Snowmelt runoff prediction method and system based on environmental feature analysis

Abstract

The application provides a snowmelt runoff yield prediction method and system based on environmental feature analysis, relates to the technical field of artificial intelligence, acquires environmental feature information of a preset area, performs evaluation and prediction to generate runoff yield prediction information and runoff yield direction prediction information, further performs flow path prediction optimization, acquires a flow path prediction result, performs loss analysis on the runoff yield prediction information, generates runoff yield backlog and a runoff yield backlog area, and adds the runoff yield backlog and the runoff yield backlog area into the snowmelt runoff yield prediction result, so that the technical problem that the prediction method for snowmelt runoff yield in the prior art is insufficient in intelligence, is incomplete in reference, and is not strict in execution mode, resulting in a certain deviation of the prediction result compared with the actual situation is solved, the prediction deviation is weakened by level progression through prediction of the criterion of snowmelt runoff yield, and the accuracy of the prediction result is further ensured through path optimization screening.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, specifically to a method and system for predicting snowmelt runoff based on environmental feature analysis. Background Technology

[0002] When snowfall is heavy, snow melting should be carried out in a timely manner to avoid safety hazards and maintain the normal ecological environment. At the same time, the runoff caused by snow melting should be effectively treated. To ensure timely and accurate treatment efficiency, snow melting runoff can be predicted in advance to predict the possibility of subsequent floods.

[0003] Currently, the main method for predicting snowmelt runoff is to continuously observe snowmelt areas, collect data, and conduct simulation analysis of runoff generation mechanisms. However, due to the existence of multiple influencing factors and the complexity of the environment, some uncontrollable factors cannot be avoided, affecting the final prediction results. Further technological innovation is needed.

[0004] In existing technologies, the prediction methods for snowmelt runoff lack intelligence, have incomplete reference surfaces, and are not rigorous enough in their execution, resulting in a certain deviation between the prediction results and the actual situation. Summary of the Invention

[0005] This application provides a snowmelt runoff prediction method and system based on environmental feature analysis, which addresses the technical problems in existing snowmelt runoff prediction methods, such as insufficient intelligence, incomplete reference surfaces, and unrigorous execution methods, leading to a certain deviation between the prediction results and the actual situation.

[0006] In view of the above problems, this application provides a method and system for predicting snowmelt runoff based on environmental feature analysis.

[0007] In a first aspect, this application provides a snowmelt runoff prediction method based on environmental characteristic analysis, the method comprising:

[0008] Obtain environmental feature information of a preset area, wherein the environmental feature information includes snow cover feature information, landform feature information and climate feature information;

[0009] Based on the climate characteristic information and the snow accumulation characteristic information, the snow melting rate is evaluated, and snow melting rate parameters are generated;

[0010] Based on the snow melting rate parameters, flow rate analysis is performed to generate flow rate prediction information.

[0011] Based on the geomorphological feature information, the flow direction is analyzed to generate flow direction prediction information;

[0012] Based on the predicted flow direction and the predicted flow rate, the flow path prediction is optimized to obtain the flow path prediction result.

[0013] Based on the flow path prediction results, loss analysis is performed on the production flow prediction information to generate production flow accumulation and production flow accumulation area;

[0014] The runoff accumulation amount and the runoff accumulation area are added to the snowmelt runoff prediction results.

[0015] Secondly, this application provides a snowmelt runoff prediction system based on environmental characteristic analysis, the system comprising:

[0016] An information acquisition module is used to acquire environmental feature information of a preset area, wherein the environmental feature information includes snow cover feature information, landform feature information and climate feature information;

[0017] A snowmelt rate assessment module is used to assess the snowmelt rate based on the climate characteristic information and the snow accumulation characteristic information, and generate snowmelt rate parameters.

[0018] A snowmelt rate prediction module is used to perform snowmelt rate analysis based on the snowmelt rate parameter and generate snowmelt rate prediction information.

[0019] A runoff direction prediction module is used to analyze the runoff direction based on the geomorphological feature information and generate runoff direction prediction information.

[0020] The path prediction module is used to perform flow path prediction optimization based on the flow direction prediction information and the flow rate prediction information, and to obtain the flow path prediction result.

[0021] The loss analysis module is used to perform loss analysis on the production flow prediction information based on the flow path prediction results, and generate production flow accumulation amount and production flow accumulation area.

[0022] An information addition module is used to add the runoff accumulation amount and the runoff accumulation area to the snowmelt runoff prediction results.

[0023] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0024] This application provides a snowmelt runoff prediction method based on environmental feature analysis. The method acquires environmental feature information of a preset area, including snow cover features, landform features, and climate features; performs snowmelt rate assessment to generate snowmelt rate parameters; analyzes runoff based on the snowmelt rate parameters to generate runoff prediction information; analyzes runoff direction based on the landform features to generate runoff direction prediction information; and optimizes flow path prediction based on the runoff direction and runoff prediction information to obtain flow path prediction results. The method also performs loss analysis on the runoff prediction information to generate runoff accumulation and runoff accumulation areas, which are then added to the snowmelt runoff prediction results. This method solves the technical problems in existing snowmelt runoff prediction methods, such as insufficient intelligence, incomplete reference surfaces, and imprecise execution methods, leading to deviations between prediction results and reality. By predicting snowmelt runoff based on criteria, the method progressively weakens prediction bias and further ensures the accuracy of prediction results through path optimization. Attached Figure Description

[0025] Figure 1 This application provides a schematic diagram of a snowmelt runoff prediction method based on environmental feature analysis;

[0026] Figure 2 This application provides a schematic diagram of the snowmelt rate parameter acquisition process in a snowmelt runoff prediction method based on environmental feature analysis;

[0027] Figure 3 This application provides a schematic diagram of the process for obtaining runoff direction prediction information in a snowmelt runoff prediction method based on environmental feature analysis;

[0028] Figure 4 This application provides a schematic diagram of a snowmelt runoff prediction system based on environmental feature analysis.

