Hydrological forecast correlation early warning method based on multiple factors

Abstract

The invention relates to the technical field of hydrological forecasting, in particular to a hydrological forecasting correlation early warning method based on multiple factors, which comprises the following steps: acquiring historical basic parameters of a to-be-forecasted site and historical basic parameters of other water conservancy projects in the same drainage basin, inputting the historical basic parameters of the to-be-forecasted site and the historical basic parameters of other water conservancy projects in the same drainage basin into a preset multi-factor associated site acquisition model, and acquiring associated sites of the to-be-forecasted site; real-time basic parameters of a to-be-forecasted site and real-time basic parameters of an associated site of the to-be-forecasted site are obtained, data preprocessing is carried out on the real-time basic parameters, and preprocessed data are obtained; inputting the preprocessed data into a trained forecasting model, and obtaining forecasting data of the to-be-forecasted site; comparing the forecast data of the to-be-forecasted site with the corresponding data threshold value, and issuing early warning information when the forecast data exceeds the data threshold value; according to the invention, the efficiency and accuracy of hydrological forecasting in the same drainage basin are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrological forecasting, and in particular to a hydrological forecasting association warning method based on multiple factors. Background Art

[0002] At present, my country's flood control and dispatching work is characterized by refinement and comprehensiveness. Hydrological forecasts can provide important reference information for my country's current disaster reduction work such as flood control, drought resistance and moisture prevention. Its accuracy directly affects whether flood control and dispatching can achieve the greatest economic and social benefits. In practical problems, the hydrological system is a complex system. This characteristic is reflected in the influence of various factors such as rainfall, evaporation, temperature, and vadose zone on the runoff process. How to correctly select these input factors will directly affect the forecast effect and forecast cost. Insufficient consideration of influencing factors will lead to insufficient forecast accuracy. Excessive selection of influencing factors will lead to long modeling time, significant increase in model complexity, and no obvious improvement or even decline in model forecast accuracy.

[0003] Generally, in the analysis process of hydrological regression forecasting, only the dimension of the influencing factor space is reduced to simplify the budget and improve efficiency while ensuring accuracy. The hydrological forecast model needs to consider multiple influencing factors at the same time. Although the traditional hydrological forecast model has taken into account factors such as precipitation, evaporation, and infiltration, it does not consider the impact of important water conservancy projects such as associated rain gauges and reservoirs in the basin on the formulation of forecast plans, as well as the relationship between the operation of important water conservancy projects such as associated rain gauges and reservoirs, resulting in low forecast accuracy.

[0004] In addition, my country's current water conservancy infrastructure is relatively weak, and it is imperative to develop modern water conservancy construction. Informatization is the only way to transform traditional water conservancy into modern water conservancy. With the rapid development of national information technology, the use of emerging information technology such as "Internet +" and "big data" to enhance government management and supervision capabilities has become a trend of the times and social requirements. However, the level of informatization development in the water conservancy industry is low and far from sufficient.

[0005] Therefore, there is an urgent need for a hydrological forecast correlation warning method based on multiple factors. Summary of the invention

[0006] 1. Technical issues to be resolved In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a hydrological forecast association warning method based on multiple factors, which solves the technical problems of slow data processing efficiency and low accuracy in the prior art.

[0007] (II) Technical solution In order to achieve the above object, the main technical solutions adopted by the present invention include: In a first aspect, an embodiment of the present invention provides a hydrological forecast association warning method based on multiple factors, comprising: S100, obtaining historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin, inputting the historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin into a pre-set multi-factor associated site acquisition model, and obtaining associated sites of the site to be predicted; The other water conservancy projects are water conservancy projects that can obtain flow data and rainfall data; S200, obtaining real-time basic parameters of the site to be forecasted and real-time basic parameters of the sites associated with the site to be forecasted, and performing data preprocessing on them to obtain preprocessed data; S300, inputting the preprocessed data into the trained forecast model to obtain forecast data of the site to be forecasted; comparing the forecast data of the site to be forecasted with the corresponding data threshold, and issuing warning information when the data threshold is exceeded; The basic parameters include: flow data and rainfall data; The pre-set multi-factor associated site acquisition model includes: a correlation coefficient acquisition unit, a Copula entropy acquisition unit, a histogram intersection value acquisition unit and an associated site acquisition unit.

