A method for constructing a dynamic model for geographical analysis and evaluation of flood disasters in power facilities

Abstract

The present invention discloses a method for constructing a dynamic model for geographical analysis and evaluation of flood disasters of electric power facilities, and the present invention relates to the technical field of risk evaluation. The method comprises the following steps: collecting meteorological data, geographical information data and average water depth of the area where the electric power facilities to be analyzed are located in the past when flood disasters occurred, establishing a flood disaster water depth prediction model, and outputting the flood disaster water depth; correcting the average water depth value by hydrological and topographic characteristic parameters; calculating the impact force of water flow on the electric power facilities and the water flow pressure received according to the accurate value of the average water depth and the real-time flow velocity of the flood, and calculating the bearing coefficient of the electric power facilities in combination with the real-time load data of the electric power facilities; constructing a risk dynamic evaluation mathematical model based on the accurate value of the average water depth and the bearing coefficient of the electric power facilities to obtain a real-time risk score of the electric power facilities; outputting different risk prompts and corresponding processing requirements according to the real-time risk score of the electric power facilities.

Description

Technical Field

[0001] The present invention relates to the technical field of risk assessment, and in particular to a method for constructing a dynamic model for geographical analysis and assessment of flood disasters in power facilities. Background Art

[0002] With global climate change and the increasing frequency of extreme weather events, the threat of floods to power infrastructure has become increasingly prominent. When power facilities encounter floods, they may not only cause equipment damage and power outages, but may also cause power supply interruptions on a larger scale, leading to significant social and economic losses. Traditional flood disaster assessment methods rely on historical data and experience for assessment and prediction, lacking the ability to dynamically respond to real-time changing factors, resulting in low efficiency in emergency management and post-disaster recovery. In addition, existing assessment models often fail to fully consider the impact of multiple factors such as geographical characteristics, soil properties, and real-time hydrological parameters, resulting in insufficient prediction accuracy for flood disasters.

[0003] At present, there are still some important technical problems in the risk assessment of flood disasters for power facilities. First, there is a lack of integrated flood disaster models, and it is impossible to comprehensively consider multi-dimensional data such as historical meteorology, topography and real-time hydrology for dynamic analysis. Secondly, traditional water depth prediction models are usually based on static historical data, and fail to effectively reflect the real-time impact of current meteorological conditions and environmental changes on flood disasters. In addition, when analyzing the impact force and water flow pressure of water flow on power facilities, existing models usually ignore the complex effects of factors such as terrain undulation and soil type, resulting in inaccurate carrying capacity assessment and unable to provide reliable support for decision-making. Therefore, developing a dynamic assessment model based on multi-source data integration that can monitor and evaluate the impact of flood disasters on power facilities in real time will be the key to improving the safety of power facilities and the effectiveness of emergency management.

[0004] In the prior art, the publication number CN103927389B discloses a method for constructing a dynamic model for geographical analysis and assessment of flood disasters, including: collecting initial data of the area to be analyzed, compressing and storing it in a created lightweight spatial database; retrieving the initial data, initializing the constructed flood inundation analysis model, and using the seed point spread algorithm to obtain the final flood inundation area range map and water depth classification map; inputting the integrated final flood inundation area range map and water depth classification map data into the flood loss assessment mathematical model and the flood risk assessment mathematical model for calculation, and obtaining the economic losses and potential disaster risk of each village in the flood storage area after the flood inundation. However, when collecting the initial data of the area to be analyzed, if this scheme fails to fully cover all relevant factors (such as meteorological data, topography, soil type, etc.), the output results of the model may be inaccurate. In addition, in practical applications, there is a lack of real-time monitoring and dynamic update mechanisms. Flood disasters are often sudden, and if the model or data cannot be updated in time, the evaluation results may lag. Therefore, general predictions made through this model may waste resources and reduce the real-time and effectiveness of the supervision model.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a method for constructing a dynamic model for geographical analysis and evaluation of flood disasters in power facilities, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for constructing a dynamic model for geographical analysis and assessment of flood disasters in power facilities, the specific steps comprising:

[0009] Collect meteorological data, geographic information data and average water depth when flood disasters occurred in the area where the power facilities to be analyzed are located in the past, establish a flood disaster water depth prediction model, use the collected historical meteorological data and geographic information data as the input of the prediction model, and use the corresponding average water depth as a label to train the prediction model;