[0029] Explanation of reference numerals in the attached diagram: Information acquisition module 11, snow melting rate assessment module 12, flow rate prediction module 13, flow rate direction prediction module 14, path prediction module 15, loss analysis module 16, information addition module 17. Detailed Implementation

[0030] This application provides a snowmelt runoff prediction method and system based on environmental feature analysis to address the technical problems in existing snowmelt runoff prediction methods, such as insufficient intelligence, incomplete reference surfaces, and unrigorous execution methods, which lead to deviations between prediction results and actual conditions.

[0031] Example 1

[0032] like Figure 1As shown, this application provides a snowmelt runoff prediction method based on environmental characteristic analysis, the method comprising:

[0033] Step S100: Obtain environmental feature information of a preset area, wherein the environmental feature information includes snow cover feature information, landform feature information and climate feature information;

[0034] Specifically, when snowfall is heavy, timely snow melting is necessary to avoid safety hazards and maintain the normal ecological environment. Simultaneously, the runoff generated by snow melting needs effective management. To ensure timely and accurate management efficiency, snow melt runoff prediction can be conducted in advance for targeted treatment. This application provides a snow melt runoff prediction method based on environmental feature analysis. Based on real-time snow accumulation characteristics, it analyzes regional runoff volume and direction to predict accumulation areas and compression amounts, achieving intelligent and accurate analysis and prediction. Specifically, a predetermined area to be snow melted is defined as the preset area. Snow thickness, snow density, and other data are collected in the preset area as snow accumulation feature information; landform morphology and distribution information are collected as landform feature information; wind speed, temperature, and other environmental information are collected as climate feature information. The snow accumulation feature information, landform feature information, and climate feature information are integrated and standardized as environmental feature information. This environmental feature information is an influencing factor on the efficiency of snow melting management and provides a basic basis for subsequent runoff prediction.

[0035] Step S200: Evaluate the snowmelt rate based on the climate characteristic information and the snow accumulation characteristic information, and generate snowmelt rate parameters;

[0036] Furthermore, such as Figure 2 As shown, the step S200 of this application further includes: estimating the snowmelt rate based on the climate characteristic information and the snow accumulation characteristic information to generate snowmelt rate parameters.

[0037] Step S210: The climate characteristic information includes temperature characteristic information and wind speed characteristic information;

[0038] Step S220: The snow accumulation feature information includes snow thickness information and snow density information;

[0039] Step S230: Collect snow melting data based on the temperature characteristic information, wind speed characteristic information, snow thickness information, and snow density information to obtain snow melting record data;

[0040] Step S240: Based on the snow melting record data, train a snow melting rate evaluation model using an integrated neural network;

[0041] Step S250: Input the temperature characteristic information, the wind speed characteristic information, the snow thickness information, and the snow density information into the snow melting rate evaluation model to obtain the snow melting rate parameters.

[0042] Furthermore, in the step S240 of this application, which involves training a snowmelt rate evaluation model based on the snowmelt record data and an integrated neural network, the following additional steps are included:

[0043] Step S241: Based on the snow melting record data, perform correlation analysis by traversing the temperature feature information, the wind speed feature information, the snow thickness information, and the snow density information to obtain the correlation degree between the associated snow melting record data and the feature indicators;

[0044] Step S242: The associated snow melting record data includes temperature characteristic associated record data, wind speed characteristic associated record data, snow thickness associated record data, and snow density associated record data;

[0045] Step S243: The correlation degree of the feature indicators includes the correlation degree of temperature feature, the correlation degree of wind speed feature, the correlation degree of snow thickness, and the correlation degree of snow density;

[0046] Step S244: Based on the temperature characteristics associated with the recorded data, train the first sub-model for evaluating the snow melting rate;

[0047] Step S245: Train the second sub-model for snowmelt rate evaluation based on the wind speed characteristics associated with the recorded data;

[0048] Step S246: Train a third sub-model for evaluating snow melting rate based on the snow thickness correlation record data;

[0049] Step S247: Based on the snow density correlation record data, train the fourth sub-model for snow melting rate evaluation;

[0050] Step S248: Based on the correlation degree of the temperature feature, the correlation degree of the wind speed feature, the correlation degree of the snow thickness, and the correlation degree of the snow density, merge the first sub-model of the snow melting rate assessment, the second sub-model of the snow melting rate assessment, the third sub-model of the snow melting rate assessment, and the fourth sub-model of the snow melting rate assessment to generate the snow melting rate assessment model, wherein the output weight ratio of any sub-model is equal to the correlation degree ratio.