[0008] Optionally, the S300 includes: S310, inputting the preprocessed data into the correlation coefficient acquisition unit, the Copula entropy acquisition unit and the histogram intersection value acquisition unit respectively to acquire the corresponding correlation coefficient, Copula entropy and histogram intersection value; S320, inputting the correlation coefficient, Copula entropy and histogram intersection value into an associated site acquisition unit to obtain a comprehensive correlation index; S330, sorting the comprehensive correlation indicators, and obtaining the top three water conservancy projects as associated sites of the site to be forecasted.

[0009] Optionally, the S310 includes: S311. Input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the Pearson correlation coefficient: ; in, x i is the flow data of the station to be predicted at a certain time. is the average flow data of the site to be predicted at that time, y i is the flow data of other water conservancy projects in the same basin at that time, is the average flow data of other water conservancy projects in the same basin at that time, z i is the rainfall data of the forecast station at that time, is the average rainfall data of the forecast station at that time, w i is the rainfall data of other water conservancy projects in the same basin at that time, is the average rainfall data of other water conservancy projects in the same basin at that time, n is the number of basic parameters obtained at that time, K 1 is the weight of traffic data, K 2 is the weight of rainfall data.

[0010] Optionally, the S310 further includes: S312. Input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the Copula entropy: ; in, c(u, v, z, w) For four variables u, v, z, w The probability density function of u is the flow data of the station to be predicted at a certain time. v is the rainfall data of the forecast station at a certain time. z is the flow data of other water conservancy projects in the same basin at that time, w It is the rainfall data of other water conservancy projects in the same basin at that moment.

[0011] Optionally, the S310 further includes: S313, input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the histogram intersection value: ; in, k is the number of intervals of the histogram, h Ai For the site to be forecasted, i The frequency in the interval, h Bi For other water conservancy projects in the same basin i The frequency in the interval, min( h Ai ，h Bi ) is in the i The minimum value of the two histograms in the interval.

[0012] Optionally, the S320 includes: S321, input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain a comprehensive correlation index: ; in, S is the correlation coefficient, They are the weights corresponding to the Pearson correlation coefficient, Copula entropy and histogram intersection values, r is the Pearson correlation coefficient, H is the Copula entropy, H max is the maximum value of Copula entropy, H min is the minimum value of Copula entropy, I is the histogram intersection value, I max is the maximum value of the histogram intersection value.

[0013] Optionally, the value of the comprehensive correlation index is between 0 and 1. The closer the value of the comprehensive index is to 1, the stronger the correlation between the site to be forecasted and other water conservancy projects in the same basin.

[0014] Optionally, after S200 and before S300, the following steps are further included: S200a, using the training data set to train the forecast site to obtain a trained forecast model; Specifically, the training data set is classified according to parameter type and site, each classified parameter is input into the prediction model, a cross entropy function and an Adam optimizer are used, and a learning rate of 0.001 is used until the cross entropy regression loss function is minimized, thereby obtaining a trained training model; The training data set includes: historical basic parameters of the site to be predicted, historical basic parameters of the associated sites, and forecast data of the site to be predicted and the associated sites at corresponding historical moments.

[0015] Optionally, in S300, The prediction model is a convolutional neural network model; The convolutional neural network model includes: an input layer, a convolution layer, a batch normalization layer, a pooling layer, a flattening layer, a fully connected layer and an output layer; The convolutional layer uses 64 filters, a kernel size of 3, and an activation function of ReLU.

[0016] Optionally, the method further comprises: The real-time basic parameters of the site to be predicted, the corresponding associated sites and the real-time basic parameters of the associated sites are uniformly stored in a database constructed in advance; Furthermore, if the regional division within the current basin does not match the historical hydrological conditions, that is, the current basin division is expanded or reduced in scope compared with the historical basin division, the associated sites of the current site are modified and configured according to actual needs.

