Intelligent monitoring system of secondary flood and sediment based on multi-source data

Abstract

The present application provides a sub-flood and sediment intelligent monitoring system based on multi-source data, which relates to the field of data processing technology. The system includes: a data acquisition unit for constructing a historical data set; a spatial data analysis unit for constructing a flood and sediment spatial analysis model; a time series analysis unit for constructing a flood and sediment trend analysis model; a spatiotemporal analysis information parsing unit for generating an intelligent monitoring module; and a monitoring and analysis unit for obtaining flood and sediment monitoring results. The present application can solve the technical problem that the traditional monitoring system has incomplete monitoring results and cannot respond to flood and sediment changes in a timely manner due to a single data source and a lack of comprehensive spatiotemporal analysis, thereby achieving the technical effect of improving the accuracy, comprehensiveness and real-time performance of flood and sediment monitoring by integrating spatiotemporal analysis information and embedding it into a real-time monitoring system.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an intelligent monitoring system for secondary floods and sediments based on multi-source data. Background Art

[0002] With the frequent occurrence of global climate change and extreme weather events, flood and soil erosion problems are becoming increasingly serious, posing a major challenge to watershed management, flood prevention and mitigation, and ecological and environmental protection. Monitoring and management of secondary flood sediment (i.e., sediment carried by water flow during secondary floods) is crucial for assessing river health, predicting flood disaster risks, and formulating soil and water conservation measures. However, with the increase in watershed area, complex changes in rainfall patterns, and the profound impact of human activities on the natural environment, monitoring and controlling secondary flood sediment has become increasingly complex and difficult. Traditional flood sediment monitoring systems are often limited by a single data source and lack of comprehensive spatiotemporal analysis, resulting in incomplete monitoring results and difficulty in fully reflecting the dynamic changes of flood sediment. At the same time, these systems often fail to respond to sudden changes in flood sediment in a timely manner, posing huge challenges to flood prevention and mitigation and soil and water conservation. Summary of the invention

[0003] The purpose of this application is to provide an intelligent monitoring system for floods and sediments based on multi-source data, in order to solve the technical problems that traditional monitoring systems have incomplete monitoring results and cannot respond to flood sand changes in a timely manner due to a single data source and lack of comprehensive spatiotemporal analysis.

[0004] In view of the above problems, the present application provides an intelligent monitoring system for secondary flood and sediment based on multi-source data.

[0005] The present application provides a sub-flood and sediment intelligent monitoring system based on multi-source data, and the system includes: a data acquisition unit, which is used to collect historical flood and sediment data of the target basin, and construct a historical data set including flow, sediment content, rainfall, basin topography, and land use data; a spatial data analysis unit, which is used to perform spatial data analysis based on the historical data set, and construct a flood and sediment spatial analysis model; a time series analysis unit, which is used to perform time series analysis based on the historical data set, and construct a flood and sediment trend analysis model; a spatiotemporal analysis information parsing unit, which is used to integrate the analysis results of the flood and sediment spatial analysis model and the flood and sediment trend analysis model, parse the flood and sediment spatiotemporal analysis information, and generate an intelligent monitoring module; a monitoring and analysis unit, which is used to fit the intelligent monitoring module to the real-time monitoring system, perform monitoring and analysis based on the real-time monitoring data of the target basin, and obtain flood and sediment monitoring results.

[0006] The technical solution provided in this application has at least the following technical effects or advantages:

[0007] The above-mentioned sub-flood sand intelligent monitoring system based on multi-source data collects historical data of the target basin, which covers key indicators such as flow, sediment content, rainfall during the flood, as well as the topography and land use of the basin. Subsequently, the spatial data analysis technology is used to deeply explore the collected historical data, understand the spatial distribution and change law of flood sand, and construct a flood sand spatial analysis model. This model can reveal the complex spatial relationship of flood sand and provide a spatial perspective for understanding the flood sand phenomenon. Through the time series analysis technology, the same data set is analyzed in the time dimension to explore the evolution trend and periodic characteristics of flood sand in time, and then a flood sand trend analysis model is constructed. This model can help predict the possible changes of flood sand in the future and provide an important reference for making preventive preparations in advance. Afterwards, the results of the spatial analysis model and the time trend model are integrated, and the spatiotemporal change information of flood sand is obtained through comprehensive analysis, and an intelligent monitoring module is generated based on this. This module integrates the analysis results of the two dimensions of space and time, and can comprehensively and accurately reflect the actual situation of flood sand. Then, the intelligent monitoring module is embedded in the real-time monitoring system to realize real-time monitoring and analysis of flood sand in the target basin. The real-time monitoring system will quickly generate flood sand monitoring results based on the real-time data obtained and combined with the intelligent monitoring module to improve the efficiency and accuracy of flood sand monitoring.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0010] Figure 1 This is a structural schematic diagram of the secondary flood and sediment intelligent monitoring system based on multi-source data in this application;

[0011] Figure 2 This is a schematic diagram of the process of constructing a flood and sediment spatial analysis model for the secondary flood and sediment intelligent monitoring system based on multi-source data in this application.