[0051] Furthermore, step S241 of this application further includes: performing correlation analysis on the temperature characteristic information, wind speed characteristic information, snow thickness information, and snow density information based on the snow melting record data to obtain the correlation degree between the snow melting record data and the characteristic indicators;

[0052] Step S2411: Using temperature feature recorded data and snow melting rate recorded data as variables, and wind speed feature recorded data, snow thickness recorded data and snow density recorded data as quantitative data, the snow melting record data is filtered to obtain the temperature feature associated record data;

[0053] Step S2412: Using the wind speed characteristic record data and the snow melting rate record data as variables, and the temperature characteristic record data, the snow thickness record data, and the snow density record data as quantitative data, the snow melting record data is filtered to obtain the wind speed characteristic associated record data;

[0054] Step S2413: Using the snow thickness feature record data and the snow melting rate record data as variables, and the temperature feature record data, the wind speed feature record data, and the snow density record data as quantitative data, the snow melting record data is filtered to obtain the snow thickness associated record data;

[0055] Step S2414: Using the snow density characteristic record data and the snow melting rate record data as variables, and the temperature characteristic record data, the wind speed characteristic record data, and the snow thickness record data as quantitative data, the snow melting record data is filtered to obtain the snow density associated record data;

[0056] Step S2415: Traverse the temperature feature correlation record data, the wind speed feature correlation record data, the snow thickness correlation record data, and the snow density correlation record data to perform correlation analysis and generate the correlation degree of the feature index.

[0057] Specifically, the climate characteristic information and the snow accumulation characteristic information are used as factors influencing snowmelt, and feature profiling and data collection analysis are performed to assess the snowmelt rate. The climate characteristic information includes temperature characteristic information and wind speed characteristic information. Multiple data collection points are evenly distributed in the snow accumulation area of ​​the preset region. For example, sensing devices can be deployed at these multiple data collection points to detect and collect characteristic information, such as collecting temperature data at each location and performing hierarchical clustering analysis. Finally, the characteristic temperature in multiple clusters, the number of clustered locations in each cluster, and the weighted average are calculated as the temperature characteristic to weaken the uniformity of temperature differences and ensure the universality and regional coverage of the final determined temperature characteristic. Similarly, wind speed data from multiple data collection points are collected, and data processing analysis is performed to determine the wind speed characteristic information. Specifically, the snow accumulation characteristic information includes snow thickness information and snow density information. Snow accumulation data from the multiple data collection points are collected, and clustering and weighted average calculations are performed respectively to determine the snow thickness information and snow density information.

[0058] The determination and processing methods for the temperature characteristic information, wind speed characteristic information, snow thickness information, and snow density information are the same, and they are normalized to form the snowmelt record data. Further, based on the snowmelt record data, a snowmelt rate evaluation model is generated through neural network training.

[0059] Specifically, based on the snowmelt record data, the temperature characteristic information, wind speed characteristic information, snow thickness information, and snow density information are adjusted and limited to perform single-factor correlation analysis on the snowmelt rate, accurately measuring the correlation degree of each characteristic. Using the wind speed characteristic data, snow thickness data, and snow density data as quantitative data, a preset number of correlated record data pairs that meet the quantitative criteria are selected from the snowmelt record data. These pairs map the corresponding temperature characteristic data to the snowmelt rate data and are considered the temperature-related influence record data. The difference between the absolute value of the change in snowmelt rate and the absolute value of the change in temperature is calculated and then divided by the preset number to measure the degree of correlation, denoted as the temperature correlation degree. The larger the difference, the greater the change in snowmelt rate caused by the temperature change, and thus the greater the correlation degree.

[0060] Similarly, using the temperature characteristic data, snow thickness data, and snow density data as quantitative data, a preset number of pairs of correlated records that meet the quantitative criteria are selected from the snowmelt record data. This means mapping the corresponding wind speed characteristic data to the snowmelt rate data, which are then used as the wind speed characteristic correlated record data. The difference between the absolute value of the snowmelt rate change and the absolute value of the wind speed change is calculated and then divided by the preset number to obtain the wind speed correlation degree. Similarly, using the temperature characteristic data, wind speed characteristic data, and snow density data as quantitative data, a preset number of snow thickness characteristic data and snowmelt rate data that meet the quantitative criteria are selected from the snowmelt record data to obtain the snow thickness data. The snow thickness correlation is calculated by comparing the absolute value of the change in snow melting rate with the absolute value of the change in snow thickness, and then dividing by a preset number. The temperature, wind speed, and snow thickness records are used as quantitative data. From the snow melting records, a preset number of snow density records and snow melting rate records that meet the quantitative criteria are selected as the snow density correlation records. The difference between the absolute value of the change in snow melting rate and the absolute value of the change in snow density is calculated and then divided by a preset number to obtain the snow density correlation. The temperature correlation, wind speed correlation, snow thickness correlation, and snow density correlation are integrated to form the feature index correlation.