[0017] (III) Beneficial effects The beneficial effects of the present invention are as follows: a hydrological forecast association warning method based on multiple factors of the present invention, by virtue of big data technology, realizes parallel analysis and calculation of a large amount of hydrological data, thereby ensuring data processing efficiency and accuracy, and meeting the warning needs of hydrological forecast; at the same time, the present invention utilizes a variety of influencing factors to associate the site to be forecast with its related associated sites, thereby realizing the associated display function of the hydrological forecast warning and improving the accuracy of the hydrological forecast.

[0018] Furthermore, if the present invention completes system construction, it can quickly form a new technical field of hydrological forecasting and warning technology with hydrological forecasting and warning technology as the core, lead new domestic and international scientific and technological trends, and generate economies of scale. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a flowchart of a multi-factor based hydrological forecast association warning method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0021] Hydrological forecast and warning: Hydrological forecast and warning is a forecast and analysis of the development of water resources in a certain period of time in the future based on the formation and movement of water resources in nature, using historical measured hydrological and meteorological data. It is an important non-engineering measure for flood prevention and disaster reduction.

[0022] Big data technology: The big data technology system is large and complex. The basic technologies include data collection, data preprocessing, distributed storage, NoSQL database, data warehouse, machine learning, parallel computing, visualization and other technical categories and different technical levels. The general big data processing framework mainly includes data collection and preprocessing, data storage, data cleaning, data query analysis and data visualization.

[0023] Multi-factor association analysis technology: Association analysis is an unsupervised data set analysis and simplification technology. It characterizes the overall correlation of the two sets of variable spaces and their respective principal component variables by extracting the components with the largest overall correlation and variance between the effect size and the influencing factor in turn, thereby achieving decorrelation and spatial dimensionality reduction of the variables and obtaining the basic fitting vector for the next step of regression. Multi-factor association analysis takes into account multiple factors related to the decision value at the same time, and seeks the relationship between multiple factor variables and the different effects of multiple influencing factors on the decision value.

[0024] Basin hydrological forecast refers to the qualitative or quantitative prediction of the hydrological conditions of a water body, a region or a hydrological station in the future based on previous or real-time hydrological and meteorological data. Basin hydrological models are usually used for forecasting. The most commonly used hydrological forecast model in China is the Sanshuiyuan Xin'anjiang model, which is a decentralized structure and is divided into four modules: evapotranspiration calculation, runoff calculation, water source division and confluence calculation. The evapotranspiration calculation adopts a three-layer evapotranspiration model; the runoff calculation adopts the full runoff theory; the free water reservoir structure is used to divide the total runoff into three types: surface runoff, subsoil flow and underground runoff; the basin slope confluence calculation adopts a linear reservoir; the river confluence adopts Muskingum segmented continuous calculation.

[0025] The embodiment of the present invention proposes a hydrological forecast correlation warning method based on multiple factors. In order to solve the technical problem in the prior art that insufficient consideration of influencing factors leads to insufficient forecast accuracy and low efficiency, the embodiment of the present invention uses a multi-factor associated site acquisition model including a correlation coefficient acquisition unit, a Copula entropy acquisition unit, a histogram intersection value acquisition unit and an associated site acquisition unit to obtain the associated sites of the site to be forecasted. At the same time, according to the data of the site to be forecasted and the associated sites, the final forecast data is obtained through a trained forecast model, thereby ensuring the efficiency and accuracy of data processing and meeting the needs of hydrological forecasting.

[0026] In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0027] Example 1

[0028] See also Figure 1 , a hydrological forecast association warning method based on multiple factors in this embodiment includes: Step S100, obtaining historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin, inputting the historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin into a pre-set multi-factor associated site acquisition model, and obtaining associated sites of the site to be predicted; Other water conservancy projects are water conservancy projects for which flow data and rainfall data can be obtained; Step S200, obtaining the real-time basic parameters of the site to be forecasted and the real-time basic parameters of the sites associated with the site to be forecasted, and performing data preprocessing on them to obtain preprocessed data; Step S300: input the preprocessed data into the trained forecast model to obtain the forecast data of the site to be forecasted; compare the forecast data of the site to be forecasted with the corresponding data threshold, and issue warning information when the data threshold is exceeded; Basic parameters include: flow data and rainfall data; The pre-set multi-factor associated site acquisition model includes: a correlation coefficient acquisition unit, a Copula entropy acquisition unit, a histogram intersection value acquisition unit and an associated site acquisition unit.

