Hydrological early warning method and system, program product and readable storage medium

Abstract

The invention provides a hydrological early warning method and system, a program product and a readable storage medium, and relates to the field of hydrological data filling. By identifying missing data items in a hydrological table and analyzing the time span and spatial distribution of the data items, a foundation is laid for subsequent filling. According to the method, a comparison mechanism of a data volume ratio and a preset threshold value is introduced, full filling is executed when the missing data volume is small, and a selective filling strategy is adopted when the missing data volume is large. According to the dynamic adjustment method, estimation errors possibly introduced by large-scale filling are effectively avoided, and meanwhile filling of key data points is guaranteed. The inverse relation between the filling proportion and the data volume ratio avoids introduction of excessive estimation errors. The corresponding filling method is selected according to the data type, and the filling accuracy is improved. The completeness and reliability of the hydrological data are remarkably improved through the filling strategy with comprehensive consideration, and a high-quality data basis is provided for subsequent abnormal data recognition and early warning.

Description

Technical Field

[0001] The present application relates to the field of hydrological data filling, and in particular to a hydrological early warning method, system, program product and readable storage medium. Background Art

[0002] Hydrological monitoring in remote areas is often subject to great uncertainty due to factors such as remote geographical location and imperfect infrastructure. Specifically, the data collection points are unevenly distributed and the collection frequency is unstable, resulting in discontinuity of the hydrological data in time and space. Some areas or time periods have relatively abundant data, while other areas or time periods have scarce data.

[0003] In order to deal with this problem of data discontinuity, related technologies usually use data filling methods. All missing data are estimated and filled. Try to complete the entire data sequence. The filling amount is usually equal to the amount of missing data, that is, fill all missing data points as much as possible.

[0004] However, this comprehensive filling method has certain limitations in practical applications. When the amount of missing data is large, large-scale data filling may introduce too many estimation errors, causing the filled data to deviate from the actual hydrological situation. In particular, for a large amount of data that is continuously missing in time or space, subsequent analysis and decision-making based on these data lack accuracy. Summary of the invention

[0005] The present application provides a hydrological early warning method, system, program product and readable storage medium for avoiding estimation errors that may be introduced by large-scale filling.

[0006] In a first aspect, the present application provides a hydrological early warning method, comprising: filling hydrological data into a preset hydrological table, and identifying missing data items in the hydrological table; determining the data filling amount and filling area based on the time span and spatial distribution of the continuous missing data in the missing data items, wherein the continuous missing data is any data in the hydrological table that is continuously missing in time span or spatial distribution, wherein: when the ratio of the data amount of the continuous missing data to the corresponding total data amount is less than a preset threshold, performing a full filling measurement; when the ratio of the data amount of the continuous missing data to the corresponding total data amount is not less than a preset threshold, executing a selective filling strategy, selecting The filling strategy is to select several points in the front part, middle part and back part of the continuous missing data for filling; calculate the data filling amount according to the filling ratio and the data amount of the continuous missing data; the ratio of the data amount of the continuous missing data to the corresponding total data is inversely proportional to the filling ratio; select the corresponding data filling method according to the type of continuous missing data; transmit the filled hydrological data to the cloud server; enable the cloud server to identify the abnormal data in the filled hydrological data; receive the early warning signal sent by the cloud server, which is sent by the cloud server when the abnormal data exceeds the threshold; send early warning information to the terminal of relevant personnel.

[0007] By adopting the above technical scheme, the missing data items in the hydrological table are identified and their time span and spatial distribution are analyzed to lay the foundation for subsequent filling. Secondly, the method introduces a comparison mechanism between the data volume ratio and the preset threshold. When the amount of missing data is small, full filling is performed, and when the amount of missing data is large, a selective filling strategy is adopted. This dynamic adjustment method effectively avoids the estimation errors that may be introduced by large-scale filling, while ensuring the filling of key data points. In addition, the inverse relationship between the filling ratio and the data volume ratio avoids the introduction of too many estimation errors. Finally, the method selects the corresponding filling method according to the data type to improve the accuracy of filling. This comprehensive filling strategy significantly improves the integrity and reliability of hydrological data, and provides a high-quality data foundation for subsequent abnormal data identification and early warning.

[0008] In combination with some embodiments of the first aspect, in some embodiments, after the step of selecting a corresponding data filling method to fill the missing data according to the type of missing data, the method also includes: obtaining the filled hydrological data through a joint query of the database; applying a data fusion algorithm to process the filled hydrological data to generate an integrated data set; transmitting the filled hydrological data to a cloud server; and enabling the cloud server to identify abnormal data in the filled hydrological data. The steps specifically include: transmitting the integrated data set to the cloud server; and enabling the cloud server to identify abnormal data in the integrated data set.

[0009] By adopting the above technical solution, the padded hydrological data is obtained through joint query of the database, so that different types of hydrological data can be uniformly accessed and processed, overcoming the query efficiency problem caused by decentralized data storage. The padded hydrological data are processed by applying the data fusion algorithm to generate an integrated data set. The hydrological data from different sources and types are integrated to improve the consistency of the data, and the integrated data set is transmitted to the cloud server instead of directly transmitting the original padded data. This method significantly reduces the amount of data that needs to be transmitted and processed, and alleviates the computing load of the cloud server.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the step of obtaining the padded hydrological data through a database joint query specifically includes: creating a joint query, the joint query connecting the padded hydrological data in all databases through a common field, each database storing one type of padded hydrological data; extracting the padded hydrological data corresponding to the preset selection conditions according to the preset selection conditions in the joint query, the selection conditions including: specifying a time range, selecting specific hydrological parameters, and setting geographic location filtering conditions.

