduce the error. The specific formula is: and min =σE(y j ) Where σ is the variance of the weight vector, y min is the minimum value of the weighted sum error, E(y j ) represents the relative error function; construct the limit vector error model output o j , the specific formula is as follows: Among them, the value range of j is [1, m], and the subscripts j and y j The value range of the subscript j is consistent, g(x) is the activation function, b i is the deviation value when the number of neurons is i; Taking the deviation value as the objective function, a river runoff risk warning model can be constructed. The specific formula is as follows: Among them, Q is the flood flow, P r is the rainfall in the river basin, Δt c Represents the interval between different time periods. The value range of x and y is x∈[0,1], y∈[1,0]. f is a nonlinear activation mapping function; A river section warning value is calculated according to the river section runoff risk warning model, and the river section warning value is compared with a preset threshold to output a river section runoff risk level; wherein the river section runoff risk level includes level one risk, level two risk, level three risk, and level four risk.

4. The basin flood early warning method based on regional data according to claim 3 is characterized in that: The step of obtaining the sediment inflow pattern and sediment content of the set area and determining the potential sediment risk of the river section based on the sediment inflow pattern and sediment content includes: When there are traces of landslides, bank collapses, or debris flows on the river bank, it is determined that sediment has entered the river section in the form of landslides, collapses, or debris flows, and the area within the set range upstream and downstream of the traces is defined as a potential high-risk sediment section; When the sediment concentration reaches the set threshold, it causes riverbed siltation. It is determined that the sediment enters the river section in the form of suspended load and bed load from upstream. The river section with siltation characteristics is defined as a potential medium-risk sediment section. After the potential high-risk sediment river sections and potential medium-risk sediment river sections are determined, the remaining river sections are determined as potential low-risk sediment river sections.

5. The basin flood early warning method based on regional data according to claim 4 is characterized in that: The step of determining the water and sediment risk of a river section based on the rainfall, sediment content, runoff risk level of the river section, and potential sediment risk of the river section in the set area includes: For river sections with potential high sediment risk, if the rainfall reaches the critical value for triggering debris flow and landslide disasters, the river section water and sediment risk level is determined to be level 4 risk; if the rainfall does not reach the critical value for triggering debris flow and landslide disasters, the river section water and sediment risk level is determined to be equal to the river section runoff risk level; For a river section with potential medium sediment risk, if the sediment concentration reaches the preset critical value, the river section water and sediment risk level is determined to be one level higher than the river section runoff risk level; if the sediment concentration does not reach the preset critical value, the river section water and sediment risk level is determined to be equal to the river section runoff risk level; For river sections with potential low sediment risk, the water and sediment risk level of the river section is determined to be consistent with the runoff risk level of the river section.

6. The basin flood early warning method based on regional data according to claim 5, characterized in that: The steps of using actual monitored rainfall and forecast rainfall as driving data, obtaining predicted water levels through a distributed hydrological model, and issuing warnings in the set area based on the predicted water levels and river section water and sediment risks include: For river sections with potential high sediment risk, when the water and sediment risk level of the river section is level 4, a danger warning will be issued to residents nearby and they will be required to evacuate immediately; when the water and sediment risk level of the river section is level 1, 2, or 3, a warning will be issued to residents nearby and they will be required to prepare for evacuation; Using actual monitored rainfall and forecast rainfall as driving data, simulated water levels and predicted water levels are obtained through distributed hydrological models; For river sections with potential medium sediment risk, the predicted water level is adjusted based on the difference between the simulated water level obtained by actual rainfall and the measured water level as the target predicted water level; When the water and sediment risk level of the river section rises by one level, if the target predicted water level is greater than or equal to the disaster level, a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the target predicted water level is less than the disaster level and greater than or equal to 0.8 times the disaster level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the target predicted water level is less than 0.8 times the disaster level, no warning will be issued; For river sections with potential low risk of sediment, if the predicted water level is greater than or equal to 1.2 times the disaster water level, a danger warning will be issued to residents around the river section and they will be required to evacuate immediately; if the predicted water level is less than 1.2 times the disaster water level and greater than or equal to the disaster water level, a warning will be issued to residents around the river section and they will be required to prepare to evacuate; if the predicted water level is less than the disaster water level, no warning will be issued.

7. The basin flood early warning method based on regional data according to claim 6, characterized in that: The step of adjusting the predicted water level as the target predicted water level based on the difference between the simulated water level obtained by actual rainfall and the measured water level comprises: Input the actual rainfall and run the distributed hydrological model to get the simulated water level, which is denoted as H sim Input the forecast rainfall and run the distributed hydrological model to obtain the predicted water level, denoted as H pred_raw ; Compare the simulated water level H sim and the measured water level H obs , we get the error sequence ΔH=H obs -H sim ; According to the error sequence, the predicted water level is adjusted to the target predicted water level. The target predicted water level H pred_adj =H pred_raw +μ(ΔH); where μ(ΔH) is the historical error mean.

8. A basin flood early warning system based on regional data, characterized in that: include: Acquisition module, used to obtain the rainfall duration t in the set area p , and according to the rainfall duration t p Calculate the riverbed runoff rate of a given area using the runoff curve method; A construction module is used to integrate the riverbed confluence rate as an auxiliary function into a preset BP neural network to construct a river section runoff risk warning model, and output a river section runoff risk level using the river section runoff risk warning model; The first determination module is used to obtain the sediment inflow pattern and sediment content of a set area, and determine the potential sediment risk of the river section based on the sediment inflow pattern and sediment content; The second determination module is used to determine the water and sediment risk of a river section based on the rainfall, sediment content, river section runoff risk level and potential sediment risk of the river section in the set area; The early warning module is used to use actual monitored rainfall and forecast rainfall as driving data, obtain predicted water levels through a distributed hydrological model, and issue early warnings in the set area based on the predicted water levels and water and sediment risks in the river section.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images