[0010] Input the current meteorological data and geographic information data of the area where the power facility to be analyzed is located into the prediction model to obtain the average water depth corresponding to the current flood disaster, and correct the average water depth value output by the flood disaster water depth prediction model based on the hydrological and topographic characteristic parameters of the area where the power facility to be analyzed is located to obtain the accurate value of the average water depth, wherein the hydrological and topographic characteristic parameters include: terrain undulation, land use, soil texture and elevation;

[0011] The impact force of water flow on power facilities and the water flow pressure on the submerged parts of power facilities are calculated based on the accurate average water depth and the real-time flow velocity of flood disaster water flow. Based on the obtained impact force and water flow pressure data, combined with the real-time load data of power facilities, the load factor of power facilities is calculated;

[0012] Based on the accurate value of average water depth and the load factor of power facilities, a dynamic risk assessment mathematical model is constructed. According to the accurate value of average water depth and the load factor of power facilities, a logical expression for calculating risk scores of the dynamic assessment mathematical model is formulated to obtain real-time risk scores of power facilities.

[0013] The real-time risk score of power facilities is compared with the pre-set risk factor threshold, and different risk warnings and corresponding processing requirements are output based on different comparison results.

[0014] Furthermore, the meteorological data of historical flood disasters collected specifically include the rainfall at various moments during the historical flood disasters, the average water depth is the average water depth at the corresponding moment, and the geographic information data includes terrain index, river network density, slope and vegetation normalization index, wherein the vegetation normalization index is calculated through remote sensing images, and the formula for calculating the vegetation normalization index is:

[0015] ;

[0016] In the formula, is the normalized difference vegetation index, is the average near-infrared reflectivity of the area where the power facilities to be analyzed are located, is the average red band reflectance of the area where the power facilities to be analyzed are located;

[0017] The collected data is enhanced, including interpolation enhancement and time series expansion. The data volume is increased based on data enhancement to generate a sample data set. The meteorological data and geographic information data of historical flood disasters are mapped one by one with the corresponding average water depth of flood disasters and stored in the sample data set.

[0018] Furthermore, a flood disaster water depth prediction model is established based on the LSTM long short-term memory network model, and the activation function and optimization algorithm are selected. The Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model; the formula of the Tanh function is:

[0019] ;

[0020] In the formula, Represents the Tanh function, independent variable Represents the weighted sum of the neuron's input, that is, the result of weighted summation of the input received by the neuron from the previous layer;

[0021] At the same time, the hyperparameters of the LSTM model are set, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing number, and the number of hidden layer neurons;

[0022] The number of network layers is set to a four-layer network structure, the number of iterations is set to 300, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 300, the batch size is set to 256, and the number of hidden layer neurons is 32.

[0023] Furthermore, the average water depth value output by the flood disaster water depth prediction model is corrected by the hydrological and topographic characteristic parameters to obtain the accurate value of the average water depth. The specific calculation formula of the accurate value of the average water depth is:

[0024] ;

[0025] In the formula, is the precise value of the average water depth, is the average water depth value output by the water depth prediction model, represents the correction function of terrain relief to water depth, is the correction function of land use to water depth, is the correction function of soil texture to water depth, is the correction function of elevation to water depth, where is the terrain relief, For land use, Indicates soil texture, For elevation.

[0026] Furthermore, the correction function of the terrain relief to the water depth is The specific calculation is based on the formula:

[0027] ;

[0028] In the formula, is the maximum terrain height in the area, is the influence weight coefficient of terrain relief, Indicates the attenuation coefficient of the depth affected by the terrain relief, which is used to describe the The rate at which the effect of water flow accumulation gradually weakens with the increase of

[0029] Correction function of land use on water depth The specific expression is:

[0030] ;

[0031] In the formula, is the influence coefficient of land use, is the land use type classification factor;

[0032] Soil texture correction function for water depth The specific expression based on is:

[0033] ;

[0034] In the formula, is the influence coefficient of soil texture, is the soil texture index, where soil texture index is calculated The formula is:

[0035] ;

[0036] In the formula, and are the weight coefficients for the percentage of sand, silt and clay in the soil, respectively, , and are the percentages of sand, silt, and clay in the soil, respectively, where and and All are greater than 0;

[0037] Correction function of elevation to water depth The specific formula is:

[0038] ;

[0039] In the formula, is the influence coefficient of elevation, Indicates the maximum elevation of the detection area.