[0011] By adopting the above technical solution, a joint query is created to connect the filled hydrological data in all databases through common fields, and preset selection conditions are introduced in the joint query, including specifying the time range, selecting specific hydrological parameters, and setting geographic location screening conditions. It can quickly and accurately obtain key hydrological data within a specific time and space range, greatly improving the efficiency and pertinence of data analysis.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the step of applying a data fusion algorithm to process the padded hydrological data to generate an integrated data set specifically includes: standardizing different types of padded hydrological data in the same space and at the same time, so that the padded hydrological data of different dimensions and units are converted into a unified dimensionless numerical range; assigning probability values ​​to each subset in the identification framework for each type of padded hydrological data according to preset rules, and obtaining basic probability distribution of multiple evidence sources; the identification framework contains a set of mutually exclusive hypotheses related to hydrological warnings; obtaining a fused data set based on the basic probability distribution of multiple evidence sources, wherein: identifying all possible combinations of the target set obtained by intersecting multiple evidence sources; calculating the sum of the probability products of the possible combinations; applying a normalization factor to adjust the sum of the probability products; combining the fused data set with the padded hydrological data to generate an integrated data set.

[0013] By adopting the above technical scheme, different types of filled hydrological data are standardized, the problem of different dimensions and units is solved, and the foundation is laid for subsequent fusion. Secondly, a data fusion algorithm based on evidence theory is introduced to assign probability values ​​to each type of hydrological data, and construct the basic probability distribution of multiple evidence sources. This step fully considers the reliability and uncertainty of different data sources. Then, by identifying the intersection of multiple evidence sources, calculating the sum of probability products and applying normalization factors, the fused data set is finally obtained. It can effectively handle the uncertainty and conflict of data and significantly improve the reliability of the fusion results. Finally, the fused data set is combined with the original filled data, and the fused data set is combined with the original filled hydrological data, which not only retains the comprehensive advantages brought by fusion, but also retains the detailed information of the original data. This method effectively solves the problems of information loss and anomaly concealment that may be caused by relying solely on integrated data.

[0014] In combination with some embodiments of the first aspect, in some embodiments, according to the type of continuous missing data, the step of selecting a corresponding data filling method to fill the data specifically includes: identifying the type of continuous missing data and classifying it into time series data and spatial series data; for time series data, using time series analysis methods to fill the data; for spatial series data, using spatial statistics methods to fill the data.

[0015] By adopting the above technical solution, the types of continuous missing data are identified and classified into time series data and space series data. This classification lays the foundation for the subsequent selection of appropriate filling methods. Secondly, for time series data, time series analysis methods are used for filling, which makes full use of the time correlation of the data. For spatial series data, spatial statistical methods are applied for filling, taking into account the spatial distribution characteristics of the data. This strategy of using special filling methods for different types of data significantly improves the accuracy and effect of filling. By distinguishing between time and space series and applying corresponding statistical methods respectively, this technical solution can more accurately capture the temporal and spatial changes of data, thereby generating filling results that are more in line with the actual situation.

[0016] In combination with some embodiments of the first aspect, in some embodiments, for time series data, the step of filling in using a time series analysis method specifically includes: when the ratio of the amount of consecutive missing data to the corresponding total amount of data is less than a preset threshold, filling in using an interpolation method; when the ratio of the amount of consecutive missing data to the corresponding total amount of data is not less than a preset threshold, filling in using a time series decomposition method.

[0017] By adopting the above technical solution, a comparison mechanism between the data volume ratio and the preset threshold is introduced, which provides a basis for the selection of filling strategies in different situations. When the amount of continuous missing data is small, interpolation is used for filling. The interpolation method can quickly and effectively handle small-scale data missing and maintain the continuity of the data. Secondly, when the amount of continuous missing data is large, the method turns to time series decomposition for filling. Time series decomposition can capture the long-term trend, seasonal changes and random fluctuations of the data, and is particularly suitable for handling large-scale time series data missing. This strategy of dynamically selecting the filling method according to the scale of missing data ensures both the rapid processing of small-scale missing and the accuracy of filling large-scale missing.

[0018] In combination with some embodiments of the first aspect, in some embodiments, according to the type of continuous missing data, a corresponding data filling method is selected to perform filling steps, specifically including: for mixed time and space data, obtaining a set of known data points including time coordinates, space coordinates and observation values; based on time and space distance, selecting adjacent known data points for continuous missing data; using a predetermined time and space variation function to construct a group of Kriging equations; solving the group of Kriging equations to obtain a weight vector; using the weight vector and the observation values ​​of the known data points to calculate the filling value of the continuous missing data.

[0019] By adopting the above technical solution, a comprehensive data foundation is laid for subsequent analysis by obtaining a set of known data points containing time coordinates, spatial coordinates and observations. Secondly, adjacent known data points are selected based on the time and space distance to ensure that the filling process takes into account the local time and space correlation. Then, the Kriging equation group is constructed using the predetermined time and space variation function. This step makes full use of the advanced theory of geostatistics and can effectively capture the time and space variation characteristics of the data. The weight vector is obtained by solving the Kriging equation group, and the filling value is calculated using the weight vector and the observation value of the known data point, so that the continuous missing data can be accurately estimated. This spatiotemporal data filling method based on Kriging interpolation not only takes into account the spatial correlation of the data, but also takes into account the information of the time dimension, so it can generate more accurate and reliable filling results.

[0020] In a second aspect, the present application provides a hydrological warning model, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the hydrological warning model to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, the present application provides a computer program product comprising instructions, which, when run on a hydrological warning model, enables the hydrological warning model to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, the present application provides a computer-readable storage medium comprising instructions, which, when executed on a hydrological warning model, causes the hydrological warning model to execute the method described in the first aspect and any possible implementation of the first aspect.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. By identifying the missing data items in the hydrological table and analyzing their time span and spatial distribution, the foundation for subsequent filling is laid. Secondly, the method introduces a comparison mechanism between the data volume ratio and the preset threshold. When the amount of missing data is small, full filling is performed, and when the amount of missing data is large, a selective filling strategy is adopted. This dynamic adjustment method effectively avoids the estimation errors that may be introduced by large-scale filling, while ensuring the filling of key data points. In addition, the inverse relationship between the filling ratio and the data volume ratio avoids the introduction of too many estimation errors. Finally, the method selects the corresponding filling method according to the data type, which improves the accuracy of filling. This comprehensive filling strategy significantly improves the integrity and reliability of hydrological data, and provides a high-quality data foundation for subsequent abnormal data identification and early warning.