[0012] Description of reference numerals:

[0013] Data collection unit 1, spatial data analysis unit 2, time series analysis unit 3, spatiotemporal analysis information parsing unit 4, monitoring and analysis unit 5. DETAILED DESCRIPTION

[0014] This application provides an intelligent monitoring system for floods and sediments based on multi-source data, which solves the technical problems of traditional monitoring systems, such as incomplete monitoring results and inability to respond to flood and sediment changes in a timely manner due to a single data source and lack of comprehensive spatiotemporal analysis. It achieves the technical effect of improving the accuracy, comprehensiveness and real-time nature of flood and sediment monitoring by integrating spatiotemporal analysis information and embedding it into a real-time monitoring system.

[0015] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0016] For examples, please see the attached Figure 1 The present application provides a secondary flood and sediment intelligent monitoring system based on multi-source data, and the system specifically includes the following modules:

[0017] The data collection unit is used to collect historical flood and sediment data of the target basin and construct a historical data set including flow, sediment content, rainfall, basin topography, and land use data.

[0018] Specifically, in the data collection unit, the system terminal collects flood events and related sediment data that have occurred in the target basin in the past. These data cover multiple key indicators during the flood period, including flow, sediment content, rainfall, basin topography, land use data, etc. Among them, flow refers to the volume of water flowing through the target basin per unit time, usually in cubic meters per second (m 3 / s) or liters per second (L / s). Flow is an important indicator for measuring the size of water flow and is of great significance for flood monitoring and prediction. Sediment content refers to the mass or volume ratio of sediment in a unit volume of water, usually expressed in kilograms per cubic meter (kg / m 3) or percentage (%). The level of sand content directly affects the erosion capacity of water flow, the sedimentation rate and the impact on downstream areas. Precipitation refers to the depth of rain falling on the target basin within a certain period of time, usually expressed in millimeters (mm). Precipitation is an important basis for assessing the possibility and scale of floods. Basin topography refers to the landform characteristics within the target basin, including natural geographical elements such as mountains, hills, plains, rivers, and lakes. Basin topography has an important impact on hydrological processes. For example, the slope, height and shape of the terrain will affect the convergence speed and path of rainfall, and then affect the formation and spread of floods. Land use data is information about land cover and utilization status, including the area, distribution and use of different land types (such as cultivated land, forest land, grassland, construction land, etc.). Land use data is of great significance for assessing flood risks. By collating and analyzing these historical data, the system terminal constructs a comprehensive historical data set. This historical data set stores rich information on flood sand phenomena in the target basin, providing a solid foundation for subsequent monitoring.

[0019] The spatial data analysis unit is used to perform spatial data analysis based on the historical data set and construct a flood and sediment spatial analysis model.

[0020] Specifically, in the spatial data analysis unit, the system terminal performs spatial data analysis on the collected historical data sets. Specifically, the system terminal performs terrain identification based on the basin topography in the historical data set, performs impact identification based on land use data, and performs spatial distribution identification based on flow, sediment content, and rainfall. Subsequently, based on these identification results, the system terminal trains and verifies the fully connected neural network to construct a water-sediment spatial analysis model. This model can simulate and predict the distribution and change trend of flood sand under different spatial conditions, helping to better understand the spatial characteristics of flood sand phenomena.

[0021] Further, if Figure 2 As shown, the present application provides a method for performing spatial data analysis based on the historical data set to construct a flood and sediment spatial analysis model, including:

[0022] According to the basin topography, the slope, slope direction, and basin boundary are identified to obtain the terrain characteristics; the land use data are used for classification to analyze the influence of different land use types on the spatial process of flood sand; according to the flow, sediment content, and rainfall, the spatial distribution relationship of flood sand distribution, concentration, and water flow path is fitted; the terrain characteristics, the influence of the spatial process of flood sand, and the spatial distribution relationship are used to train a fully connected neural network to construct the flood sand spatial analysis model.

