Hydrological simulation method and system for simulating hydrological process of hydropower station water body

Abstract

The invention provides a hydrological simulation method and system for simulating the hydrological process of a hydropower station water body, and relates to the field of data processing. The method comprises the following steps: determining a plurality of warehousing flow influence factors, an optimal warehousing influence time lag corresponding to the plurality of warehousing flow influence factors, a plurality of ex-warehouse flow influence factors and an optimal ex-warehouse influence time lag corresponding to the plurality of ex-warehouse flow influence factors, and establishing a hydrological simulator, the residual structure is used for introducing an input cell state and an output cell state of a previous time step, the cell state updating layer is used for updating the cell state according to the output of the residual structure and the output of the input gate, and the linear layer is used for performing water level prediction according to the output of the output gate; training a hydrological simulator; acquiring current input data of the hydropower station to be simulated; the water level of the to-be-simulated hydropower station is predicted through the trained hydrological simulator, and the method has the advantage that the simulation accuracy of the hydrological process of the hydropower station water body is improved.

Description

Technical Field

[0001] The invention relates to the field of data processing, and in particular to a hydrological simulation method and system for simulating the hydrological process of a water body in a hydropower station. Background Art

[0002] It is very necessary to generalize the complex hydrological phenomena, processes and external limiting factors through hydrological simulation to give an approximate scientific model to carry out simulated dispatching of hydropower station reservoirs (groups). It not only has important theoretical significance, but also has greater economic value. In this way, the dispatching feedback optimization of the dispatching results of the economically reasonable dispatching of the runoff process can make full use of water energy resources and water resources, increase the power generation of hydropower stations, and improve the comprehensive utilization benefits of hydropower station reservoirs under the premise of meeting the safety of engineering and production, and at the same time provide strong support for preventing and responding to water resource disasters.

[0003] In recent years, the research on hydrological simulators is mainly based on the unit flow characteristic curve and discharge curve calibrated when the hydropower station is built, and then the upstream outflow time is evaluated by experience. Finally, the unit load and gate opening are converted into inflow and outflow flow, and then the water level changes of the downstream power station are deduced. This method ignores the complex nonlinear relationship and time-varying characteristics of the control variables of the power station, and the entire simulation process introduces additional errors caused by various curve inaccuracies.

[0004] Therefore, it is necessary to provide a hydrological simulation method and system for simulating the hydrological process of the water body of a hydropower station, so as to improve the simulation accuracy of the hydrological process of the water body of the hydropower station. Summary of the invention

[0005] The present invention provides a hydrological simulation method for simulating the hydrological process of a water body in a hydropower station, comprising: determining a plurality of inflow influencing factors and the optimal inflow influence time lags corresponding to the plurality of inflow influencing factors; determining a plurality of outflow influencing factors and the optimal outflow influence time lags corresponding to the plurality of outflow influencing factors; establishing a hydrological simulator based on the plurality of inflow influencing factors and the optimal inflow influence time lags corresponding to the plurality of inflow influencing factors and the plurality of outflow influencing factors and the optimal outflow influence time lags corresponding to the plurality of outflow influencing factors, wherein the hydrological simulator at least comprises a residual structure, an input gate, a cell state update layer, an output gate and a linear layer, the residual structure being used to introduce the input cell state and the output cell state of the previous time step, and the input gate being used to calculate the degree to which new information is accepted and to generate a new candidate memory state. , the cell state update layer is used to update the cell state according to the output of the residual structure and the output of the input gate, the output gate is used to determine the information in the output cell state of the current time step output by the cell state update layer that needs to be output to the hidden state, and the linear layer is used to predict the water level according to the output of the output gate; establish multiple training samples; train the hydrological simulator through the multiple training samples to generate a trained hydrological simulator; based on the multiple inflow influencing factors and the optimal inflow influence time lags corresponding to the multiple inflow influencing factors and the multiple outflow influencing factors and the optimal outflow influence time lags corresponding to the multiple outflow influencing factors, obtain the current input data of the hydropower station to be simulated; predict the water level of the hydropower station to be simulated according to the current input data of the hydropower station to be simulated through the trained hydrological simulator.

[0006] Furthermore, the multiple factors affecting the inflow flow include at least the unit load of the upstream hydropower station, the gate opening, the water level in front of the dam, the water level behind the dam and multiple environmental factors.

[0007] Furthermore, determining the optimal time lag of inflow influence corresponding to the multiple inflow flow influencing factors includes: determining the first correlation coefficients between the multiple inflow flow influencing factors and the water level corresponding to the multiple first candidate time lags; and selecting the optimal time lag of inflow influence corresponding to the multiple inflow flow influencing factors from the multiple first candidate time lags based on the first correlation coefficients between the multiple inflow flow influencing factors and the water level corresponding to the multiple first candidate time lags.