[0024] 2. Standardization of different types of filled hydrological data solves the problem of different dimensions and units, laying the foundation for subsequent fusion. Secondly, a data fusion algorithm based on evidence theory is introduced to assign probability values ​​to each type of hydrological data, and construct the basic probability distribution of multiple evidence sources. This step fully considers the reliability and uncertainty of different data sources. Then, by identifying the intersection of multiple evidence sources, calculating the sum of probability products and applying normalization factors, the fused data set is finally obtained. It can effectively handle the uncertainty and conflict of data and significantly improve the reliability of the fusion results. Finally, the fused data set is combined with the original filled data, and the fused data set is combined with the original filled hydrological data, which not only retains the comprehensive advantages brought by the fusion, but also retains the detailed information of the original data. This method effectively solves the problems of information loss and anomaly concealment that may be caused by relying solely on integrated data. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a flow chart of hydrological early warning in the embodiment of the present application; Figure 2 This is another schematic diagram of the hydrological warning process in the embodiment of the present application; Figure 3 yes Figure 2 A specific flow chart of step S201 in FIG. Figure 4 yes Figure 2 A specific flow chart of step S202; Figure 5 It is a schematic diagram of an exemplary hardware structure of the hydrological early warning model in the embodiment of the present application. DETAILED DESCRIPTION

[0026] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more listed items.

[0027] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0028] The hydrological early warning method in the embodiment of the present application is described below: See also Figure 1 , which is a flow chart of the hydrological early warning method in an embodiment of the present application.

[0029] S101, filling the hydrological data into a preset hydrological table, and identifying the missing data items in the hydrological table; Among them, hydrological data refers to various measurements and observation data related to hydrology, including precipitation, river flow, water level, etc. The preset hydrological table refers to a pre-designed table structure for recording and storing hydrological data.

[0030] It should be noted that the purpose of the hydrological table is to facilitate the identification of items that should have data but do not.

[0031] S102. Determine the data filling amount and filling area based on the time span and spatial distribution of the continuous missing data in the missing data items, wherein the continuous missing data is any data that is continuously missing in the time span or spatial distribution in the hydrological table, wherein: Among them, continuous missing data refers to the situation where multiple consecutive data points are missing in the time series or spatial distribution. The data filling amount refers to the number of data points that need to be supplemented. The filling area refers to the time period or geographical area where data filling is required.

[0032] It should be noted that the following description takes a single continuous missing data point as an example for detailed description. This process can be extended to all continuous missing data points in the data set, and each missing point follows the same processing flow.

[0033] S103, when the ratio of the amount of consecutive missing data to the corresponding total amount of data is less than a preset threshold, perform full filling measurement; The preset threshold refers to a preset ratio value, and the full filling measurement represents a strategy of filling all missing data points.

[0034] In some embodiments, the ratio of the amount of consecutive missing data to the total amount of data in the corresponding time or space range is calculated. If this ratio is less than a preset threshold, it means that the missing data is relatively small and will not seriously affect the overall data quality. In this case, the system chooses to perform a full filling measurement, that is, fill all the identified missing data points.

[0035] In some embodiments, linear interpolation or nearest neighbor filling may be used for this continuous missing data; It should be noted that there is a single missing data item in the table. This single point missing can be regarded as a special continuous missing data, and its time and space range is compressed to a specific time point and spatial position. For this single point missing situation, all filling measurements are also performed, which will not be repeated here.

[0036] S104, when the ratio of the amount of the continuous missing data to the corresponding total amount of data is not less than a preset threshold, a selective filling strategy is executed, wherein the selective filling strategy is to select a number of points in the front part, the middle part and the back part of the continuous missing data for filling; When the proportion of consecutive missing data is large, full filling may introduce too much uncertainty. The system judges whether the ratio of the amount of missing data to the total amount of data is not less than the preset threshold. If so, the system adopts a selective filling strategy. This strategy does not fill all missing points, but selects key points in different parts of the missing data for filling. The system usually selects some points in the front, middle and back of the missing data for filling. This method can reduce the potential errors caused by filling while maintaining the overall trend of the data.

[0037] In some embodiments, the continuous missing data segment is first divided into three equal parts: front, middle, and back. In each part, the system uses the calculated interval (obtained in step S105) to select the filling point. For the selected points, the system uses an interpolation method (such as linear interpolation or spline interpolation) to fill. Finally, the system smoothes the filled data to ensure the continuity of the overall trend, which is not limited here.

[0038] S105, calculating the data filling amount according to the filling ratio and the data amount of the continuous missing data; the ratio of the data amount of the continuous missing data to the corresponding total data amount is inversely proportional to the filling ratio; Among them, the filling ratio indicates the proportion of data points actually filled in the continuous missing data to the total amount of missing data, and the ratio of the amount of continuous missing data to the corresponding total amount of data indicates the severity of the missing data.

[0039] In some embodiments, the ratio of the amount of consecutive missing data to the total amount of data is considered. The larger the ratio, the more serious the missing data is, and the system will reduce the filling ratio accordingly to avoid introducing too much uncertainty. The system then calculates the specific amount of data filling based on this dynamically adjusted filling ratio and the total amount of consecutive missing data. This ensures that a more cautious filling strategy is adopted when the data missing is serious, and a more comprehensive filling can be performed when the data missing is minor.