[0023] In a preferred embodiment, in the process of constructing the flood and sand spatial analysis model, the system terminal extracts the basin terrain data, i.e., the basin terrain elevation data (DEM, digital elevation model) from the historical data set. DEM is a digital model representing ground elevation information, which exists in the form of a grid. Subsequently, using these DEM data, the system terminal extracts the basin boundary through a hydrological analysis method, which involves filling depression processing (removing local low-lying areas in DEM), flow direction analysis (determining the water flow direction of each grid), confluence accumulation calculation (calculating the water flow accumulation of each grid) and basin segmentation (determining the basin boundary according to the water flow direction and water flow accumulation). Through these steps, the boundary range of the basin can be accurately delineated. Afterwards, the slope of the basin terrain is calculated by calculating the ratio of the elevation difference between adjacent grids in the DEM to the horizontal distance, and then the slope direction of the basin terrain is determined by the direction in which the water flow leaves each grid. Then, the obtained slope, slope direction, and basin boundary are summarized to obtain the terrain characteristics.

[0024] The system terminal divides the land use data into different types according to the land classification standard, and obtains multiple land use types, including cultivated land, forest land, grassland, construction land, water area, unused land, etc. Among them, the land classification standard is determined based on the classification of the current land use status. For each land use type, its key features are extracted, such as vegetation coverage, soil type, etc. These features will be used for subsequent impact analysis. Subsequently, based on the characteristics of the land use type, the potential impact of different land use types on the spatial process of flood and sand, such as flood propagation speed, sediment deposition, erosion intensity, etc., is judged, for example, slowing down the flow rate, increasing soil permeability, etc.

[0025] The system terminal uses GIS (Geographic Information System) technology to overlay and analyze flood sand distribution data with flow and sediment content data. Analyze the distribution characteristics of flood sand under different flow and sediment content conditions, such as inundation range, deposition thickness, etc. Subsequently, construct the flood sand concentration field based on the sediment content data and the corresponding spatial location information. Then use interpolation methods, such as IDW, to generate continuous concentration distribution maps to show the spatial variation of flood sand concentration. After that, use the basin topography and water flow direction analysis tools to determine the water flow path in the basin. Combined with flow data and terrain characteristics, simulate the water flow path during the flood and analyze its impact on flood sand distribution and concentration. Then, according to the research purpose and data characteristics, select a suitable model to fit the spatial distribution relationship between flood sand distribution, concentration and water flow path. Common models include water-sediment relationship models, such as the exponential relationship model between suspended sediment concentration and flow, and water flow simulation models, such as the water flow path simulation model based on DEM. Then use the least squares method to estimate the model parameters. These parameters include the correlation coefficient between flow and sediment content, the resistance coefficient of the water flow path, etc.

[0026] After obtaining the influence of terrain characteristics and different land use types on the spatial process of flood sand, as well as the spatial distribution relationship of flood sand distribution, concentration and water flow path, the system terminal uses terrain characteristics, the influence of water and sand spatial process, and spatial distribution relationship to segment the data set and obtain training data set, verification data set, and test data set. Subsequently, the fully connected neural network is trained using the training data set, and the model is tested for performance verification using the verification data set and the test data set. After that, the system terminal outputs the fully connected neural network that meets the test objectives to obtain a flood sand spatial analysis model. This flood sand spatial analysis model can predict possible changes in future flood sand and provide strong support for flood prevention and disaster reduction and watershed management.

[0027] Furthermore, the present application provides a method of using the terrain features, the influence of the flood and sand spatial process, and the spatial distribution relationship to train a fully connected neural network to construct the flood and sand spatial analysis model, including:

[0028] The terrain features, land use type, flow, sand content, and precipitation are used as input data, and the input data are standardized to a uniform range; based on the input data, the influence of flood and sand spatial processes, and the spatial distribution relationship, the data set is segmented to obtain a training data set, a validation data set, and a test data set; a fully connected neural network is initialized, the network is trained using the training data set, and the parameter difference is calculated using a loss function; according to the parameter difference gradient of the loss function, the network weights are updated using an optimization algorithm until the parameter difference of the loss function reaches a reduction target; the model is performance verified using the validation data set and the test data set, and the model parameters are adjusted until the test objectives are met to obtain the flood and sand spatial analysis model, which is used to predict the spatial distribution and spatial changes of flood and sand processes.