[0040] Furthermore, the impact force of the water flow on the power facilities and the water flow pressure on the submerged parts of the power facilities are calculated based on the accurate value of the average water depth and the real-time flow velocity of the flood disaster water flow. The formulas for calculating the impact force of the water flow on the power facilities and the water flow pressure on the submerged parts of the power facilities are:

[0041] ;

[0042] ;

[0043] In the formula, The impact of water flow on power facilities. is the density of water, is the load-bearing area of ​​the power facility, is the real-time flow rate of water, The water pressure on the submerged part of the power facility. is the gravitational acceleration, where the force-bearing area of ​​the power facility The calculation is based on the formula:

[0044] ;

[0045] In the formula, It is the perimeter of the base of the power facility.

[0046] Furthermore, based on the obtained impact force and water flow pressure data, combined with the real-time load data of the power facility, the formula for calculating the load factor of the power facility is as follows:

[0047] ;

[0048] In the formula, is the load factor of the power facility, The real-time load of the power facility. , and are the weight coefficients of the impact force of water flow on power facilities, the water flow pressure on the submerged part of power facilities, and the real-time load of power facilities, respectively. and , and Both are greater than 0.

[0049] Furthermore, the formula for calculating the real-time power facility risk score is:

[0050] ;

[0051] In the formula, Scoring the risk of power facilities, and is the weight coefficient of the average water depth and the load factor of the power facilities, where and and All are greater than 0;

[0052] According to the real-time power facility risk score, it is compared with the pre-set risk factor threshold. According to different comparison results, different risk warnings and corresponding processing requirements are output. The logic is as follows: is a pre-set risk factor threshold;

[0053] when When the flood disaster is detected, it is judged that the current flood disaster in the area where the power facilities are located is a level one risk and no action is needed at present;

[0054] when When the current flood disaster in the area where the power facilities are located is judged to be a secondary risk, operations to reduce the workload of the power facilities are carried out, and the development of the flood disaster is continuously monitored;

[0055] when When the current flood disaster in the area where the power facilities are located is judged to be a level 3 risk, the current power equipment should be stopped.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] By comprehensively analyzing historical meteorological data, geographic information, and hydrological characteristics, the accuracy of flood depth prediction has been significantly improved. Using a neural network model, machine learning technology, and finite element analysis methods, it is possible to dynamically evaluate flood risks based on real-time data. By establishing a finite element physical analysis model based on hydrological and topographic characteristic parameters such as terrain undulation, land use, soil texture, and elevation, the model can correct the flood depth prediction results to obtain a more accurate average water depth value. This not only improves the response speed and prediction capabilities for flood events, but also makes power facilities more resilient to flood disasters, reducing potential economic losses and safety risks.

[0058] In addition, based on the calculated impact force and water flow pressure, combined with the real-time load data of the power facilities, the load factor is dynamically calculated. It provides a real-time assessment of the load-bearing capacity of power facilities under flood conditions. Through real-time monitoring and dynamic scoring mechanisms, the model can promptly issue risk alerts to managers and provide targeted response strategies based on preset thresholds. The data-driven decision support system can effectively improve the ability of power facilities to respond to natural disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION

[0060] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0061] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0062] Example:

[0063] See also Figure 1 , the present invention provides a technical solution:

[0064] A method for constructing a dynamic model for geographical analysis and assessment of flood disasters in power facilities, the specific steps comprising:

[0065] Step 1: Collect meteorological data, geographic information data and average water depth when flood disasters occurred in the area where the power facilities to be analyzed are located in the past, establish a flood disaster water depth prediction model, use the collected historical meteorological data and geographic information data as inputs of the prediction model, and use the corresponding average water depth as a label to train the prediction model;

[0066] The meteorological data of historical flood disasters collected specifically include the rainfall at various times during the historical flood disasters. The average water depth is the average water depth at the corresponding time. The geographic information data includes terrain index, river network density, slope and vegetation normalization index, wherein the vegetation normalization index is calculated through remote sensing images, and the formula for calculating the vegetation normalization index is:

[0067] ;

[0068] In the formula, is the normalized difference vegetation index, is the average near-infrared reflectivity of the area where the power facilities to be analyzed are located, is the average red band reflectance of the area where the power facilities to be analyzed are located;

[0069] The collected data is enhanced, including interpolation enhancement and time series expansion. The data volume is increased based on data enhancement to generate a sample data set. The meteorological data and geographic information data of historical flood disasters are mapped one by one with the corresponding average water depth of flood disasters and stored in the sample data set.