[0040] S106, selecting a corresponding data filling method to fill in the missing data according to the type of the missing data; The type of continuous missing data indicates the nature of the missing data, such as time series data, spatial distribution data, or spatiotemporal mixed data. The data filling method refers to the specific algorithm or technology used to estimate and fill missing data.

[0041] In some embodiments, step S106 specifically includes: S501, identifying the type of continuous missing data, and classifying it into time series data and space series data; In some embodiments, it is possible to determine in the hydrological table whether the missing values ​​are distributed along the time dimension or the space dimension. For time series data, it is usually manifested as the missing of the same monitoring point in a continuous time period; for spatial series data, it is manifested as the missing of adjacent spatial locations at the same time. Then, the missing data is divided into corresponding categories based on these characteristics.

[0042] S502, for time series data, fill in the data using a time series analysis method; In some embodiments, a suitable time series analysis method is selected, such as moving average, exponential smoothing, ARIMA model or machine learning method. Then, the historical data before the missing data point is used to train or fit the selected model. Next, the trained model is used to predict or estimate the value of the missing time point. Finally, these predicted values ​​are filled into the missing positions of the original data set. This process takes into account the time dependency and trend characteristics of the data and can generate fill-in values ​​consistent with the original data pattern.

[0043] In some embodiments, step S502 specifically includes: S5021, when the ratio of the amount of consecutive missing data to the corresponding total amount of data is less than a preset threshold, using interpolation to fill in; In some embodiments, the ratio of the number of consecutive missing data points to the total number of data points is calculated. Then, this ratio is compared with a preset threshold. If the ratio is less than the threshold, it means that the missing data is relatively small, and the overall structure and trend of the data remain intact. In this case, interpolation is selected for filling. Interpolation uses known data points around the missing points to estimate the missing values. This method is simple and fast, and can maintain the continuity and smoothness of the data when there is less missing data.

[0044] S602: When the ratio of the amount of consecutive missing data to the corresponding total amount of data is not less than a preset threshold, time series decomposition is used for filling.

[0045] Among them, time series decomposition refers to the process of decomposing time series data into multiple components, usually including trend, seasonality and residual (random) components. The trend component indicates the long-term change direction of the data. The seasonal component reflects the cyclical change pattern of the data. The residual component represents random fluctuations that cannot be explained by trend and seasonality.

[0046] In some embodiments, this step is performed when it is determined that the proportion of missing data is large and is not suitable for simple interpolation. The complete historical data is decomposed into a time series and the data is divided into three parts: trend, seasonality, and residual. Then, these three parts are analyzed and modeled separately. For the trend component, regression methods may be used for prediction; for the seasonal component, historical seasonal patterns may be used for filling; for the residual component, statistical methods or simple interpolation may be used. Then, the predicted or estimated values ​​of these three parts are recombined to obtain a comprehensive estimate of the missing time points. Finally, these estimates are filled in the missing positions of the original data set. This method can better capture the long-term trends and cyclical characteristics of the data, and is particularly suitable for dealing with missing data for longer time periods.

[0047] It can be seen that the introduction of a comparison mechanism between the data volume ratio and the preset threshold provides a basis for the selection of filling strategies in different situations. When the amount of continuous missing data is small, interpolation is used for filling. The interpolation method can quickly and effectively handle small-scale data missing and maintain the continuity of the data. Secondly, when the amount of continuous missing data is large, the method turns to time series decomposition for filling. Time series decomposition can capture the long-term trend, seasonal changes and random fluctuations of the data, and is particularly suitable for handling large-scale time series data missing. This strategy of dynamically selecting the filling method according to the scale of missing data ensures both the rapid processing of small-scale missing and the accuracy of filling large-scale missing.

[0048] S503. For spatial sequence data, apply spatial statistical methods to fill in the gaps.

[0049] In some embodiments, an appropriate spatial interpolation method is selected, such as inverse distance weighted method (IDW), Kriging method or spline interpolation method. Then, the spatial distribution characteristics and correlation of known data points are analyzed. Next, based on the selected method and analysis results, the estimated values ​​of the missing positions are calculated. Finally, these estimated values ​​are filled into the missing positions of the original data set. This process takes into account the spatial dependence of the data and can generate filling values ​​that are spatially continuous and consistent with the surrounding observation points.

[0050] It can be seen that the types of continuous missing data are identified and classified into time series data and space series data. This classification lays the foundation for the subsequent selection of appropriate filling methods. Secondly, for time series data, time series analysis methods are used for filling, which makes full use of the time correlation of the data. For spatial series data, spatial statistical methods are used for filling, taking into account the spatial distribution characteristics of the data. This strategy of using special filling methods for different types of data significantly improves the accuracy and effect of filling. By distinguishing between time and space series and applying corresponding statistical methods respectively, this technical solution can more accurately capture the temporal and spatial changes of data, thereby generating filling results that are more in line with the actual situation.

[0051] In some other embodiments, after step S503, the method further includes: S504, for mixed time and space data, obtaining a set of known data points including time coordinates, space coordinates and observation values; Among them, spatiotemporal mixed data refers to data sets that have both temporal and spatial attributes, such as hydrological data measured at different times and locations. The time coordinate represents the observation time of the data point, such as a date or timestamp.

[0052] In some embodiments, all non-missing data points are screened out from the original data set. Then, the time information of each data point is extracted, which may include year, month, day, hour, etc., and converted into a unified time format. Next, the spatial information of each data point is extracted, usually the geographic coordinates (longitude and latitude) or coordinates in other spatial reference systems. Finally, the time coordinates, spatial coordinates, and corresponding observations are combined into a structured data set. This step lays the foundation for subsequent spatiotemporal analysis and interpolation, ensuring that all necessary information is properly collected and organized.