[0029] In an optional implementation, in order to ensure the accuracy and efficiency of model training, the system terminal uses data such as terrain features, land use type, flow, sand content, precipitation, etc. as input data, and standardizes these data through the maximum-minimum method. This step is to convert all data into the same numerical range to eliminate the influence of different dimensions and magnitudes on data analysis and model training. Subsequently, based on these input data and their impact analysis and spatial distribution relationship on the flood sand spatial process, 70% of the input data and the corresponding flood sand spatial process impact and spatial distribution relationship are extracted to form a training data set, 50% is extracted from the remaining data as a verification data set, and the remaining data is used as a test data set. Then, according to the specific needs of flood sand spatial analysis, a fully connected neural network is initialized as the basic architecture of the flood sand spatial analysis model, including the number of neural network layers, the number of neurons in each layer, and the activation function, etc., and random numbers are used to determine the weights and biases, and the mean square error is used as the loss function. After completing the initialization of the model, the system terminal inputs the input data in the training data set into the neural network, and obtains the output prediction value through the calculation of each layer of neurons. The defined loss function is used to calculate the difference between the predicted value and the corresponding flood and sediment spatial process's influence and spatial distribution relationship. According to the gradient information of the loss function, the chain rule is used to calculate the gradient of each parameter in the network layer by layer. The gradient descent method is used as the optimization algorithm, and the weights and biases of the network are updated according to the gradient information to reduce the value of the loss function until the parameter difference of the loss function reaches the descent target. During the training process, the validation data set is used regularly to evaluate the model performance, and the model parameters such as learning rate and network structure are adjusted according to the validation results. After the training is completed, the final performance of the model is evaluated using the test data set to ensure that the model meets the test objectives, that is, the prediction accuracy. If not, the parameters of the model are further adjusted to improve the accuracy and generalization ability of the model. Otherwise, the current flood and sediment spatial analysis model is output to predict the spatial distribution and spatial changes of flood and sediment processes.

[0030] The time series analysis unit is used to perform time series analysis based on the historical data set and construct a flood sand trend analysis model.

[0031] Specifically, in the time series analysis unit, the system terminal performs time series analysis on the historical data set. This step aims to reveal the laws and trends of flood and sediment processes over time. Analysis methods include trend analysis, periodicity analysis, and autocorrelation analysis. Through these analyses, the temporal characteristics and dynamic behaviors of flood and sediment processes can be understood. Subsequently, based on the results of the time series analysis, a flood and sediment trend analysis model is constructed using the time series model to predict possible trends and changes in future flood and sediment processes.

[0032] Furthermore, the present application provides a method for performing time series analysis based on the historical data set to construct a flood sediment trend analysis model, including:

[0033] Obtain the time series data of the flow, sediment content, and rainfall; visualize the time series data of the flow, sediment content, and rainfall, and fit the change trend and periodic change characteristics; perform autocorrelation analysis based on the change trend and periodic change characteristics to identify the trend change relationship; use the time series model to perform trend prediction based on the trend change relationship, and construct the flood sediment trend analysis model.

[0034] In a preferred embodiment, in the process of constructing a flood sand trend analysis model, the system terminal collects time series data on flow, sand content and rainfall. These data change over time, record the specific values ​​at each time point, and provide a basis for analyzing the dynamic behavior of flood sand. In order to more intuitively understand the laws and characteristics of these data, the system terminal uses visualization technology to display these time series data. By drawing a time series graph, observe the changing trend of the data over time, for example, whether there is a trend of growth or decrease, and whether there are obvious periodic changes, such as seasonal fluctuations. This visual analysis helps to preliminarily grasp the overall characteristics of the data. On the basis of the visual analysis, the system terminal performs autocorrelation analysis. Autocorrelation analysis is a statistical method used to study the relationship between current values ​​and past values ​​in time series data. Through autocorrelation analysis, the trend change relationship in the data is identified, that is, how the current value is affected by the past value. This is crucial to understanding the dynamic behavior of the flood sand process. Subsequently, a suitable time series model is selected according to the obtained trend change relationship, such as an ARIMA model, a seasonal ARIMA model (SARIMA), an exponential smoothing model, etc. Then use the historical data set to estimate the parameters of the selected time series model, such as the p, d, and q parameters in the ARIMA model. The performance of the model is evaluated by residual analysis, model goodness of fit test and other methods, and the model is adjusted and optimized as necessary. After that, the determined time series model is used to predict the trend of flow, sediment content, rainfall, etc. in the future. The prediction results include the predicted value and its confidence interval to evaluate the uncertainty of the prediction. Then, the trend prediction results are combined with the physical mechanism of the flood and sand process to construct a flood and sand trend analysis model. Because of factors such as nonlinear relationships and hysteresis effects in the flood and sand process, the system terminal makes necessary corrections and improvements to the flood and sand trend analysis model. The final flood and sand trend analysis model can reflect the dynamic change characteristics of the flood and sand process and provide a scientific basis for flood prevention and disaster reduction.