[0070] Among them, rainfall is the direct factor causing floods. The average rainfall over many years provides a long-term trend of precipitation in a region, while the maximum 24-hour rainfall reflects the impact of extreme weather events. The greater the rainfall, the more water flows into the surface and rivers, causing the water level to rise, thereby increasing the average water depth of flood disasters.

[0071] Areas with high topographic index may cause water to converge faster, thereby increasing water depth, while areas with low topographic index may drain faster, reducing water depth.

[0072] A higher river network density generally allows for better drainage and reduces the average depth of floods; whereas areas with a low river network density experience poor water drainage, which may lead to increased water depth.

[0073] The steeper the terrain slope, the faster the water flows, which usually reduces the water retention time and thus reduces the average flood depth. However, areas with gentle slopes have slower water flows and are prone to water accumulation, leading to increased water depth.

[0074] The better the vegetation cover (higher NDVI value), the stronger the soil's water retention capacity and the higher the rainfall infiltration rate, which can reduce surface runoff and thus reduce the average water depth of floods.

[0075] A flood disaster water depth prediction model is established based on the LSTM long short-term memory network model, and the activation function and optimization algorithm are selected. The Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model; the formula of the Tanh function is:

[0076] ;

[0077] In the formula, Represents the Tanh function, independent variable Represents the weighted sum of the neuron's input, that is, the result of weighted summation of the input received by the neuron from the previous layer;

[0078] At the same time, the hyperparameters of the LSTM model are set, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing number, and the number of hidden layer neurons;

[0079] The number of network layers is set to a four-layer network structure, the number of iterations is set to 300, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 300, the batch size is set to 256, and the number of hidden layer neurons is 32.

[0080] LSTM can effectively process time series data, especially data with long-term dependencies. The water depth prediction of flood disasters usually depends on the time series of variables such as historical rainfall and meteorological conditions. LSTM can capture such temporal dynamics. The structure of LSTM allows it to maintain information over multiple time steps. Compared with traditional RNN (recurrent neural network), it can better learn long-term dependencies. LSTM can also learn and update in real time by continuously inputting new data. It is suitable for dealing with dynamically changing flood conditions and can provide timely forecast information for emergency management.

[0081] Step 2: Input the current meteorological data and geographic information data of the area where the power facilities to be analyzed are located into the prediction model to obtain the average water depth corresponding to the current flood disaster. Based on the hydrological and topographic characteristic parameters of the area where the power facilities to be analyzed are located, correct the average water depth value output by the flood disaster water depth prediction model to obtain the accurate value of the average water depth. The hydrological and topographic characteristic parameters include: terrain undulation, land use, soil texture and elevation.

[0082] The average water depth value output by the flood disaster water depth prediction model is corrected by the hydrological and topographic characteristic parameters to obtain the accurate value of the average water depth. The specific calculation formula of the accurate value of the average water depth is:

[0083] ;

[0084] In the formula, is the precise value of the average water depth, is the average water depth value output by the water depth prediction model, represents the correction function of terrain relief to water depth, is the correction function of land use to water depth, is the correction function of soil texture to water depth, is the correction function of elevation to water depth, where is the terrain relief, For land use, Indicates soil texture, For elevation.

[0085] For the four correction functions, their designs are all consistent with the average water depth accuracy. The larger the correction function value, the more accurate the average water depth. The larger the number, the deeper the water in the flood disaster.

[0086] The correction function of the terrain relief to the water depth is The specific calculation is based on the formula:

[0087] ;

[0088] In the formula, Indicates the undulation of the local terrain, usually referring to the height difference of the terrain or the height change of the local terrain. Water flow may be more affected in places with large terrain undulations. In areas with large terrain undulations, water flow will encounter more obstacles, which may cause the water flow speed to slow down or gather. This effect will gradually increase as the undulation increases, so the function contains a The term is proportional to the terrain; it indicates that the greater the terrain undulation, the deeper the water depth may be. The correction function The larger the value.

[0089] For very strong terrain relief, that is, The accumulation effect of water flow will reach a saturation point and no longer The increase of the depth continues to increase, so the attenuation coefficient that affects the depth of the terrain relief is introduced. As the terrain becomes more rugged, the impact will gradually slow down and eventually stabilize.

[0090] Indicates the maximum terrain height in the entire area. It is used to measure the terrain undulation Normalization is performed to ensure that the model has scale consistency.