[0029] In a specific embodiment, the data threshold is determined according to specific application scenarios and requirements. In actual operation, it is also necessary to continuously adjust and optimize the data threshold setting to adapt to changing environmental conditions and improve reliability.

[0030] In this embodiment, the forecast data of the site to be forecasted is compared with the corresponding data threshold. When the data threshold is exceeded, the warning information is issued. The specific warning rules are: The alert levels include: Yellow warning: Mild risk, remind relevant personnel to pay attention; Orange warning: moderate risk, preventive measures are recommended; Red alert: severe risk, take emergency measures immediately.

[0031] Set data thresholds for each level based on rules of thumb: Yellow warning threshold: forecast results exceed 110% of the historical average; Orange warning threshold: forecast results exceed 90% of the historical maximum value; Red warning threshold: The forecast result exceeds 110% of the historical maximum.

[0032] A hydrological forecast association warning method based on multiple factors in this embodiment ensures the efficiency and accuracy of data processing and meets the needs of hydrological forecast and warning. At the same time, this embodiment deeply integrates technologies such as neural network models and multi-factor association analysis, broadening the breadth and depth of application in solving hydrological forecast and warning problems.

[0033] Example 2

[0034] A hydrological forecast association warning method based on multiple factors in this embodiment includes: Step S100, obtaining historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin, inputting the historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin into a pre-set multi-factor associated site acquisition model, and obtaining associated sites of the site to be predicted; Other water conservancy projects are water conservancy projects for which flow data and rainfall data can be obtained; Step S200, obtaining the real-time basic parameters of the site to be forecasted and the real-time basic parameters of the sites associated with the site to be forecasted, and performing data preprocessing on them to obtain preprocessed data; Step S300: input the preprocessed data into the trained forecast model to obtain the forecast data of the site to be forecasted; compare the forecast data of the site to be forecasted with the corresponding data threshold, and issue warning information when the data threshold is exceeded; Basic parameters include: flow data and rainfall data; The pre-set multi-factor associated site acquisition model includes: a correlation coefficient acquisition unit, a Copula entropy acquisition unit, a histogram intersection value acquisition unit and an associated site acquisition unit.

[0035] In this embodiment, historical basic parameters of the site to be forecasted and historical basic parameters of other water conservancy projects in the same basin are obtained, and the data sources are hydrological station records, meteorological station records, etc. Other water conservancy projects in the same basin include important water conservancy projects such as associated rainfall stations and reservoirs in the basin.

[0036] In a specific embodiment, the forecast data of the site to be forecasted is input into the grid layer of the site area to be forecasted. At the same time, the grid layer also includes the data threshold in the area. Through this network layer, users can quickly and intuitively understand the information data and data trends of various water and rainfall conditions.

[0037] Furthermore, in step S200, the real-time basic parameters of the site to be forecasted and the implementation basic parameters of the associated sites of the site to be forecasted are obtained, and data preprocessing is performed on them. The acquired and processed data specifically includes: Missing value processing: use interpolation or other methods to fill missing values; Abnormal processing: Use statistical methods, such as the 3σ principle, to detect and process abnormal values; Data format unification: Ensure that the timestamps and units of the data used are consistent.

[0038] In this embodiment, step S300 includes: Step S310: input the preprocessed data into the correlation coefficient acquisition unit, the Copula entropy acquisition unit and the histogram intersection value acquisition unit respectively to acquire the corresponding correlation coefficient, Copula entropy and histogram intersection value; Step S320, inputting the correlation coefficient, Copula entropy and histogram intersection value into the associated site acquisition unit to obtain a comprehensive correlation index; Step S330: sort the comprehensive correlation indicators and obtain the top three water conservancy projects as associated sites of the site to be forecasted.