[0035] Furthermore, the present application provides an autocorrelation analysis based on the change trend and periodic change characteristics, including:

[0036] Autocorrelation analysis is performed according to the change trend and periodic change characteristics, where the function expression is:

[0037]

[0038] In an optional implementation manner, the system terminal performs autocorrelation analysis on the change trend and periodic change characteristics based on an autocorrelation analysis function, and the autocorrelation analysis function is specifically as follows: Among them, X t is the tth observation of the time series. The time series consists of a series of data points arranged in chronological order, and each data point corresponds to a specific time point. is the average value of the time series, which is calculated by adding all the observations in the time series and then dividing by the total number of observations, that is, the length of the time series T. The average provides a measure of the central tendency of the time series. T is the length of the time series, that is, the total number of observations in the time series, which helps to understand the scale and scope of the time series. k is the lag order, which is a key parameter in autocorrelation analysis. The lag order refers to the number of time units that the past observations are pushed back when comparing the current observations with the past observations, that is, when calculating the correlation between the t-th observation and the tk-th observation, k is the time interval between the two observations. By changing the value of k, you can explore the autocorrelation at different time intervals.

[0039] The spatiotemporal analysis information parsing unit is used to integrate the analysis results of the flood sand spatial analysis model and the flood sand trend analysis model, parse the flood sand spatiotemporal analysis information, and generate an intelligent monitoring module.

[0040] Specifically, in the spatiotemporal analysis information parsing unit, the system terminal sets up a fusion decision layer to deeply integrate the analysis results of the flood sand spatial analysis model and the flood sand trend analysis model. Subsequently, the system terminal connects the flood sand spatial analysis model and the flood sand trend analysis model in parallel, connects the output layer of the flood sand spatial analysis model and the flood sand trend analysis model to the fusion decision layer, and encapsulates the connected results to build an intelligent monitoring module. This intelligent monitoring module can parse the spatiotemporal analysis information of flood sand and analyze the status of flood sand in the basin.

[0041] Furthermore, the present application provides an analysis module that integrates the analysis results of the flood sand spatial analysis model and the flood sand trend analysis model, analyzes the flood sand spatiotemporal analysis information, and generates an intelligent monitoring module, including:

[0042] A fusion decision layer is set up, and the analysis results of the flood sand spatial analysis model and the flood sand trend analysis model are fed back to the fusion decision layer; the flood sand location probability is obtained according to the flood sand spatial analysis model, and the predicted time and trend probability of flood sand events are obtained according to the flood sand trend analysis model, the flood sand location probability is fused and superimposed with the predicted time and trend probability of flood sand events, the state of flood sand in the basin is analyzed, and the intelligent monitoring module is generated.

[0043] In a preferred embodiment, in the process of building the intelligent monitoring module, the system terminal sets a fusion decision layer. The core function of the fusion decision layer is to comprehensively process the analysis results of the flood sand spatial analysis model and the flood sand trend analysis model, and is responsible for integrating scattered information into valuable decision-making basis. The flood sand spatial analysis model provides the system terminal with the probability of the location where flood sand may occur through the learned knowledge. These probability values ​​reflect the possibility of different regions being invaded by flood sand in the future, and are the quantitative expression of spatial distribution characteristics. At the same time, the flood sand trend analysis model predicts the probability of the occurrence time of flood sand events and their development trend based on time series analysis and statistical methods. These prediction information not only includes when flood sand may occur, but also reveals its possible change trend, such as increased flow and increased sand content, which is a manifestation of the dynamic change of time series. In the fusion decision layer, the system terminal fuses and superimposes the flood sand location probability with the predicted time and trend probability of flood sand events. This step combines the spatial possibility with the temporal prediction to form a complete spatiotemporal analysis framework. Through superposition analysis, it is possible to more accurately judge when and where flood sand events are most likely to occur, as well as the intensity of their development trend. Based on the results of fusion and superposition, the fusion decision layer conducts a comprehensive analysis of the state of flood sand in the basin. It comprehensively considers information from multiple dimensions such as spatial distribution, time prediction and trend changes, providing more accurate and comprehensive decision support for the final monitoring and analysis of the intelligent monitoring module.