[0091] Indicates the attenuation coefficient of the depth affected by the terrain relief, which is used to describe the The rate at which the influence of water flow accumulation gradually weakens with the increase of The value of directly affects the velocity and distribution characteristics of water flow under different terrain conditions. A larger value will increase the sensitivity of the water flow to terrain changes, forming a more complex flow pattern; while a smaller Lower values ​​may cause water to flow along a more singular path, with less accumulation.

[0092] The specific method for determining is as follows: In the relevant hydrology and hydrological model literature, some studies have given the The recommended range may be obtained through historical data analysis, field observation or computational fluid dynamics simulation. The specific value range of may vary depending on the region, terrain features and environmental conditions, but is usually between 0 and 1.

[0093] The influence weight coefficient of the terrain relief , The value of is usually related to the type of terrain and hydrological conditions. Different types of terrain (such as mountains, hills, plains, etc.) have different effects on water flow, so For example, the terrain in mountainous areas is more undulating, which may lead to more complex water flow distribution and require higher To reflect the strong influence of terrain on water flow. In flat areas, the terrain changes little and the influence is weaker. The value may be lower.

[0094] For flat areas, The value of can be small, typically between 0.01 and 0.1.

[0095] For hilly areas, The value of may be larger, perhaps between 0.1 and 1.

[0096] For mountainous areas or areas with large terrain undulations, The value of may be large, perhaps between 1 and 5.

[0097] Correction function of land use on water depth The specific expression is:

[0098] ;

[0099] In the formula, is the influence coefficient of land use, is the land use type classification factor.

[0100] Since the larger the land use type classification factor, the weaker its permeability and regulation capacity, resulting in a greater depth of waterlogging, the correction function It is proportional to the land use type classification factor. The larger the land use type classification factor, the deeper the water depth may be. The correction function The larger the value.

[0101] Land use type classification factor It is set according to the land use type, generally: the land use type classification factor of farmland is 0.5, the land use type classification factor of city is 2.0, the land use type classification factor of wetland is 0.3, etc.; its size determines the overall impact intensity of land use on water flow or water depth.

[0102] Wetlands have a lower classification factor because they usually have strong water storage and regulation capabilities and have a greater mitigation effect on water flow. Farmlands usually have higher permeability and therefore have less obstruction to water flow. Although farmlands have higher permeability, their hydrological regulation effect is relatively limited, especially in the face of extreme hydrological events. The classification factor is usually set to a larger value (such as 0.5) because they rely on the regulation of soil and irrigation systems and do not have the natural regulation ability of wetlands. Cities have a higher classification factor because urbanization is usually accompanied by a large number of impermeable surfaces (such as roads and buildings), which leads to water accumulation and intensified runoff. For other land types, the factor setting range is generally within 0-3, and is set according to the ground permeability and the obstruction effect on water flow.

[0103] The influence coefficient of land use ,like A larger value indicates that the land use has a stronger impact on water flow; conversely, a smaller value indicates a weaker impact. For farmland, the coefficient is generally 0.5 to 1.0; for cities, the coefficient is generally 1.0-1.5; and for wetlands, the coefficient is generally 0.3-0.8. The value is between 0 and 2.

[0104] Soil texture correction function for water depth The specific expression based on is:

[0105] ;

[0106] In the formula, is the influence coefficient of soil texture, is the soil texture index, where soil texture index is calculated The formula is:

[0107] ;

[0108] In the formula, and are the weight coefficients for the percentage of sand, silt and clay in the soil, respectively, , and are the percentages of sand, silt, and clay in the soil, respectively.

[0109] The texture of the soil directly affects hydrological characteristics such as water permeability, retention capacity, and flow rate. Different soil compositions (sand, silt, clay) can significantly change the characteristics of water flow, which in turn affects water depth. Generally, the more permeable the soil (the easier it is for water to flow through), the less water the soil can hold, and the less variation in water depth; whereas areas with less permeable soil (where water cannot easily flow through) may result in water retention and greater variation in water depth.

[0110] Therefore, the soil has a higher proportion of sand, is more permeable, and may have less variation in water depth, indicating a soil texture index of The larger the value, the higher the proportion of clay in the soil, the poorer the permeability, the more difficult it is for water to penetrate, and the more likely it is for water to accumulate, resulting in a greater water depth. That is, it is inversely proportional to the water depth, the soil texture index The larger the water depth, the smaller the value, so the correction function of soil texture to water depth is The smaller the value.