[0008] Furthermore, the multiple outflow influencing factors include at least the current power generation load plan and gate opening plan of the hydropower station.

[0009] Further, determining the optimal time lag of outbound impact corresponding to the multiple outbound flow influencing factors includes: determining the second correlation coefficients between the multiple outbound flow influencing factors and the water level corresponding to the multiple second candidate time lags; and selecting the optimal time lag of outbound impact corresponding to the multiple outbound flow influencing factors from the multiple second candidate time lags based on the second correlation coefficients between the multiple outbound flow influencing factors and the water level corresponding to the multiple second candidate time lags.

[0010] Furthermore, the residual structure introduces the input cell state and output cell state of the previous time step based on the following formula: ,in, is the output of the residual structure, is the output cell state at the previous time step, is the input cell state at the previous time step.

[0011] Furthermore, the cell state update layer updates the cell state according to the output of the residual structure and the output of the input gate based on the following formula: ,in, is the output cell state at the current time step, It is a value between 0 and 1 output by the input gate, indicating the degree to which the new information is accepted. is the new candidate memory state output by the input gate, is element-wise multiplication.

[0012] Furthermore, the output gate determines which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state based on the following formula: , ,in, It is a value between 0 and 1 for the output of the output gate, which is used to determine which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state. is the weight matrix of the output gate, is the hidden state of the previous time step, is the feature matrix, is the bias term of the output gate, is the sigmoid function, is the tangent value of the output cell state at the current time step, is the tangent function, is the hidden state at the current time step.

[0013] Furthermore, the linear layer predicts the water level according to the output of the output gate based on the following formula: ,in, is the predicted water level output by the linear layer, is the weight of the linear layer, is the bias of the linear layer.

[0014] The present invention provides a hydrological simulation system for simulating the hydrological process of a water body in a hydropower station, comprising: a parameter determination module, used to determine a plurality of inflow influencing factors and the optimal inflow influence time lags corresponding to the plurality of inflow influencing factors; and also used to determine a plurality of outflow influencing factors and the optimal outflow influence time lags corresponding to the plurality of outflow influencing factors; a model establishment module, used to establish a hydrological simulator based on the plurality of inflow influencing factors and the optimal inflow influence time lags corresponding to the plurality of inflow influencing factors and the plurality of outflow influencing factors and the optimal outflow influence time lags corresponding to the plurality of outflow influencing factors, wherein the hydrological simulator at least comprises a residual structure, an input gate, a cell state update layer, an output gate and a linear layer, the residual structure is used to introduce the input cell state and the output cell state of the previous time step, the input gate is used to calculate the degree to which new information is accepted and to generate new candidate memory states, and the cell state update layer is used to generate new candidate memory states. The new layer is used to update the cell state according to the output of the residual structure and the output of the input gate, the output gate is used to determine which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state, and the linear layer is used to predict the water level according to the output of the output gate; a sample establishment module is used to establish multiple training samples; a model training module is used to train the hydrological simulator through the multiple training samples to generate a trained hydrological simulator; a data acquisition module is used to obtain the current input data of the hydropower station to be simulated based on the multiple inflow influencing factors and the optimal inflow influence time lags corresponding to the multiple inflow influencing factors and the multiple outflow influencing factors and the optimal outflow influence time lags corresponding to the multiple outflow influencing factors; a hydrological simulation module is used to predict the water level of the hydropower station to be simulated according to the current input data of the hydropower station to be simulated through the trained hydrological simulator.