[0061] Further, the temperature feature-related recorded data is extracted, and hierarchical identification nodes are determined based on the temperature feature recorded data. Hierarchical decision nodes are determined based on the snow melting rate recorded data. Node mapping and association are performed, and these are used as sample data for neural network training to generate the first sub-model for snow melting rate evaluation. Similarly, the wind speed feature-related recorded data is extracted, and the second sub-model for snow melting rate evaluation is constructed through neural network training; the snow thickness-related recorded data is extracted, and the third sub-model for snow melting rate is generated through neural network training; the snow density-related recorded data is extracted, and the fourth sub-model for snow melting rate is generated through neural network training.

[0062] Furthermore, the correlation ratios of the temperature feature correlation, wind speed feature correlation, snow thickness correlation, and snow density correlation are measured separately to determine the feature weight distribution. Then, the first, second, third, and fourth sub-models for snowmelt rate assessment are weighted, and the models are merged to generate the snowmelt rate assessment model. The output of the snowmelt rate assessment model is a weighted calculation of the outputs of each sub-model, further improving the objectivity and accuracy of the model analysis.

[0063] Furthermore, the temperature characteristic information, wind speed characteristic information, snow thickness information, and snow density information are input into the snowmelt rate evaluation model. By identifying and matching the information to the corresponding sub-model, data analysis is performed to output the corresponding single snowmelt rate. Then, the snowmelt rate parameters are obtained by weighted calculation to ensure the real-time state fit of the snowmelt rate parameters. The snowmelt rate parameters are used as a reference indicator for predicting snowmelt runoff.

[0064] Step S300: Analyze the snowmelt rate parameters to generate snow production flow and generate snow production flow prediction information;

[0065] Step S400: Analyze the runoff direction based on the geomorphic feature information to generate runoff direction prediction information;

[0066] Specifically, based on the snowmelt rate parameter, the snowmelt volume per unit time in the preset area is measured to determine the volume ratio before and after snowmelt. This ratio is used as a benchmark to predict the runoff generation per unit time, which is then used as the runoff generation prediction information. Furthermore, the topographic distribution of the preset area is a decisive factor in determining the runoff generation information. Elevation characteristic information of the preset area is collected, including elevation data from multiple data collection points. Neighborhood elevation deviation is calculated, and an elevation deviation threshold is set for regional clustering. The regional clustering results are then analyzed for geomorphological features to identify concave regions, i.e., runoff convergence areas, which are used as the final runoff direction. The distribution information of these concave regions is used as the runoff direction prediction result. Based on the runoff direction prediction result, flow path prediction is further performed.

[0067] Furthermore, such as Figure 3 As shown, the step S400 of this application further includes: analyzing the runoff direction based on the geomorphic feature information to generate runoff direction prediction information.

[0068] Step S410: The geomorphic feature information includes elevation feature information;

[0069] Step S420: Perform cluster analysis on the preset region based on the altitude feature information to generate region clustering results;

[0070] Step S430: Obtain the distribution information of the concave region based on the region clustering results;

[0071] Step S440: Determine the flow direction prediction information based on the concave region distribution information.

[0072] Specifically, multiple data collection points are determined, their corresponding altitudes are collected, and their locations are identified. This information is then organized to serve as the altitude feature information. Further, an altitude deviation threshold is set, which is the critical value for altitude difference in regional clustering. The altitude deviation of any adjacent points within the preset region is calculated and compared with the altitude deviation threshold. If the altitude deviation is less than or equal to the threshold, the points are clustered into one region; otherwise, they belong to two regions. This threshold determination and region assignment is repeated multiple times for multiple data collection points within the preset region to obtain multiple clustered regions, which serve as the regional clustering result.

[0073] Furthermore, based on the regional clustering results, the distribution information of the concave regions is extracted through geomorphological recognition. For example, multi-angle, full-coverage images of the preset region can be collected for location positioning and feature recognition to determine the concave regions and accurately locate the possible runoff origin. Generally, runoff after snowmelt converges towards lower elevations, i.e., towards the concave regions. The snowmelt direction is determined based on the distribution information of the concave regions, serving as the predicted runoff direction information and providing a basis for subsequent flow path analysis.

[0074] Step S500: Optimize the flow path prediction based on the flow direction prediction information and the flow rate prediction information, and obtain the flow path prediction result;

[0075] Step S600: Based on the flow path prediction results, perform loss analysis on the production flow prediction information to generate production flow accumulation and production flow accumulation area;

[0076] Step S700: Add the runoff accumulation amount and the runoff accumulation area to the snowmelt runoff prediction results.

[0077] Specifically, the predicted flow direction information and the predicted flow rate information are used as prediction references. Based on the distribution information of the concave region, path prediction is performed according to the flow path optimization fitness function to obtain the flow path prediction result. The flow path prediction result is a possible path that meets the fitness definition and is not unique.

[0078] Furthermore, during the flow process, there will be certain flow loss, such as flow diffusion and soil absorption, which will affect the final accumulated flow. Based on the flow path prediction results, loss analysis can be performed. Historical flow absorption data for each path in the prediction results can be evaluated and analyzed, discarding flow loss. Based on the flow prediction information, the accumulated flow in each concave region can be determined, and a threshold for the accumulated flow can be set, i.e., a critical value limiting the capacity of the accumulated flow. For example, a critical capacity with the necessity of accumulation treatment can be used as the threshold to filter areas with high accumulation and ignore irrelevant areas, thereby reducing the workload of subsequent accumulation treatment. The concave regions that meet the threshold are designated as the accumulated flow regions, and the corresponding accumulation volume is designated as the accumulated flow. The accumulated flow and the accumulated flow regions are added to the snowmelt runoff prediction results. Based on the snowmelt runoff prediction results, the accumulated flow corresponding to the accumulated flow regions can be targeted for treatment to achieve the optimal treatment effect.