[0111] Since the greater the sand content in the soil, the stronger the permeability, the water can easily penetrate into the deep soil layer, and the water depth does not change much. Silt is slightly more permeable than clay, so it is set and and All are greater than 0, where the percentage of sand, silt and clay in the soil can be obtained by conducting a soil content analysis test;

[0112] The influence coefficient of soil texture It can be determined in a variety of ways, such as through experimental data, literature references, model calibration or theoretical derivation.

[0113] Since sand has high water permeability, water penetrates faster in sand. Take a lower value. The value range is generally 0.1 to 0.5. The permeability and water holding capacity of silt are moderate, and its influence coefficient It will be slightly higher than sand. The general value range is 0.5 to 1.0. Since clay has poor permeability, water tends to stay on the surface, resulting in large changes in water depth. Usually a higher value is taken. The value range is approximately 1.0 to 2.0.

[0114] The general value range is 0.1 to 2.0, and the specific value depends on the soil type (such as sand, clay, silt, etc.) and its permeability and hydrological characteristics.

[0115] Correction function of elevation to water depth The specific formula is:

[0116] ;

[0117] In the formula, is the influence coefficient of elevation, Indicates the maximum elevation of the detection area.

[0118] use This ratio is used to normalize the elevation data and compress it into This normalization operation prevents the influence of elevation on water depth correction from being too large or too small, and also avoids the correction coefficient being too extreme due to the large differences in terrain in different regions. Water depth correction function There is a linear proportional relationship between and elevation. That is, when the elevation increases, the correction function The value of will also increase. This means that at higher elevations, the water is deeper and the depth correction will be greater. Conversely, when the elevation is lower, the correction is smaller. By putting the ratio of the current elevation to the maximum elevation into the formula, you are actually adjusting the depth of the water based on the ratio of the elevation of that point to the maximum elevation of the area. Therefore, at elevations close to When the water depth correction function will approach the maximum value, and when the elevation approaches zero, the depth correction function Is smaller.

[0119] Among them, the maximum terrain height in the area and the maximum elevation of the detection area Obtained through a built finite element geographic model or through publicly available terrain quality.

[0120] In the natural environment, elevation has a certain impact on water depth. For example, waters located in higher places may be affected by upstream currents, or the water level and depth may change due to the ups and downs of the terrain. By performing elevation correction on the depth data, the deviation caused by terrain changes on the depth measurement can be eliminated, thereby obtaining more accurate depth information.

[0121] In many existing hydrological studies, corresponding influence coefficients may have been set for different types of terrain and elevation. By consulting relevant literature, it is possible to obtain the influence coefficients applicable to a specific area. For example, some areas may have determined standard coefficients of influence of elevation on water depth through long-term hydrological studies. Flat areas: In areas with small elevation changes (such as plains or low hills), the influence of elevation on water depth is small, so Values ​​are typically low, perhaps between 0.1 and 0.5.

[0122] Medium slope areas: In areas with a certain slope (such as hilly areas), the effect of elevation on water depth may be more significant due to the large terrain undulations. Possible values ​​are between 0.5 and 1.5.

[0123] Steep areas: In high mountains or steep slopes, water flow is greatly affected by elevation due to the dramatic changes in terrain. The value of may be higher, perhaps between 1.0 and 3.0.

[0124] Step 3: Calculate the impact force of water flow on power facilities and the water flow pressure on the submerged parts of power facilities based on the precise average water depth and the real-time flow velocity of flood disaster water flow. Calculate the bearing coefficient of power facilities based on the obtained impact force and water flow pressure data and the real-time load data of power facilities.

[0125] The impact force of water flow on power facilities and the water flow pressure on the submerged parts of power facilities are calculated based on the accurate value of average water depth and the real-time flow velocity of flood disaster water flow. The formulas for calculating the impact force of water flow on power facilities and the water flow pressure on the submerged parts of power facilities are:

[0126] ;

[0127] ;

[0128] In the formula, The impact of water flow on power facilities. is the density of water, is the load-bearing area of ​​the power facility, is the real-time flow rate of water, The water pressure on the submerged part of the power facility. is the gravitational acceleration, where the force-bearing area of ​​the power facility The calculation is based on the formula:

[0129] ;

[0130] In the formula, It is the perimeter of the base of the power facility.

[0131] Based on the obtained impact force and water flow pressure data, combined with the real-time load data of the power facilities, the formula for calculating the load factor of the power facilities is:

[0132] ;

[0133] In the formula, is the load factor of the power facility, The real-time load of the power facility. , and are the weight coefficients of the impact force of water flow on power facilities, the water flow pressure on the submerged part of power facilities, and the real-time load of power facilities, respectively. and , and Both are greater than 0.