[0015] Compared with the prior art, the hydrological simulation method and system for simulating the hydrological process of the water body of a hydropower station provided by the present invention have at least the following beneficial effects: By introducing complex neural network structures such as residual structure, input gate, cell state update layer, output gate and linear layer, the hydrological simulator can capture the complex nonlinear and long-term dependencies in time series data. This enables it to simulate and predict water level changes more accurately, especially for complex systems affected by multiple inflow and outflow factors. Through training, the hydrological simulator can learn the specific impact of different inflow and outflow factors on water level changes and their interactions. This makes the model highly adaptable to water level prediction under different conditions and improves the generalization ability of the model. The training process is based on a large amount of sample data, automatically learning and optimizing model parameters, reducing errors caused by human intervention and subjective judgment. With the continuous addition of new data, the model can also be learned and updated online to further improve the prediction performance. Using real-time or recent inflow and outflow data, the model can quickly predict future water level changes and provide timely and accurate reference for the management and dispatch of hydropower stations. This helps hydropower stations formulate more scientific operation plans and improve the efficiency of water energy utilization. Accurate water level forecasting helps hydropower stations to promptly detect potential water level anomalies or dangerous situations, and take appropriate measures to prevent or respond, thereby reducing operational risks and ensuring the safe and stable operation of hydropower stations. By comprehensively considering multiple factors affecting inflow and outflow flows, the model can provide a more comprehensive and in-depth analysis of water resource conditions. This helps to achieve optimal allocation and rational use of water resources, improve water resource utilization efficiency, and promote sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein: Figure 1 is a schematic flow chart of a hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to some embodiments of this specification; Figure 2 is a schematic diagram of a calculation process of a hydrological simulator according to some embodiments of this specification; Figure 3 is a schematic diagram of the correlation coefficient between the factors affecting the inflow flow and the water level according to some embodiments of this specification; Figure 4 It is a relationship diagram between water level and main influencing factors according to some embodiments of this specification; Figure 5 is a schematic diagram of the structure of a hydrological simulator according to some embodiments of this specification; Figure 6 is a schematic diagram of water level simulation error according to some embodiments of this specification; Figure 7 It is a module schematic diagram of a hydrological simulation system for simulating the hydrological process of a water body in a hydropower station according to some embodiments of this specification. DETAILED DESCRIPTION

[0017] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0018] Figure 1 is a flow chart of a hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to some embodiments of this specification, such as Figure 1 As shown, the hydrological simulation method for simulating the hydrological process of the water body of a hydropower station may include the following process.

[0019] Step 110, determining a plurality of inbound flow influencing factors and optimal inbound flow influencing time lags corresponding to the plurality of inbound flow influencing factors.

[0020] The core work of hydrological simulation is to deduce the change of water level in front of the hydropower station dam according to the changes of power station units, gates and meteorological factors. Its essence is to determine the main factors affecting water level changes and the relationship between each factor and the simulation target. Therefore, it is necessary to first determine the physical model that the water level simulation relies on, explore the core influencing factors, and determine the input of hydrological simulation through this analysis. Specifically, given the initial water level , find the final water level , its mathematical expression is The main factors affecting the future water level are transmitted through two variables: the current water level and the future water level change. Based on this, the input of the hydrological simulation is specifically constructed: (1) Current water level

[0021] Current water level It represents the latest observation value of the water level. However, due to the influence of factors such as water waves and the installation location of the water level meter, the instantaneous water level measurement value cannot represent the true water level value. After noise reduction, the observation value over a period of time can better represent the true value of the water level. Therefore, the historical water level over the past period of time is introduced ( represents the historical duration) as the input for hydrological simulation; (2) Water level change value

[0022] The direct factors affecting the change of water level are the inflow and outflow. However, since the inflow and outflow data are obtained by fitting the load and gate opening respectively in combination with the unit flow characteristic (NHQ) curve and the gate discharge curve, errors of the NHQ curve and the discharge curve themselves will be introduced. In the Dadu River Basin, the power generation load value, water level and gate opening value are observed data, which have lower systematic errors than the inflow and outflow. Therefore, the following main influencing factors of inflow and outflow are introduced: (3) Main factors affecting inbound traffic The inflow is mainly affected by the power generation load and gate opening of the upstream power station. Since the water level in front of the dam and the water level behind the dam of the upstream power station imply the movement relationship between the upstream inflow and outflow, the historical unit load, historical gate opening, and historical water level in front of and behind the dam of the upstream power station are introduced as simulator inputs in combination with the range of flow arrival time of the upstream power station. At the same time, considering the natural flow generation in the basin, regional weather forecast data are introduced as inputs for hydrological simulation.

[0023] Preferably, multiple factors affecting the inflow flow include at least the unit load of the upstream hydropower station, the gate opening, the water level in front of the dam, the water level behind the dam and multiple environmental factors, among which the environmental factors may include at least the ambient temperature, ambient wind direction, ambient wind speed, etc.

[0024] Preferably, determining the optimal time lag of inbound impact corresponding to multiple inbound flow influencing factors includes: Determine first correlation coefficients between a plurality of inflow influencing factors and water level corresponding to a plurality of first candidate time lags, for example, the plurality of first candidate time lags may include 5 minutes ahead, 10 minutes ahead, 30 minutes ahead, one hour ahead, 5 minutes behind, 10 minutes behind, 30 minutes behind, one hour behind, etc.; Based on the first correlation coefficients between the multiple inflow influencing factors corresponding to the multiple first candidate time lags and the water level, the optimal inflow influencing time lags corresponding to the multiple inflow influencing factors are selected from the multiple first candidate time lags.