[0044] The monitoring and analysis unit is used to fit the intelligent monitoring module to the real-time monitoring system, perform monitoring and analysis based on the real-time monitoring data of the target watershed, and obtain flood and sand monitoring results.

[0045] Specifically, in the monitoring and analysis unit, the system terminal fits the intelligent monitoring module into the real-time monitoring system. This process ensures that the intelligent monitoring module can receive, process and analyze the monitoring data from the target basin in real time, so as to quickly generate the monitoring results of flood sand. As the front end of data collection, the real-time monitoring system continuously obtains the latest monitoring data from various sensors, remote sensing equipment or other data sources in the target basin. These data cover key indicators such as water flow velocity, water level, and sand content, and are an important basis for evaluating the flood sand status. Once the real-time monitoring system collects the data, the intelligent monitoring module uses the flood sand spatial analysis model and combines the spatial location information in the real-time data to evaluate the locations and probabilities that may be affected by flood sand. Using the flood sand trend analysis model, the real-time data is analyzed in time series to predict the development trend and possible time of flood sand events. After obtaining these analysis results, the intelligent monitoring module merges and superimposes the flood sand location probability with the predicted time and trend probability to form a comprehensive assessment of the flood sand status of the basin. This process fully considers the complex changes in the flood sand phenomenon in space and time, and improves the accuracy and reliability of the monitoring results.

[0046] Furthermore, the present application provides the monitoring and analysis based on the real-time monitoring data of the target watershed to obtain flood and sand monitoring results, including:

[0047] A fixed time window and a dynamic time window are set, wherein the fixed time window is an equal time window and the dynamic time window is a non-equal time window; data segmentation is performed on the real-time monitoring data according to the fixed time window and the dynamic time window to obtain fixed window monitoring data and dynamic window monitoring data; the fixed window monitoring data and the dynamic window monitoring data are monitored and analyzed respectively by the intelligent monitoring module to obtain a first monitoring result and a second monitoring result; the flood and sand monitoring result is obtained according to the first monitoring result and the second monitoring result.

[0048] In a preferred embodiment, when setting the data processing strategy, the system terminal uses two methods, fixed time window and dynamic time window, to segment the real-time monitoring data. Fixed time window means that the duration of each window is equal, for example, every 5 minutes or every 30 minutes is a window. This method is suitable for situations where the data changes relatively smoothly. The dynamic time window is more flexible, and its duration is not fixed. It can be dynamically adjusted according to the actual situation of the data to better capture the changing characteristics in the data. Subsequently, the system terminal uses these two time windows to segment the real-time monitoring data respectively to obtain fixed window monitoring data and dynamic window monitoring data. These two sets of data represent the observation results of the watershed state at different time scales. Afterwards, the intelligent monitoring module will analyze these two sets of data respectively. For fixed window monitoring data, the analysis focuses on the stability and periodic changes of the data; while for dynamic window monitoring data, the analysis focuses more on the mutation points and abnormal behaviors in the data. Through these two different analysis methods, the system terminal obtains the first monitoring result and the second monitoring result. Then, the system terminal compares and integrates the two monitoring results to obtain more accurate and reliable flood and sand monitoring results.

[0049] Furthermore, the present application provides obtaining the flood sand monitoring result according to the first monitoring result and the second monitoring result, including:

[0050] Compare the deviation information of the first monitoring result and the second monitoring result; trace back and locate the monitoring data based on the deviation information, perform periodic stability evaluation based on the monitoring data, configure fusion weights, and the fusion weights correspond to the periodic stability evaluation; locate the first monitoring result and the second monitoring result according to the periodic fusion weights, perform result fusion evaluation, and obtain the flood and sand monitoring result.