[0134] Among them, the bearing coefficient of the power facility is proportional to the impact force of the power facility, the water flow pressure on the submerged part of the power facility and the real-time load of the power facility. The greater the impact force and water flow pressure, the greater the bearing coefficient of the power facility, the greater the real-time load, which leads to instability of the power facility, so it is also proportional to the bearing coefficient of the power facility. And because the real-time load has the greatest impact on the bearing capacity of the power facility, the water flow impact force and water flow pressure are slightly smaller than the real-time load, and the impact of the water flow impact force on the bearing capacity is greater than the water flow pressure, so it is set .

[0135] Step 4: Construct a dynamic risk assessment mathematical model based on the precise value of the average water depth and the bearing coefficient of the power facilities. Formulate a logical expression for calculating the risk score using the dynamic assessment mathematical model based on the precise value of the average water depth and the bearing coefficient of the power facilities to obtain a real-time risk score for the power facilities.

[0136] The real-time power facility risk score calculation is based on the formula:

[0137] ;

[0138] In the formula, Scoring the risk of power facilities, and is the weight coefficient of the average water depth and the load factor of the power facilities, where and and Both are greater than 0.

[0139] Among them, the risk score of power facilities is proportional to the average water depth accuracy and the bearing coefficient of power facilities, that is, the deeper the flood, the higher the risk, and the larger the bearing coefficient of power facilities, the higher the risk. At the same time, since the bearing coefficient of power facilities is more serious in judging the risk than the average water depth accuracy, the risk score is set. .

[0140] Step 5: Compare the real-time risk score of power facilities with the pre-set risk factor threshold, and output different risk warnings and corresponding processing requirements based on different comparison results. The logic is as follows: Calibration is a pre-set risk factor threshold.

[0141] when When the flood disaster is detected, it is judged that the current flood disaster in the area where the power facilities are located is a level one risk and no action is needed at present;

[0142] when When the current flood disaster in the area where the power facilities are located is judged to be a secondary risk, operations to reduce the workload of the power facilities are carried out, and the development of the flood disaster is continuously monitored;

[0143] when When the current flood disaster in the area where the power facilities are located is judged to be a level 3 risk, the current power equipment should be stopped.

[0144] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0145] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0146] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0147] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for constructing a dynamic model for geographical analysis and evaluation of flood disasters in power facilities, characterized in that: The specific steps include: Collect meteorological data, geographic information data and average water depth when flood disasters occurred in the area where the power facilities to be analyzed are located in the past, establish a flood disaster water depth prediction model, use the collected historical meteorological data and geographic information data as the input of the prediction model, and use the corresponding average water depth as a label to train the prediction model; Input the current meteorological data and geographic information data of the area where the power facility to be analyzed is located into the prediction model to obtain the average water depth corresponding to the current flood disaster, and correct the average water depth value output by the flood disaster water depth prediction model based on the hydrological and topographic characteristic parameters of the area where the power facility to be analyzed is located to obtain the accurate value of the average water depth, wherein the hydrological and topographic characteristic parameters include: terrain undulation, land use, soil texture and elevation; The average water depth value output by the flood disaster water depth prediction model is corrected by the hydrological and topographic characteristic parameters to obtain the accurate value of the average water depth. The specific calculation formula of the accurate value of the average water depth is: In the formula, is the precise value of the average water depth, is the average water depth value output by the water depth prediction model, represents the correction function of terrain relief to water depth, is the correction function of land use to water depth, is the correction function of soil texture to water depth, is the correction function of elevation to water depth, where is the terrain relief, For land use, Indicates soil texture, is the elevation; The impact force of water flow on power facilities and the water flow pressure on the submerged parts of power facilities are calculated based on the accurate average water depth and the real-time flow velocity of flood disaster water flow. Based on the obtained impact force and water flow pressure data, combined with the real-time load data of power facilities, the load factor of power facilities is calculated; The impact force of water flow on power facilities and the water flow pressure on the submerged parts of power facilities are calculated based on the accurate value of average water depth and the real-time flow velocity of flood disaster water flow. The formulas for calculating the impact force of water flow on power facilities and the water flow pressure on the submerged parts of power facilities are: In the formula, The impact of water flow on power facilities. is the density of water, is the load-bearing area of ​​the power facility, is the real-time flow rate of water, The water pressure on the submerged part of the power facility. is the acceleration due to gravity, is the precise value of the average water depth, where the stress area of ​​the power facility The calculation is based on the formula: In the formula, The perimeter of the base of the power facility; Based on the obtained impact force and water flow pressure data, combined with the real-time load data of the power facilities, the formula for calculating the load factor of the power facilities is: In the formula, is the load factor of the power facility, The real-time load of the power facility. , and are the weight coefficients of the impact force of water flow on power facilities, the water flow pressure on the submerged part of power facilities, and the real-time load of power facilities, respectively. and , and All are greater than 0; Based on the accurate value of average water depth and the load factor of power facilities, a dynamic risk assessment mathematical model is constructed. According to the accurate value of average water depth and the load factor of power facilities, a logical expression for calculating risk scores of the dynamic assessment mathematical model is formulated to obtain real-time risk scores of power facilities. The real-time risk score of power facilities is compared with the pre-set risk factor threshold, and different risk warnings and corresponding processing requirements are output based on different comparison results.