[0079] Furthermore, the step S500 of this application, which involves optimizing the flow path prediction based on the flow direction prediction information and the flow rate prediction information to obtain the flow path prediction result, further includes:

[0080] Step S510: Arrange the concave region distribution information according to the predicted flow direction information to obtain the concave region distribution sequence;

[0081] Step S520: Obtain the i-th concave region according to the concave region distribution sequence;

[0082] Step S530: Based on the i-th concave region, extract the (i+1)-th concave region set from the concave region distribution sequence;

[0083] Step S540: Obtain the fitness function for flow path optimization:

[0084]

[0085] l i ∩l (i+1)j ≥l0

[0086] H i -H (i+1)j >0

[0087] Among them, l i Characterizing the width of the flow outlet in the i-th concave region, l (i+1)j H represents the width of the flow inlet of the j-th concave region in the (i+1)-th concave region set. i H represents the elevation of the runoff outlet in the i-th concave region. (i+1)j The elevation of the runoff inlet of the j-th concave region in the (i+1)-th concave region set, w1 represents... The weights, w2 represents H i -H (i+1)j The weights, D (i→i+1)j The fitness of the j-th concave region in the (i+1)-th concave region set is represented by l0, which is the preset minimum overlap threshold.

[0088] Step S550: Based on the flow path optimization fitness function, traverse the (i+1)th concave region set to perform flow path prediction optimization and obtain the flow path prediction result.

[0089] Furthermore, the step S540 of this application further includes: traversing the (i+1)th concave region set to perform flow path prediction optimization based on the flow path optimization fitness function, and obtaining the flow path prediction result.

[0090] Step S551: Obtain the j-th concave region based on the (i+1)-th concave region set;

[0091] Step S552: Based on the i-th concave region and the j-th concave region, evaluate the fitness of the j-th concave region using the flow path optimization fitness function to generate the fitness of the j-th concave region;

[0092] Step S553: ​​Determine whether the fitness of the j-th concave region meets the fitness threshold;

[0093] Step S554: If satisfied, add the j-th concave region to the flow path prediction result.

[0094] Specifically, based on the geomorphic features of the preset area, runoff direction prediction is performed to obtain the runoff direction prediction information. Then, the distribution information of the concave regions is sorted based on their distribution location to generate a concave region distribution sequence. Preferably, the concave regions can be regionalized and labeled for differentiation. Further, based on the concave region distribution sequence, a region is randomly extracted as the i-th concave region, and multiple regions bordering the i-th concave region are extracted as the (i+1)-th concave region set. Further, a flow path optimization fitness function is constructed to optimize the flow path prediction. A comprehensive analysis of the bordering regions can be performed, taking into account the diversion situation during the flow process, thus improving the actual fit of the prediction results.

[0095] Specifically, obtain the fitness function for the flow path optimization: l i ∩l (i+1)j ≥l0, H i -H (i+1)j >0, where l i Characterizing the width of the flow outlet in the i-th concave region, l (i+1)j H represents the width of the flow inlet of the j-th concave region in the (i+1)-th concave region set. i H represents the elevation of the runoff outlet in the i-th concave region. (i+1)j The elevation of the runoff inlet of the j-th concave region in the (i+1)-th concave region set, w1 represents... The weights, w2 represents H i -H (i+1)j The weights, D (i→i+1)j The fitness of the j-th concave region in the (i+1)-th concave region set is represented by l0, which is a preset minimum overlap threshold. These parameters can be obtained directly through data collection and statistics. Characterizing interface overlap, H i -H (i+1)jThe interface elevation deviation is characterized. The flow path optimization fitness function is used as the optimization prediction execution tool to calculate the fitness of the (i+1)th concave region set, thereby determining the flow path prediction result.

[0096] Specifically, based on the (i+1)th concave region set, any adjacent concave region, i.e., the jth concave region, is extracted. Regional parameters are collected for the ith and jth concave regions and input into the flow path optimization fitness function to calculate the fitness of the jth concave region, indicating the feasibility of a flow path from the ith to the jth concave region. A fitness threshold is further set, i.e., the critical value for determining the feasibility of the flow path. It is determined whether the fitness of the jth concave region meets the fitness threshold. If it does not meet the threshold, the path has low feasibility and is excluded. If it does meet the threshold, the flow path has some feasibility, and the jth concave region is the final convergence region, which is added to the flow path prediction result. Flow path fitness analysis is performed on the adjacent regions of each concave region to ensure the coverage completeness and accuracy of the flow path prediction results.