[0053] S505, selecting adjacent known data points for the continuous missing data based on the spatiotemporal distance; Among them, spatiotemporal distance refers to a distance metric that comprehensively considers the time and space dimensions, and is used to indicate the degree of proximity between two data points in time and space.

[0054] In some embodiments, it is necessary to define a suitable spatiotemporal distance metric, which typically involves a standardized and weighted combination of temporal and spatial distances. Then, for each missing data point, the spatiotemporal distance between it and all known data points is calculated. Next, the closest known data point is selected based on a pre-set criterion (such as a distance threshold or a fixed number of neighboring points). Finally, these selected neighboring points and their related information (such as coordinates and observations) are saved for subsequent interpolation. This process ensures that the most spatiotemporally relevant data points are used when interpolating, thereby improving the accuracy of the interpolation.

[0055] For example: first define the space-time distance function, standardize the time difference and space distance and then linearly combine them, then for each missing point, calculate its space-time distance with all known points, then use the quick sort algorithm to sort the distances, and then select the first N points with the smallest distance (N is the preset number of neighboring points), and finally store the information of these N points in a data structure, which is not limited here.

[0056] S506, constructing a Kriging equation group using a predetermined spatiotemporal variation function; Among them, the spatiotemporal variogram (also called the spatiotemporal covariance function or spatiotemporal semivariogram) is a mathematical function that describes the correlation structure of data in time and space. The Kriging equations are a set of linear equations used to solve the optimal weights in the Kriging interpolation method.

[0057] In some embodiments, it is necessary to ensure that there is a spatiotemporal variogram model suitable for the current data. Then, use this variogram to calculate the covariance between the missing point and each selected neighboring point, as well as the covariance between the neighboring points. Next, fill these covariance values ​​into the coefficient matrix of the Kriging equations. At the same time, construct the right-hand side vector of the equations, that is, the covariance between the missing point and each neighboring point. Finally, add the Lagrange multiplier term to ensure the unbiased estimation condition that the weight sum is 1. This process encodes the correlation structure of the spatiotemporal data into the equations, preparing for solving the optimal interpolation weights.

[0058] For example: select a suitable spatiotemporal variation function model (such as a product type or sum type model), then calculate the time distance and space distance for each pair of data points, then substitute these distances into the variation function to calculate the covariance value, and then fill the calculated covariance value into the corresponding positions of the matrix and vector, and finally add Lagrange multiplier rows and columns to complete the construction of the equation group.

[0059] S507, solving the Kriging equations to obtain a weight vector; In some embodiments, it is necessary to select a suitable numerical solution method. Considering that the Kriging equations are usually symmetric and positive, methods such as Gaussian elimination, LU decomposition or conjugate gradient method can be used. Then, the selected method is applied to solve the system of equations. During the solution process, numerical stability processing may be required, such as using rotation or scaling to improve the condition number of the matrix. Then, the solution obtained is verified to ensure that it meets the conditions of unbiasedness and minimum variance. Finally, the weight vector obtained by the solution is extracted to prepare for the next interpolation calculation. The key to this process is to solve the system of equations accurately and efficiently, because the accuracy of the weights directly affects the quality of the final interpolation result.

[0060] For example: using Gaussian elimination, first transform the augmented matrix into an upper triangular form, then solve the unknowns by back substitution, then perform a numerical stability check, if the result is unstable, use a partial principal component selection strategy to re-solve, and finally normalize the solution to ensure that the sum of the weights is 1, which is not limited here.

[0061] It can be seen that by obtaining a set of known data points containing time coordinates, spatial coordinates and observations, a comprehensive data foundation is laid for subsequent analysis. Secondly, adjacent known data points are selected based on the time and space distance to ensure that the filling process takes into account the local time and space correlation. Then, the Kriging equation group is constructed using the predetermined time and space variation function. This step makes full use of the advanced theory of geostatistics and can effectively capture the time and space variation characteristics of the data. The weight vector is obtained by solving the Kriging equation group, and the filling value is calculated using the weight vector and the observation value of the known data point, so that the continuous missing data can be accurately estimated. This spatiotemporal data filling method based on Kriging interpolation not only considers the spatial correlation of the data, but also considers the information of the time dimension, so it can generate more accurate and reliable filling results.

[0062] S505. Calculate the filling value of the continuous missing data using the weight vector and the observed values ​​of the known data points.

[0063] The weight vector is the optimal interpolation coefficient set obtained by solving the Kriging equations. The known data points refer to the neighboring data points selected in step S502. The observed values ​​are the actual measured values ​​of these known data points. The filled values ​​refer to the estimated values ​​used to replace the missing data.

[0064] In some embodiments, each weight in the weight vector is multiplied by the observed value of the corresponding known data point. Then, all these products are added together to obtain an estimate of the missing point. Then, some post-processing may be required, such as checking whether the estimated value is within a reasonable range, or applying some constraints (such as non-negativity). Finally, the calculated fill-in value is inserted into the missing position of the original data set. This process implements the estimation of missing values ​​based on a weighted average of surrounding known points, where the weight reflects the relative influence of each known point on the missing point.

[0065] It can be seen that by obtaining a set of known data points containing time coordinates, spatial coordinates and observations, a comprehensive data foundation is laid for subsequent analysis. Secondly, adjacent known data points are selected based on the time and space distance to ensure that the filling process takes into account the local time and space correlation. Then, the Kriging equation group is constructed using the predetermined time and space variation function. This step makes full use of the advanced theory of geostatistics and can effectively capture the time and space variation characteristics of the data. The weight vector is obtained by solving the Kriging equation group, and the filling value is calculated using the weight vector and the observation value of the known data point, so that the continuous missing data can be accurately estimated. This spatiotemporal data filling method based on Kriging interpolation not only considers the spatial correlation of the data, but also considers the information of the time dimension, so it can generate more accurate and reliable filling results.