[0051] In an optional embodiment, the system terminal compares the deviation information of the first monitoring result and the second monitoring result obtained by the fixed time window and the dynamic time window. This step is intended to evaluate the consistency of the monitoring results under the two time dimensions. If the two monitoring results show a high consistency, that is, the deviation information is less than or equal to the preset deviation information, then the system terminal determines that the data performance under the two time windows is similar, thereby enhancing the reliability of the results. In this case, the system terminal uses weighted averaging to fuse the two monitoring results to obtain a more comprehensive and robust flood sand monitoring result. This weighted averaging method can make full use of the information under the two time windows to improve the accuracy and credibility of the monitoring results. However, if there are significant differences between the two monitoring results, the system terminal locates and analyzes the reasons for the difference. By tracing back to the specific monitoring data, the periodic stability of these data is evaluated. The purpose of this step is to identify the periodic changes or abnormal fluctuations that may exist in the monitoring data, which may be caused by seasonal factors, human activities or other external conditions. Based on the results of the periodic stability evaluation, the system terminal configures corresponding fusion weights for the two monitoring results. These weights reflect the stability and reliability of the monitoring results under different time windows, as well as their contribution to the final monitoring results. When configuring weights, the system terminal gives higher weights to monitoring results that show higher periodic stability, because such results are more likely to reflect the true flood and sand conditions. Subsequently, the two monitoring results are located according to the configured fusion weights, and the result fusion evaluation is performed. This process integrates the effective information in the two monitoring results, while weakening or eliminating those deviations that may be caused by abnormal or unstable factors. Through such fusion evaluation, more accurate and reliable flood and sand monitoring results can be obtained.

[0052] In summary, the intelligent monitoring system for secondary flood and sediment based on multi-source data provided by this application has the following technical effects:

[0053] This application collects historical flood and sediment data of the basin through the data acquisition unit, and constructs a historical data set including flow, sediment content, rainfall, basin topography and land use. On this basis, the spatial data analysis unit performs spatial analysis on these data, identifies terrain features and analyzes the impact of land use types on the spatial process of flood sand, and constructs a flood sand spatial analysis model. In addition, the time series analysis unit performs time series analysis on the historical data set, fits the change trend and periodic change characteristics, and constructs a flood sand trend analysis model. Subsequently, the spatiotemporal analysis information parsing unit fuses the results of the above two analysis models, parses the spatiotemporal information of flood sand, and generates an intelligent monitoring module. The monitoring and analysis unit applies the intelligent monitoring module to the real-time monitoring system, analyzes the real-time monitoring data of the target basin, and obtains the flood sand monitoring results. The spatial analysis model is implemented through fully connected neural network training. The input data includes terrain features, land use types, flow, sediment content and precipitation, etc. After standardization and data set segmentation, the optimization algorithm is used to update the network weights until the model performance meets the test objectives. The time series analysis part identifies the trend change relationship through autocorrelation analysis and uses the time series model for trend prediction. In the spatiotemporal information analysis link, the probability of flood sand location is fused and superimposed with the event prediction time and trend probability through the fusion decision layer to generate an intelligent monitoring module to achieve a comprehensive analysis of the flood sand status in the basin. In the real-time monitoring process, fixed time windows and dynamic time windows are set to segment the real-time monitoring data, and the monitoring results are obtained through analysis by the intelligent monitoring module. By comparing the monitoring results of different time windows, the monitoring data is traced and located based on the deviation information, the periodic stability evaluation is performed and the corresponding fusion weights are configured, and the monitoring results are comprehensively evaluated to ensure the accuracy and reliability of flood sand monitoring. These technical effects jointly solve the technical problems that the traditional monitoring system has a single data source and lacks comprehensive spatiotemporal analysis, resulting in incomplete monitoring results and inability to respond to flood sand changes in a timely manner, and achieves the technical effect of improving the accuracy, comprehensiveness and real-time nature of flood sand monitoring by integrating spatiotemporal analysis information and embedding it into the real-time monitoring system.

[0054] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be 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 the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0055] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. The intelligent monitoring system for secondary flood and sediment based on multi-source data is characterized by: The secondary flood and sediment intelligent monitoring system based on multi-source data includes: Data collection unit, used to collect historical flood and sediment data of the target basin, and construct historical data sets including flow, sediment content, rainfall, basin topography, and land use data; A spatial data analysis unit, used to perform spatial data analysis based on the historical data set and construct a flood and sediment spatial analysis model; A time series analysis unit, used to perform time series analysis based on the historical data set and construct a flood sediment trend analysis model; A spatiotemporal analysis information parsing unit, used to integrate the analysis results of the flood sand spatial analysis model and the flood sand trend analysis model, parse the flood sand spatiotemporal analysis information, and generate an intelligent monitoring module; A monitoring and analysis unit, used to fit the intelligent monitoring module to the real-time monitoring system, perform monitoring and analysis based on the real-time monitoring data of the target watershed, and obtain flood and sediment monitoring results; Based on the historical data set, spatial data analysis is performed to build a flood and sediment spatial analysis model, including: According to the watershed topography, identifying the slope, slope direction, and watershed boundary, and obtaining the topographic features; Using the land use data to classify, analyze the impact of different land use types on the spatial process of flood and sediment; Fitting the spatial distribution relationship of flood sediment distribution, concentration and water flow path according to the flow, sediment content and rainfall; The flood and sand spatial analysis model is constructed by using the terrain characteristics, the influence of the flood and sand spatial process, and the spatial distribution relationship to perform fully connected neural network training.