[0097] The snowmelt runoff prediction method based on environmental feature analysis provided in this application has the following technical effects:

[0098] 1. This invention provides a snowmelt runoff prediction method and system based on environmental feature analysis, relating to the field of artificial intelligence technology. It acquires environmental feature information of a preset area, evaluates and predicts to generate runoff prediction information and runoff direction prediction information; then, it performs flow path prediction and optimization to obtain flow path prediction results; it performs loss analysis on the runoff prediction information to generate runoff accumulation amount and runoff accumulation area, adding these to the snowmelt runoff prediction results. This solves the technical problems in existing snowmelt runoff prediction methods, such as insufficient intelligence, incomplete reference surfaces, and imprecise execution methods, leading to deviations between prediction results and reality. By predicting snowmelt runoff based on criteria, it progressively weakens prediction bias and further ensures the accuracy of prediction results through path optimization screening.

[0099] 2. Model and conduct snowmelt rate analysis of multi-dimensional influence characteristics to improve the accuracy of the analysis and ensure the efficiency of subsequent analysis; divide the region and evaluate the feasibility of flow paths between regions, select those with higher adaptability, and further improve the actual fit of the final prediction results.

[0100] Example 2

[0101] Based on the same inventive concept as the snowmelt runoff prediction method based on environmental feature analysis in the foregoing embodiments, such as Figure 4As shown, this application provides a snowmelt runoff prediction system based on environmental characteristic analysis, the system comprising:

[0102] Information acquisition module 11, the information acquisition module 11 is used to acquire environmental feature information of a preset area, wherein the environmental feature information includes snow cover feature information, landform feature information and climate feature information;

[0103] Snow melting rate assessment module 12, which is used to assess the snow melting rate based on the climate characteristic information and the snow accumulation characteristic information, and generate snow melting rate parameters;

[0104] Snowmelt rate prediction module 13 is used to perform snowmelt rate analysis based on the snowmelt rate parameter and generate snowmelt rate prediction information.

[0105] Runoff direction prediction module 14, which is used to perform runoff direction analysis based on the geomorphological feature information and generate runoff direction prediction information;

[0106] Path prediction module 15, the path prediction module 15 is used to perform flow path prediction optimization based on the flow direction prediction information and the flow rate prediction information, and obtain flow path prediction results;

[0107] Loss analysis module 16 is used to perform loss analysis on the production flow prediction information based on the flow path prediction result, and generate production flow accumulation amount and production flow accumulation area.

[0108] Information addition module 17 is used to add the runoff accumulation amount and the runoff accumulation area to the snowmelt runoff prediction results.

[0109] Furthermore, the system also includes:

[0110] A climate feature information analysis module, wherein the climate feature information analysis module is used to analyze the climate feature information, including temperature feature information and wind speed feature information;

[0111] A snow cover feature information analysis module, wherein the snow cover feature information analysis module is used to analyze the snow cover feature information, including snow thickness information and snow density information;

[0112] The snow melting data acquisition module is used to acquire snow melting data based on the temperature characteristic information, the wind speed characteristic information, the snow thickness information, and the snow density information, and obtain snow melting record data.

[0113] A model training module is used to train a snowmelt rate evaluation model based on an integrated neural network according to the snowmelt record data.

[0114] The model analysis module is used to input the temperature characteristic information, the wind speed characteristic information, the snow thickness information, and the snow density information into the snowmelt rate evaluation model to obtain the snowmelt rate parameters.

[0115] Furthermore, the system also includes:

[0116] The correlation analysis module is used to perform correlation analysis on the snow melting record data by traversing the temperature feature information, the wind speed feature information, the snow thickness information and the snow density information, and to obtain the correlation degree between the snow melting record data and the feature indicators.

[0117] A snowmelt record data analysis module is used to analyze the snowmelt record data, which includes temperature characteristic related record data, wind speed characteristic related record data, snow thickness related record data, and snow density related record data.

[0118] The feature index correlation analysis module is used to analyze the correlation of the feature indicators, including temperature feature correlation, wind speed feature correlation, snow thickness correlation, and snow density correlation.

[0119] The first sub-model training module is used to train a first sub-model for evaluating snow melting rate based on the temperature characteristics associated with the recorded data.

[0120] The second sub-model training module is used to train the snowmelt rate evaluation second sub-model based on the wind speed characteristics and associated recorded data.

[0121] The third sub-model training module is used to train the snow melting rate evaluation third sub-model based on the snow thickness associated with the recorded data.

[0122] The fourth sub-model training module is used to train the snow melting rate evaluation fourth sub-model based on the snow density correlation record data.

[0123] The model merging module is used to merge the first sub-model, the second sub-model, the third sub-model, and the fourth sub-model of snowmelt rate assessment based on the correlation degree of the temperature feature, the correlation degree of the wind speed feature, the correlation degree of the snow thickness, and the correlation degree of the snow density, to generate the snowmelt rate assessment model, wherein the output weight ratio of any sub-model is equal to the correlation degree ratio.

[0124] Furthermore, the system also includes:

[0125] A temperature feature-related record data filtering module is used to filter the snow melting record data using temperature feature record data and snow melting rate record data as variables, and wind speed feature record data, snow thickness record data and snow density record data as quantitative data, to obtain the temperature feature-related record data.

[0126] A wind speed feature-related record data filtering module is used to filter the snow melting record data using the wind speed feature record data and the snow melting rate record data as variables, and the temperature feature record data, the snow thickness record data and the snow density record data as quantitative data, to obtain the wind speed feature-related record data.