[0066] S107, transmitting the padded hydrological data to the cloud server; allowing the cloud server to identify abnormal data in the padded hydrological data; The padded hydrological data refers to the hydrological data set processed and padded in the above steps. The cloud server refers to a high-performance computing device running in a remote data center.

[0067] In some embodiments, the cloud server may apply a box plot method to identify outliers and may also apply a moving average or exponential smoothing technique to compare actual values ​​with predicted values, which is not limited here.

[0068] S108, receiving a warning signal sent by the cloud server, where the warning signal is sent by the cloud server when abnormal data exceeds a threshold; When the cloud server completes the identification of abnormal data, if it finds that the abnormal situation reaches the warning standard, it needs to notify the relevant systems or personnel in time.

[0069] In some embodiments, the cloud server uses a multi-level warning mechanism, which is divided into different levels (such as mild, moderate, and severe) according to the severity of the abnormal data. When an abnormality is detected, the server first determines the warning level, and then generates a structured warning signal containing information such as the warning level, abnormality type, and impact range.

[0070] S109. Send warning information to terminals of relevant personnel.

[0071] Among them, relevant personnel refer to staff, decision makers or the public who need to receive hydrological warning information. Terminal refers to the device that receives warning information, such as a mobile phone, computer or dedicated warning device. Warning information is generated based on the warning signal and contains detailed information such as specific warning content, response measures and suggestions.

[0072] In some embodiments, first, a more detailed and easy-to-understand warning message is generated based on the content and severity of the warning signal. This information may include a description of the abnormal situation, possible impacts, recommended measures, etc. Then, the system will determine the list of relevant personnel who need to receive this information according to the pre-set distribution rules. The system then sends the warning information to the terminal devices of these personnel through various channels (such as SMS, email, dedicated app push, etc.).

[0073] It can be seen that the hydrological early warning method provided by the present application can flexibly select the filling strategy according to the characteristics of the missing data. First, the method identifies the missing data items in the hydrological table and analyzes their time span and spatial distribution to lay the foundation for subsequent filling. Secondly, the method introduces a comparison mechanism between the data volume ratio and the preset threshold. When the amount of missing data is small, full filling is performed, and when the amount of missing data is large, a selective filling strategy is adopted. This dynamic adjustment method effectively avoids the estimation errors that may be introduced by large-scale filling, while ensuring the filling of key data points. In addition, the inverse relationship between the filling ratio and the data volume ratio avoids the introduction of too many estimation errors. Finally, the method selects the corresponding filling method according to the data type to improve the accuracy of filling. This comprehensive filling strategy significantly improves the integrity and reliability of hydrological data, and provides a high-quality data foundation for subsequent abnormal data identification and early warning.

[0074] However, in actual use, the diversity of hydrological data will affect the processing efficiency of cloud servers. Each type of hydrological data may have different temporal resolution and spatial distribution characteristics. The multi-dimensional nature of this data may cause cloud servers to face problems such as excessive computing load and extended response time when processing large amounts of data.

[0075] See also Figure 2 , Figure 2 This is another flow chart of hydrological early warning in the embodiment of the present application; in some embodiments, after step S106, it also includes: S201, obtaining the filled hydrological data through database joint query; A database joint query refers to a query operation that retrieves data from multiple related database tables at the same time.

[0076] In some embodiments, each type of hydrological data is stored in a separate dedicated database, or, in the same database, logical partitions or table spaces are created for different types of hydrological data; In some embodiments, the required information is extracted from multiple database tables that may store the filled hydrological data by using SQL language or other database query tools.

[0077] See also Figure 3 , Figure 3 yes Figure 2 A specific flow chart of step S201 in FIG. 1 is shown in FIG. 1 ; in some embodiments, step S201 specifically includes: S2011, creating a joint query, the joint query connects the padded hydrological data in all databases through a common field, and each database stores one type of padded hydrological data; Among them, common fields refer to fields with the same meaning and data type in different databases or tables, which are used to establish associations between data.

[0078] In some embodiments, all databases to be queried are determined, and then common fields in each database for association are identified, such as timestamp, geographic coordinates, or site IDs. Next, an SQL query statement is constructed or a database management tool is used to connect the filled hydrological data in these databases through a JOIN operation. This joint query allows comprehensive information from multiple data sources to be obtained in one operation.

[0079] S2012. Extracting the filled hydrological data corresponding to the preset selection conditions according to the preset selection conditions in the joint query. The selection conditions include: specifying a time range, selecting specific hydrological parameters, and setting geographic location screening conditions.

[0080] Preset selection conditions refer to predefined criteria or rules for filtering data. Specifying a time range means selecting data within a specific time period. Specific hydrological parameters refer to specific hydrological indicators, such as water level and flow. Geographic location filtering conditions are used to select data based on spatial information, such as a specific watershed or monitoring site.

[0081] In some embodiments, specific selection conditions are determined according to user needs or analysis purposes, including time range (such as the past week or a hydrological year), hydrological parameters of interest (such as daily average water level or instantaneous flow), and geographic locations (such as a river or a lake). These conditions are then converted into filter statements in the query language, such as a WHERE clause or a HAVING clause, and added to the previously created joint query. Finally, the modified query is executed to extract only the filled hydrological data that meets all the preset conditions.

[0082] It can be seen that the creation of a joint query connects the hydrological data filled in all databases through common fields, and introduces preset selection conditions in the joint query, including specifying the time range, selecting specific hydrological parameters, and setting geographic location screening conditions. It can quickly and accurately obtain key hydrological data within a specific time and space range, greatly improving the efficiency and pertinence of data analysis.

[0083] S202, applying a data fusion algorithm to process the filled hydrological data to generate an integrated data set; In some embodiments, a data fusion algorithm, such as Kalman filtering, Bayesian inference or deep learning method, is selected to process the padded hydrological data.