[0039] Step S310 includes: Step S311: Input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the Pearson correlation coefficient: ; in, x i is the flow data of the station to be predicted at a certain time. is the average flow data of the site to be predicted at that time, y i is the flow data of other water conservancy projects in the same basin at that time, is the average flow data of other water conservancy projects in the same basin at that time, z i is the rainfall data of the forecast station at that time, is the average rainfall data of the forecast station at that time, w i is the rainfall data of other water conservancy projects in the same basin at that time, is the average rainfall data of other water conservancy projects in the same basin at that time, n is the number of basic parameters obtained at that time, K 1 is the weight of traffic data, K 2 is the weight of rainfall data.

[0040] Step S312: Input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the Copula entropy: ; in, c(u, v, z, w) For four variables u, v, z, w The probability density function of u is the flow data of the station to be predicted at a certain time. v is the rainfall data of the forecast station at a certain time. zis the flow data of other water conservancy projects in the same basin at that time, w It is the rainfall data of other water conservancy projects in the same basin at that moment.

[0041] Step S313: Input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the histogram intersection value: ; in, k is the number of intervals of the histogram, h Ai For the site to be forecasted, i The frequency in the interval, h Bi For other water conservancy projects in the same basin i The frequency in the interval, min( h Ai ，h Bi ) is in the i The minimum value of two histograms in the interval.

[0042] In this embodiment, step S320 includes: Step S321: Input the historical basic parameters of the site to be forecasted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain a comprehensive correlation index: ; in, S is the correlation coefficient, They are the weights corresponding to the Pearson correlation coefficient, Copula entropy and histogram intersection values, r is the Pearson correlation coefficient, H is the Copula entropy, H max is the maximum value of Copula entropy, H min is the minimum value of Copula entropy, I is the histogram intersection value, I max is the maximum value of the histogram intersection value.

[0043] The value of the comprehensive correlation index is between 0 and 1. The closer the value of the comprehensive index is to 1, the stronger the correlation between the forecast site and other water conservancy projects in the same basin.

[0044] After step S200 and before step S300, the method further includes: Step S200a, using the training data set to train the forecast site to obtain a trained forecast model; Specifically, the training data set is classified according to parameter type and site, each classified parameter is input into the prediction model, and the cross entropy function and Adam optimizer are used with a learning rate of 0.001 until the cross entropy regression loss function is minimized to obtain a trained training model; The training data set includes: historical basic parameters of the site to be predicted, historical basic parameters of the associated sites, and forecast data of the site to be predicted and the associated sites at corresponding historical moments.

[0045] In step S300, The prediction model is a convolutional neural network model; The convolutional neural network model includes: input layer, convolution layer, batch normalization layer, pooling layer, flattening layer, fully connected layer and output layer; The convolutional layer uses 64 filters, a kernel size of 3, and an activation function of ReLU.

[0046] In this embodiment, it is necessary to ensure that the format of the preprocessed data meets the input requirements of the convolutional neural network model. Generally, the input data needs to be a four-dimensional tensor with a shape of (number of samples, time steps, number of features, number of channels). For example, if each sample contains 100 time steps of flow and rainfall data, and each time step has 2 features (flow and rainfall), the shape of the input data is (number of samples, 100, 2, 1).

[0047] In addition, the pooling layer in this embodiment uses a maximum pooling layer, and the pooling window size is 2×2.

[0048] The multi-factor-based hydrological forecast association warning method in this embodiment also includes: The real-time basic parameters of the site to be predicted, the corresponding associated sites and the real-time basic parameters of the associated sites are uniformly stored in a database constructed in advance; Furthermore, if the regional division within the current basin does not match the historical hydrological conditions, that is, the current basin division is expanded or reduced in scope compared with the historical basin division, the associated sites of the current site are modified and configured according to actual needs.