2. The method for constructing a dynamic model for geographical analysis and assessment of flood disasters in power facilities according to claim 1, characterized in that: The meteorological data of historical flood disasters collected specifically include the rainfall at various times during the historical flood disasters. The average water depth is the average water depth at the corresponding time. The geographic information data includes terrain index, river network density, slope and vegetation normalization index, wherein the vegetation normalization index is calculated through remote sensing images, and the formula for calculating the vegetation normalization index is: In the formula, is the normalized difference vegetation index, is the average near-infrared reflectivity of the area where the power facilities to be analyzed are located, is the average red band reflectance of the area where the power facilities to be analyzed are located; The collected data is enhanced, including interpolation enhancement and time series expansion. The data volume is increased based on data enhancement to generate a sample data set. The meteorological data and geographic information data of historical flood disasters are mapped one by one with the corresponding average water depth of flood disasters and stored in the sample data set.

3. The method for constructing a dynamic model for geographical analysis and assessment of flood disasters in power facilities according to claim 2, characterized in that: A flood disaster water depth prediction model is established based on the LSTM long short-term memory network model, and the activation function and optimization algorithm are selected. The Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model; the formula of the Tanh function is: In the formula, Represents the Tanh function, independent variable Represents the weighted sum of the neuron's input, that is, the result of weighted summation of the input received by the neuron from the previous layer; At the same time, the hyperparameters of the LSTM model are set, and the hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing number, and the number of hidden layer neurons; The number of network layers is set to a four-layer network structure, the number of iterations is set to 300, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 300, the batch size is set to 256, and the number of hidden layer neurons is 32.

4. The method for constructing a dynamic model for geographical analysis and assessment of flood disasters in power facilities according to claim 1, characterized in that: The correction function of the terrain relief to the water depth is The specific calculation is based on the formula: In the formula, is the maximum terrain height in the area, is the influence weight coefficient of terrain relief, Indicates the attenuation coefficient of the depth affected by the terrain relief, which is used to describe the The rate at which the effect of water flow accumulation gradually weakens with the increase of Correction function of land use on water depth The specific expression is: In the formula, is the influence coefficient of land use, is the land use type classification factor; Soil texture correction function for water depth The specific expression based on is: In the formula, is the influence coefficient of soil texture, is the soil texture index, where soil texture index is calculated The formula is: In the formula, are the weight coefficients for the percentage of sand, silt and clay in the soil, respectively, , and are the percentages of sand, silt, and clay in the soil, respectively, where and and All are greater than 0; Correction function of elevation to water depth The specific formula is: In the formula, is the influence coefficient of elevation, Indicates the maximum elevation of the detection area.

5. The method for constructing a dynamic model for geographical analysis and evaluation of flood disasters in power facilities according to claim 4, characterized in that: The real-time power facility risk score calculation is based on the formula: In the formula, Scoring the risk of power facilities, and is the weight coefficient of the average water depth and the load factor of the power facilities, where and and All are greater than 0; According to the real-time power facility risk score, it is compared with the pre-set risk factor threshold. According to different comparison results, different risk warnings and corresponding processing requirements are output. The logic is as follows: is a pre-set risk factor threshold; when When the flood disaster is detected, it is judged that the current flood disaster in the area where the power facilities are located is a level one risk and no action is needed at present; when When the current flood disaster in the area where the power facilities are located is judged to be a secondary risk, operations to reduce the workload of the power facilities are carried out, and the development of the flood disaster is continuously monitored; when When the current flood disaster in the area where the power facilities are located is judged to be a level 3 risk, the current power equipment should be stopped.

Images