[0084] Step S107 is replaced by S203, transmitting the integrated data set to the cloud server; and enabling the cloud server to identify abnormal data in the integrated data set.

[0085] It can be seen that by obtaining the filled hydrological data through joint query of the database, different types of hydrological data can be uniformly accessed and processed, overcoming the query efficiency problem caused by decentralized data storage. The data fusion algorithm is applied to process the filled hydrological data to generate an integrated data set. The hydrological data from different sources and types are integrated to improve the consistency of the data, and the integrated data set is transmitted to the cloud server instead of directly transmitting the original filled data. This method significantly reduces the amount of data that needs to be transmitted and processed, and alleviates the computing load of the cloud server.

[0086] However, in actual use, relying solely on the integrated data set may cause data quality problems, which in turn affects the cloud server's ability to identify abnormal data. For example, the data integration process may cause some detailed information in the original hydrological data to be lost or distorted, affecting the accuracy of the integrated data set. During the data fusion process, some abnormal features may be averaged or masked, making the anomalies in the integrated data set less obvious than in the original data. The errors of various types of hydrological data during the integration process may accumulate and amplify, resulting in a decrease in the overall accuracy of the integrated data set. The temporal and spatial resolutions of different types of hydrological data may be unified during the integration process, resulting in the loss of some important local or short-term change features.

[0087] See also Figure 4 , Figure 4 yes Figure 2 A specific flow chart of step S202 in FIG. 1 is a schematic diagram of a specific flow chart of step S202 in FIG. 1 ; in some embodiments, step S202 specifically includes: S2021. Standardize the filled hydrological data of different types in the same space and at the same time, so that the filled hydrological data of different dimensions and units are converted into a unified dimensionless value range; A dimensionless numerical range is a pure numerical range without specific physical units, usually from 0 to 1 or -1 to 1.

[0088] In some embodiments, the type of hydrological data that needs to be standardized is determined, and for each type of data, the maximum and minimum values ​​of each type of data are extracted from the historical data. The minimum-maximum standardization method is applied. This method uses a linear transformation to map the original data to a specified new interval. Each original data point is converted to a standardized value through the minimum-maximum standardization method. This process maps all data points to a range of 0 to 1. It is achieved that different types of hydrological data are unified to the same scale. For example, rainfall and flow data that originally had different units are now converted to the same standardized range.

[0089] S2022. According to preset rules, a probability value is assigned to each subset in the identification framework for each type of filled hydrological data, and a basic probability assignment of multiple evidence sources is obtained; the identification framework includes a set of mutually exclusive hypotheses related to hydrological warning; the identification framework represents a set used to describe all possible hypotheses in evidence theory. Basic probability assignment refers to a function value assigned to each subset in the identification framework in evidence theory, indicating the support for the subset. A mutually exclusive set of hypotheses refers to a set of mutually exclusive possible situations or results.

[0090] In some embodiments, a recognition framework = {normal, mild attention, moderate attention, severe attention, extreme attention} is first formulated; According to pre-set rules (which may be based on expert knowledge or historical data analysis), the support for each hypothesis is calculated for each type of standardized hydrological data (such as water level, flow, rainfall). This process converts different types of hydrological data into a series of probability distributions, each of which reflects the judgment of different warning levels for that type of data.

[0091] For example: if the water level is 0-2 meters: m({Normal}) = 0.8, m({Normal, Mild Caution}) = 0.2; If the water level is between 2 and 4 meters: m({mild caution}) = 0.6, m({mild caution, moderate caution}) = 0.4; If the water level is between 4 and 6 meters: m({moderate caution}) = 0.7, m({moderate caution, severe caution}) = 0.3; If the water level exceeds 6 meters: m({Severe Caution, Extraordinary Caution}) = 1.0; S2023, obtaining a fused data set based on basic probability distribution of multiple evidence sources, wherein: identifying all possible combinations of the target set obtained by intersecting multiple evidence sources; calculating the sum of probability products of the possible combinations; and applying a normalization factor to adjust the sum of probability products; The steps first determine all target sets for which fusion probabilities need to be calculated, including single hypotheses and possible combinations of hypotheses. Then, for each target set, identify all possible combinations of evidence sources that may lead to the set. Next, calculate the sum of the probability products of these combinations, which represents the combined support of all evidence for the target set. Finally, adjust these probability sums by normalization factors to ensure that the final probability distribution satisfies the probability axioms. This process combines information from different hydrological data sources to obtain a more comprehensive and reliable warning probability distribution.

[0092] For example: m(A)=K*Σ(m1(B)*m2(C)), B∩C=A In the formula, m(A) is the fused combination of evidence source m1(B) and evidence source m2(C), m1(B) is one evidence source, such as water level data, m2(C) is another evidence source, such as rainfall data, and K is the normalization factor; Continuing with the previous example, we need to fuse water level data (m1) and rainfall data (m2): m1({mild alert})=0.6, m1({mild alert, moderate alert})=0.4 m2({normal, mild alert})=0.3, m2({mild alert, moderate alert})=0.7 Applying the above rules we get .

[0093] m({mild alert})=0.55, m({moderate alert})=0.28, m({mild alert, moderate alert})=0.17 S2024. Combine the fused data set with the filled hydrological data to generate an integrated data set.

[0094] It should be noted that the fused dataset is mainly used as a reference to provide preliminary indications of potential anomalies to the cloud server, rather than as a basis for final decision-making.