[0127] A snow thickness-related record data filtering module is used to filter the snow melting record data using the snow thickness feature record data and the snow melting rate record data as variables, and the temperature feature record data, the wind speed feature record data, and the snow density record data as quantitative data, to obtain the snow thickness-related record data.

[0128] A snow density correlation record data filtering module is used to filter the snow melting record data using the snow density feature record data and the snow melting rate record data as variables, and the temperature feature record data, the wind speed feature record data and the snow thickness record data as quantitative data, to obtain the snow density correlation record data.

[0129] The data correlation analysis module is used to traverse the temperature feature correlation record data, the wind speed feature correlation record data, the snow thickness correlation record data, and the snow density correlation record data to perform correlation analysis and generate the correlation degree of the feature index.

[0130] Furthermore, the system also includes:

[0131] A geomorphic feature information analysis module, wherein the geomorphic feature information analysis module is used to analyze the geomorphic feature information, including elevation feature information;

[0132] A region clustering module is used to perform cluster analysis on the preset region based on the altitude feature information and generate region clustering results;

[0133] A concave region distribution information determination module is used to obtain concave region distribution information based on the region clustering results;

[0134] A direction prediction module is used to determine the flow direction prediction information based on the distribution information of the concave region.

[0135] Furthermore, the system also includes:

[0136] A sequence acquisition module is used to arrange the concave region distribution information according to the flow direction prediction information to obtain a concave region distribution sequence;

[0137] The i-th concave region acquisition module is used to acquire the i-th concave region according to the concave region distribution sequence;

[0138] A region set extraction module is used to extract the (i+1)th concave region set from the concave region distribution sequence based on the i-th concave region.

[0139] The function acquisition module is used to acquire the fitness function for flow path optimization.

[0140]

[0141] l i ∩l (i+1)j ≥l0

[0142] H i -H (i+1)j >0

[0143] Among them, l i Characterizing the width of the flow outlet in the i-th concave region, l (i+1)j H represents the width of the flow inlet of the j-th concave region in the (i+1)-th concave region set. i H represents the elevation of the runoff outlet in the i-th concave region. (i+1)j The elevation of the runoff inlet of the j-th concave region in the (i+1)-th concave region set, w1 represents... The weights, w2 represents H i -H (i+1)j The weights, D (i→i+1)j The fitness of the j-th concave region in the (i+1)-th concave region set is represented by l0, which is the preset minimum overlap threshold.

[0144] The path prediction and optimization module is used to perform flow path prediction and optimization by traversing the (i+1)th concave region set according to the flow path optimization fitness function, and to obtain the flow path prediction result.

[0145] Furthermore, the system also includes:

[0146] The j-th concave region acquisition module is used to acquire the j-th concave region based on the (i+1)-th concave region set.

[0147] A region fitness generation module is used to evaluate the flow path optimization fitness function based on the i-th concave region and the j-th concave region to generate the fitness of the j-th concave region.

[0148] A threshold determination module is used to determine whether the fitness of the j-th concave region meets the fitness threshold.

[0149] A region addition module is used to add the j-th concave region to the flow path prediction result if a certain condition is met.

[0150] Through the foregoing detailed description of a snowmelt runoff prediction method based on environmental feature analysis, those skilled in the art can clearly understand the snowmelt runoff prediction method and system based on environmental feature analysis in this embodiment. As for the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section description.

[0151] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting snowmelt runoff based on environmental characteristic analysis, characterized in that, include: Obtain environmental feature information of a preset area, wherein the environmental feature information includes snow cover feature information, landform feature information and climate feature information; Based on the climate characteristic information and the snow accumulation characteristic information, the snow melting rate is evaluated, and snow melting rate parameters are generated; Based on the snow melting rate parameters, flow rate analysis is performed to generate flow rate prediction information. Based on the geomorphological feature information, the flow direction is analyzed to generate flow direction prediction information; Based on the predicted flow direction and the predicted flow rate, the flow path prediction is optimized to obtain the flow path prediction result. Based on the flow path prediction results, loss analysis is performed on the production flow prediction information to generate production flow accumulation and production flow accumulation area; The runoff accumulation amount and the runoff accumulation area are added to the snowmelt runoff prediction results; The step of analyzing the runoff direction based on the geomorphological feature information and generating runoff direction prediction information includes: The geomorphic feature information includes elevation feature information; Based on the altitude feature information, cluster analysis is performed on the preset region to generate region clustering results; Based on the region clustering results, obtain the distribution information of the concave regions; Based on the distribution information of the concave region, the predicted flow direction information is determined; The step of optimizing the flow path prediction based on the flow direction prediction information and the flow rate prediction information to obtain the flow path prediction result includes: Based on the predicted flow direction information, the concave region distribution information is arranged to obtain a concave region distribution sequence; Based on the distribution sequence of the concave regions, the i-th concave region is obtained; Based on the i-th concave region, extract the (i+1)-th concave region set from the concave region distribution sequence; Obtain the fitness function for flow path optimization: in, Characterizing the width of the flow outlet in the i-th concave region, Characterizes the width of the flow inlet of the j-th concave region in the (i+1)-th concave region set. Characterizing the elevation of the runoff outlet in the i-th concave region, The elevation of the runoff inlet of the j-th concave region in the (i+1)-th concave region set. Characterization The weight, Characterization The weight, Characterizes the fitness of the j-th concave region in the (i+1)-th concave region set. This is the preset minimum overlap threshold; Based on the flow path optimization fitness function, the flow path prediction optimization is performed by traversing the (i+1)th concave region set to obtain the flow path prediction result.