[0049] In the specific implementation process, when the forecast data exceeds the data threshold, the warning mechanism is triggered and the warning information is released through various channels, such as SMS, email or APP push; Different warning levels are set according to the severity of the forecast data, such as yellow warning, orange warning and red warning. When a warning occurs, the corresponding emergency plan is activated and relevant departments and personnel are notified to take action.

[0050] In this embodiment, after obtaining the forecast data of a certain period of time of the site to be forecasted, the forecast data can be corrected, and the corrected forecast data can be used by the sites associated with the site to be forecasted. The Kalman filter method is used to correct the forecast data, and its formula is: ; in, F k is the state transition model, X k－1 is the state estimation value of the forecast result in the k-1 period, X k is the forecast result after correction for the current period, B k is the covariance of the forecast process, U k is the covariance in the real-time monitoring process, W k is a value drawn from a zero-mean multivariate normal distribution.

[0051] Furthermore, the database constructed in advance includes real-time basic parameters of the site to be forecasted, corresponding associated sites, real-time basic parameters of associated sites, water level data of the site to be forecasted and associated sites, water level-flow relationship between the site to be forecasted and associated sites, forecast data of the site to be forecasted, and corresponding corrected forecast data; Among them, the relationship between water level and flow is fitted according to the least squares method to obtain the quantitative relationship between the two. The least squares method is the most commonly used method to solve curve fitting. Its basic idea is: Let: ; in It is a set of linearly independent basis functions selected in advance, x represents the water level, and a1, a2, …am are unknown coefficients.

[0052] The fitting criterion is to make the real-time observation value Q i The sum of squared errors between the fitted value f(x) is minimized: ; Where n is the number of real-time observation points, Q i is the flow rate at the ith observation point, and x is the water level.

[0053] The present embodiment provides a multi-factor based hydrological forecast association warning method, which utilizes multiple influencing factors to obtain associated sites of the sites to be forecasted, thereby realizing the association display function of the hydrological forecast and warning between the sites to be forecasted of the water conservancy project and other water conservancy projects in the same basin, thereby improving the accuracy of the hydrological forecast.

[0054] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0055] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0056] In the present invention, unless otherwise clearly specified and limited, when a first feature is “on” or “below” a second feature, it may be that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. Moreover, when a first feature is “above”, “above” or “above” a second feature, it may be that the first feature is directly above or obliquely above the second feature, or it may simply mean that the first feature is higher in level than the second feature. When a first feature is “below”, “below” or “below” a second feature, it may be that the first feature is directly below or obliquely below the second feature, or it may simply mean that the first feature is lower in level than the second feature.

[0057] In the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0058] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A hydrological forecast association warning method based on multiple factors, characterized in that: include: S100, obtaining historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin, inputting the historical basic parameters of the site to be predicted and historical basic parameters of other water conservancy projects in the same river basin into a pre-set multi-factor associated site acquisition model, and obtaining associated sites of the site to be predicted; The other water conservancy projects are water conservancy projects that can obtain flow data and rainfall data; S200, obtaining real-time basic parameters of the site to be forecasted and real-time basic parameters of the sites associated with the site to be forecasted, and performing data preprocessing on them to obtain preprocessed data; S300, inputting the preprocessed data into a trained forecast model to obtain forecast data of the site to be forecasted; Compare the forecast data of the site to be forecasted with the corresponding data threshold, and issue warning information when the data threshold is exceeded; The basic parameters include: flow data and rainfall data; The pre-set multi-factor associated site acquisition model includes: a correlation coefficient acquisition unit, a Copula entropy acquisition unit, a histogram intersection value acquisition unit and an associated site acquisition unit.

2. The hydrological forecast association warning method based on multiple factors according to claim 1 is characterized in that: The S300 includes: S310, inputting the preprocessed data into the correlation coefficient acquisition unit, the Copula entropy acquisition unit and the histogram intersection value acquisition unit respectively to acquire the corresponding correlation coefficient, Copula entropy and histogram intersection value; S320, inputting the correlation coefficient, Copula entropy and histogram intersection value into an associated site acquisition unit to obtain a comprehensive correlation index; S330, sorting the comprehensive correlation indicators, and obtaining the top three water conservancy projects as associated sites of the site to be forecasted.