2. The intelligent monitoring system for secondary flood and sediment based on multi-source data according to claim 1, characterized in that: The flood and sediment spatial analysis model is constructed by using the terrain characteristics, the influence of the flood and sediment spatial process, and the spatial distribution relationship to train a fully connected neural network, including: The terrain features, land use type, flow, sediment content, and precipitation are used as input data, and the input data are standardized to a uniform range; Based on the input data, the influence of the flood and sediment spatial process, and the spatial distribution relationship, the data set is segmented to obtain a training data set, a verification data set, and a test data set; Initializing a fully connected neural network, using the training data set to perform network training, and using a loss function to calculate parameter differences; According to the parameter difference gradient of the loss function, the optimization algorithm is used to update the network weights until the parameter difference of the loss function reaches the drop target; The model is subjected to performance verification test using the validation data set and the test data set, and the model parameters are adjusted until the test objectives are met, thereby obtaining the flood-sand spatial analysis model, which is used to predict the spatial distribution and spatial variation of flood-sand processes.

3. The intelligent monitoring system for secondary flood and sediment based on multi-source data according to claim 1, characterized in that: Based on the historical data set, time series analysis is performed to build a flood sediment trend analysis model, including: Obtaining time series data of flow, sediment content, and rainfall; Visualize the time series data of flow, sediment content and rainfall, and fit the change trend and periodic change characteristics; Performing autocorrelation analysis based on the change trend and periodic change characteristics to identify trend change relationships; The time series model is used to perform trend prediction based on the trend change relationship and construct the flood sand trend analysis model.

4. The intelligent monitoring system for secondary flood and sediment based on multi-source data according to claim 3 is characterized in that: Autocorrelation analysis is performed according to the change trend and periodic change characteristics, where the function expression is: ,in is the tth observation value of the time series, is the average value of the time series, T is the length of the time series, k is the lag order, is the degree of autocorrelation of the time series at lag order k, is the observed value of the time series at the lag k moment.

5. The intelligent monitoring system for secondary flood and sediment based on multi-source data according to claim 1, characterized in that: The analysis results of the flood sand spatial analysis model and the flood sand trend analysis model are integrated to analyze the flood sand spatiotemporal analysis information and generate an intelligent monitoring module, including: A fusion decision layer is set to feed back the analysis results of the flood sediment spatial analysis model and the flood sediment trend analysis model to the fusion decision layer; The flood sand location probability is obtained according to the flood sand spatial analysis model, the predicted time and trend probability of flood sand events are obtained according to the flood sand trend analysis model, the flood sand location probability is fused and superimposed with the predicted time and trend probability of flood sand events, the state of flood sand in the watershed is analyzed, and the intelligent monitoring module is generated.

6. The intelligent monitoring system for secondary flood and sediment based on multi-source data according to claim 1, characterized in that: The monitoring and analysis based on the real-time monitoring data of the target watershed to obtain flood and sand monitoring results includes: Setting a fixed time window and a dynamic time window, wherein the fixed time window is an equal time window and the dynamic time window is an unequal time window; The real-time monitoring data is segmented according to the fixed time window and the dynamic time window to obtain fixed window monitoring data and dynamic window monitoring data; The fixed window monitoring data and the dynamic window monitoring data are monitored and analyzed respectively by the intelligent monitoring module to obtain a first monitoring result and a second monitoring result; The flood sand monitoring result is obtained according to the first monitoring result and the second monitoring result.

7. The intelligent monitoring system for secondary flood and sediment based on multi-source data according to claim 6, characterized in that: The flood sand monitoring result is obtained according to the first monitoring result and the second monitoring result, including: Comparing deviation information between the first monitoring result and the second monitoring result; Based on the deviation information, the monitoring data is traced back and located, and a periodic stability evaluation is performed based on the monitoring data, and a fusion weight is configured, where the fusion weight corresponds to the periodic stability evaluation; The first monitoring result and the second monitoring result are located according to the periodic fusion weight, and the result fusion evaluation is performed to obtain the flood sand monitoring result.

Images