2. The method as described in claim 1, characterized in that, The step of assessing the snowmelt rate based on the climate characteristic information and the snow accumulation characteristic information, and generating snowmelt rate parameters, includes: The climate characteristic information includes temperature characteristic information and wind speed characteristic information; The snow cover feature information includes snow thickness information and snow density information; Snow melting data is collected based on the temperature characteristic information, wind speed characteristic information, snow thickness information, and snow density information to obtain snow melting record data; Based on the snow melting record data, a snow melting rate evaluation model is trained using an integrated neural network. The temperature characteristic information, the wind speed characteristic information, the snow thickness information, and the snow density information are input into the snow melting rate evaluation model to obtain the snow melting rate parameters.

3. The method as described in claim 2, characterized in that, The step of training a snowmelt rate assessment model based on the snowmelt record data and an integrated neural network includes: Based on the snow melting record data, a correlation analysis is performed on the temperature feature information, the wind speed feature information, the snow thickness information, and the snow density information to obtain the correlation degree between the associated snow melting record data and the feature indicators. The associated snow melting record data includes temperature characteristic associated record data, wind speed characteristic associated record data, snow thickness associated record data, and snow density associated record data; The correlation of the feature indicators includes temperature feature correlation, wind speed feature correlation, snow thickness correlation, and snow density correlation; Based on the temperature characteristics and associated recorded data, train a first sub-model for snow melting rate evaluation; Based on the wind speed characteristics and associated recorded data, a second sub-model for snowmelt rate evaluation is trained. Based on the snow thickness correlation record data, train a third sub-model for snow melting rate evaluation; Based on the snow density correlation record data, train the fourth sub-model for snow melting rate evaluation; Based on the correlation of temperature features, wind speed features, snow thickness, and snow density, the first sub-model, the second sub-model, the third sub-model, and the fourth sub-model for evaluating snow melting rate are merged to generate the snow melting rate evaluation model, wherein the output weight of any sub-model is equal to the correlation ratio.

4. The method as described in claim 3, characterized in that, The step involves performing a correlation analysis on the temperature characteristic information, wind speed characteristic information, snow thickness information, and snow density information based on the snow melt record data to obtain the correlation degree between the associated snow melt record data and the characteristic indicators, including: Using temperature characteristic data and snow melting rate data as variables, and wind speed characteristic data, snow thickness data, and snow density data as quantitative data, the snow melting record data is filtered to obtain the temperature characteristic associated record data. Using the wind speed characteristic record data and the snow melting rate record data as variables, and the temperature characteristic record data, the snow thickness record data, and the snow density record data as quantitative data, the snow melting record data is filtered to obtain the wind speed characteristic associated record data; Using the snow thickness characteristic record data and the snow melting rate record data as variables, and the temperature characteristic record data, the wind speed characteristic record data, and the snow density record data as quantitative data, the snow melting record data is filtered to obtain the snow thickness related record data; Using the snow density characteristic record data and the snow melting rate record data as variables, and the temperature characteristic record data, the wind speed characteristic record data, and the snow thickness record data as quantitative data, the snow melting record data is filtered to obtain the snow density associated record data; The correlation analysis is performed by traversing the temperature feature correlation record data, the wind speed feature correlation record data, the snow thickness correlation record data, and the snow density correlation record data to generate the correlation degree of the feature index.

5. The method as described in claim 1, characterized in that, The step of performing flow path prediction optimization by traversing the (i+1)th concave region set according to the flow path optimization fitness function and obtaining the flow path prediction result includes: Based on the set of (i+1)th concave regions, obtain the jth concave region; Based on the i-th concave region and the j-th concave region, the fitness of the j-th concave region is generated by evaluating the fitness based on the flow path optimization function. Determine whether the fitness of the j-th concave region meets the fitness threshold. If satisfied, the j-th concave region is added to the flow path prediction result.

6. A snowmelt runoff prediction system based on environmental characteristic analysis, characterized in that, The system for predicting snowmelt runoff based on environmental feature analysis as described in claim 1 comprises: An information acquisition module is used to acquire environmental feature information of a preset area, wherein the environmental feature information includes snow cover feature information, landform feature information and climate feature information; A snowmelt rate assessment module is used to assess the snowmelt rate based on the climate characteristic information and the snow accumulation characteristic information, and generate snowmelt rate parameters. A snowmelt rate prediction module is used to perform snowmelt rate analysis based on the snowmelt rate parameter and generate snowmelt rate prediction information. A runoff direction prediction module is used to analyze the runoff direction based on the geomorphological feature information and generate runoff direction prediction information. The path prediction module is used to perform flow path prediction optimization based on the flow direction prediction information and the flow rate prediction information, and to obtain the flow path prediction result. The loss analysis module is used to perform loss analysis on the production flow prediction information based on the flow path prediction results, and generate production flow accumulation amount and production flow accumulation area. An information addition module is used to add the runoff accumulation amount and the runoff accumulation area to the snowmelt runoff prediction results.

Images