[0095] It can be seen that the standardization of different types of filled hydrological data solves the problem of different dimensions and units, laying the foundation for subsequent fusion. Secondly, a data fusion algorithm based on evidence theory is introduced to assign probability values ​​to each type of hydrological data, and construct the basic probability distribution of multiple evidence sources. This step fully considers the reliability and uncertainty of different data sources. Then, by identifying the intersection of multiple evidence sources, calculating the sum of probability products and applying normalization factors, the fused data set is finally obtained. It can effectively handle the uncertainty and conflict of data and significantly improve the reliability of the fusion results. Finally, the fused data set is combined with the original filled data, which not only retains the comprehensive advantages brought by the fusion, but also retains the detailed information of the original data. This method effectively solves the problems of information loss and anomaly concealment that may be caused by relying solely on integrated data.

[0096] The following introduces an exemplary hydrological warning model 500 provided in an embodiment of the present application. Figure 5 It is a schematic diagram of an exemplary hardware structure of the hydrological warning model 500 provided in an embodiment of the present application.

[0097] In some embodiments, the hydrological early warning model 500 is a computer device or the hydrological early warning model 500 includes a computer device. The computer device includes a processor, a memory and a network interface connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with other external terminals or servers through a network connection. In some embodiments, the network interface can be a wired network interface, and in some embodiments, the network interface can also be a wireless network interface. When the computer program is executed by the processor, the method in the embodiment of the present application is implemented.

[0098] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0099] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0100] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0101] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media integration. The available medium can be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk), etc.

[0102] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A hydrological early warning method, characterized in that: include: Filling the hydrological data into a preset hydrological table and identifying missing data items in the hydrological table; Based on the time span and spatial distribution of the continuous missing data in the missing data items, the data filling amount and filling area are determined, wherein the continuous missing data is any data in the hydrological table that is continuously missing in time span or spatial distribution, wherein: When the ratio of the amount of the consecutive missing data to the corresponding total amount of data is less than a preset threshold, performing full filling measurement; When the ratio of the amount of the continuous missing data to the corresponding total amount of data is not less than the preset threshold, a selective filling strategy is executed, wherein the selective filling strategy is to select a number of points in the front part, the middle part and the back part of the continuous missing data for filling; The data filling amount is calculated according to the filling ratio and the data amount of the continuous missing data; the ratio of the data amount of the continuous missing data to the corresponding total data is inversely proportional to the filling ratio; According to the type of the continuous missing data, select a corresponding data filling method to fill it; Transmitting the padded hydrological data to a cloud server; allowing the cloud server to identify abnormal data in the padded hydrological data; receiving a warning signal sent by the cloud server, wherein the warning signal is sent by the cloud server when the abnormal data exceeds a threshold; Send warning information to the terminals of relevant personnel.

2. The method according to claim 1, characterized in that After the step of selecting a corresponding data filling method to fill the missing data according to the type of the missing data, the method further comprises: Obtain the filled hydrological data through database joint query; Apply data fusion algorithm to process the filled hydrological data to generate integrated data sets; The step of transmitting the padded hydrological data to a cloud server and enabling the cloud server to identify abnormal data in the padded hydrological data specifically includes: The integrated data set is transmitted to a cloud server; and the cloud server is enabled to identify abnormal data in the integrated data set.

3. The method according to claim 2, characterized in that The step of obtaining the filled hydrological data through database joint query specifically includes: Creating a joint query, wherein the joint query connects the padded hydrological data in all databases through a common field, each of the databases storing one type of the padded hydrological data; In the joint query, the filled hydrological data corresponding to the preset selection conditions are extracted according to the preset selection conditions, and the selection conditions include: specifying a time range, selecting specific hydrological parameters, and setting geographic location screening conditions.

4. The method according to claim 3, characterized in that The step of applying the data fusion algorithm to process the filled hydrological data to generate an integrated data set specifically includes: Standardizing the padded hydrological data of different types in the same space and at the same time, so that the padded hydrological data of different dimensions and units are converted into a unified dimensionless numerical range; Assigning probability values ​​to each subset in the identification framework for each type of the padded hydrological data according to preset rules, and obtaining basic probability distributions for multiple evidence sources; the identification framework includes a set of mutually exclusive hypotheses related to hydrological warnings; A fused data set is obtained based on basic probability distribution of multiple evidence sources, wherein: all possible combinations of the multiple evidence sources intersecting to obtain a target set are identified; the sum of probability products of the possible combinations is calculated; and the sum of the probability products is adjusted by applying a normalization factor; The fused data set is combined with the padded hydrological data to generate the integrated data set.

5. The method according to claim 1, characterized in that The step of selecting a corresponding data filling method to fill the missing data according to the type of the missing data specifically includes: Identify the type of the continuous missing data and classify it into time series data and space series data; For time series data, time series analysis methods are used for filling; For spatial series data, spatial statistical methods are used for filling.

6. The method according to claim 5, characterized in that The step of filling the time series data using the time series analysis method specifically includes: When the ratio of the amount of the consecutive missing data to the corresponding total amount of data is less than a preset threshold, interpolation is used to fill the missing data; When the ratio of the amount of the continuous missing data to the corresponding total amount of data is not less than a preset threshold, time series decomposition is used for filling.

7. The method according to claim 1, characterized in that The step of selecting a corresponding data filling method to fill the missing data according to the type of the missing data specifically includes: For mixed spatiotemporal data, obtain a set of known data points containing time coordinates, spatial coordinates, and observation values; Selecting adjacent known data points for the consecutive missing data based on the temporal and spatial distances; Using the predetermined spatiotemporal variogram, the Kriging equations are constructed; Solving the Kriging equations to obtain a weight vector; The weight vector and the observed values ​​of the known data points are used to calculate the filling values ​​of the continuous missing data.

8. A hydrological early warning model, characterized in that: The hydrological warning model includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the hydrological warning model to execute the method described in any one of claims 1-7.

9. A computer program product comprising instructions, characterized in that When the computer program product runs on a hydrological warning model, the hydrological warning model is enabled to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium comprising instructions, characterized in that: When the instruction is executed on the hydrological warning model, the hydrological warning model is caused to execute the method according to any one of claims 1 to 7.

Images