3. The hydrological forecast association warning method based on multiple factors according to claim 2 is characterized in that: The S310 includes: S311. Input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the Pearson correlation coefficient: ； in, x i is the flow data of the station to be predicted at a certain time. is the average flow data of the site to be predicted at that time, y i is the flow data of other water conservancy projects in the same basin at that time, is the average flow data of other water conservancy projects in the same basin at that time, z i is the rainfall data of the forecast station at that time, is the average rainfall data of the forecast station at that time, w i is the rainfall data of other water conservancy projects in the same basin at that time, is the average rainfall data of other water conservancy projects in the same basin at that moment, n is the number of basic parameters obtained at that moment, K1 is the weight of flow data, and K2 is the weight of rainfall data.

4. The hydrological forecast association warning method based on multiple factors according to claim 3 is characterized in that: The S310 further includes: S312. Input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the Copula entropy: in, c(u,v,z,w) For four variables u,v,z,w The probability density function of u is the flow data of the station to be predicted at a certain time. v is the rainfall data of the forecast station at a certain time. z is the flow data of other water conservancy projects in the same basin at that time, w It is the rainfall data of other water conservancy projects in the same basin at that moment.

5. The hydrological forecast association warning method based on multiple factors according to claim 4 is characterized in that: The S310 further includes: S313, input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain the histogram intersection value: ； in, k is the number of intervals of the histogram, h Ai For the site to be forecasted, i The frequency in the interval, h Bi For other water conservancy projects in the same basin i The frequency in the interval, min( h Ai ，h Bi ) is in the i The minimum value of two histograms in the interval.

6. The hydrological forecast association warning method based on multiple factors according to claim 2 is characterized in that: The S320 includes: S321, input the historical basic parameters of the site to be predicted and the historical basic parameters of other water conservancy projects in the same basin into the following formula to obtain a comprehensive correlation index: ； in, S is the correlation coefficient, They are the weights corresponding to the Pearson correlation coefficient, Copula entropy and histogram intersection values, r is the Pearson correlation coefficient, H is the Copula entropy, H max is the maximum value of Copula entropy, H min is the minimum value of Copula entropy, I is the histogram intersection value, I max is the maximum value of the histogram intersection value.

7. The hydrological forecast association warning method based on multiple factors according to claim 6 is characterized in that: The value of the comprehensive correlation index is between 0 and 1. When the value of the comprehensive index is closer to 1, the correlation between the forecast site and other water conservancy projects in the same basin is stronger.

8. The hydrological forecast association warning method based on multiple factors according to claim 1 is characterized in that: After S200 and before S300, the method further includes: S200a, using the training data set to train the forecast site to obtain a trained forecast model; Specifically, the training data set is classified according to parameter type and site, each classified parameter is input into the prediction model, a cross entropy function and an Adam optimizer are used, and a learning rate of 0.001 is used until the cross entropy regression loss function is minimized, thereby obtaining a trained training model; The training data set includes: historical basic parameters of the site to be predicted, historical basic parameters of the associated sites, and forecast data of the site to be predicted and the associated sites at corresponding historical moments.

9. The hydrological forecast association warning method based on multiple factors according to claim 1 is characterized in that: In the S300, The prediction model is a convolutional neural network model; The convolutional neural network model includes: an input layer, a convolution layer, a batch normalization layer, a pooling layer, a flattening layer, a fully connected layer and an output layer; The convolutional layer uses 64 filters, a kernel size of 3, and an activation function of ReLU.

10. The hydrological forecast association warning method based on multiple factors according to claim 1 is characterized in that: The method further comprises: The real-time basic parameters of the site to be predicted, the corresponding associated sites and the real-time basic parameters of the associated sites are uniformly stored in a database constructed in advance; Furthermore, if the regional division within the current basin does not match the historical hydrological conditions, that is, the current basin division is expanded or reduced in scope compared with the historical basin division, the associated sites of the current site are modified and configured according to actual needs.

Images