[0025] Specifically, for each first candidate time lag, the first correlation coefficient between each inflow influencing factor and the water level can be calculated according to the following formula:

[0026] in, For the corresponding The first correlation coefficient between the e-th inflow influencing factor and the water level of the first candidate time lag, For the corresponding The e-th inflow influencing factor of the first candidate time lag is at time point t minus the The value of the first candidate time lag at the time point, The first candidate time lag represents synchronization, the The first candidate lag < 0 indicates leading, The first candidate lag > 0 indicates a lag, is the total number of selected time points, is the water level at time point t.

[0027] Figure 3 is a schematic diagram of the correlation coefficient between the factors affecting the inflow flow and the water level according to some embodiments of this specification, such as Figure 3 As shown in the figure, the same influencing factor of inflow has different correlation coefficients with water level at different lag periods.

[0028] For each first candidate time lag, the matching value corresponding to the first candidate time lag can be calculated according to the first correlation coefficient between each inflow influencing factor and the water level based on the following formula:

[0029] in, For the The matching value corresponding to the first candidate time lag, is the total number of factors affecting the inbound flow, is the weight corresponding to the e-th inflow flow influencing factor.

[0030] The first candidate time lag with the largest matching value may be used as the optimal inbound impact time lag corresponding to multiple inbound flow influencing factors.

[0031] Step 120, determining a plurality of outbound flow influencing factors and optimal outbound flow influencing time lags corresponding to the plurality of outbound flow influencing factors.

[0032] Preferably, the multiple outflow influencing factors at least include the current power generation load plan and gate opening plan of the hydropower station. At the same time, considering the potential scheduling mode that may be implied by the historical power generation load and gate opening, the historical load and gate opening are introduced as the input of the hydrological simulation.

[0033] Preferably, determining the optimal time lag of outbound impact corresponding to multiple outbound flow influencing factors includes: Determine a second correlation coefficient between a plurality of outflow flow influencing factors and the water level corresponding to a plurality of second candidate time lags; Based on the second correlation coefficients between the multiple outbound flow influencing factors corresponding to the multiple second candidate time lags and the water level, the optimal outbound flow influencing time lag corresponding to the multiple outbound flow influencing factors is selected from the multiple second candidate time lags.

[0034] The method of determining the optimal time lag of outbound impact corresponding to multiple outbound flow influencing factors is similar to the method of determining the optimal time lag of inbound impact corresponding to multiple inbound flow influencing factors, which will not be repeated here.

[0035] Step 130, a hydrological simulator is established based on a plurality of inflow influencing factors and the optimal inflow influencing time lags corresponding to the plurality of inflow influencing factors and a plurality of outflow influencing factors and the optimal outflow influencing time lags corresponding to the plurality of outflow influencing factors.

[0036] Based on multiple inflow influencing factors and the optimal inflow influencing time lag corresponding to the multiple inflow influencing factors and multiple outflow influencing factors and the optimal outflow influencing time lag corresponding to the multiple outflow influencing factors, step 130 will examine the relationship between the influencing factors and the water level in order to establish a hydrological simulator. The water level simulation needs to deduce the water level on the downstream dam according to the factors affecting the water level upstream and downstream. The relationship between each factor and the water level below the downstream dam is relatively complex, showing the following characteristics: (1) Nonlinearity Figure 4 This is a diagram showing the relationship between water level and main influencing factors according to some embodiments of this specification, such as Figure 4 As shown in the figure, the future water level change of the downstream hydropower station and the future load change show an overall opposite relationship, that is, as the future load increases, its water level shows a downward trend, but the relationship between the two is not a linear change; the future water level change of the current hydropower station is positively correlated with the change in the load of the upstream hydropower station, that is, as the power generation load of the upstream hydropower station increases, the water level of the downstream hydropower station has an upward trend, but the overall relationship is not a linear change.

[0037] (2) Long memory Due to the influence of upstream flow time, the impact of the discharge of the upstream hydropower station will affect the water level of the downstream hydropower station after a certain period of time. From the perspective of time series, there is an obvious time difference correlation between the water level of the downstream hydropower station and the influencing factors, that is, long memory.

[0038] Based on the above research conclusions, the hydrological simulator selected the Long Short-Term Memory (LSTM) network model, and continued to optimize and improve the model level according to the characteristics of the watershed water conditions and the model characteristics; first, the flow time is affected by the outflow of the upstream hydropower station. If the outflow of the upstream hydropower station increases, the flow time decreases, and if the outflow of the upstream hydropower station decreases, the flow time increases. That is, the flow time has obvious time-varying characteristics. In the modeling stage using the long short-term memory network, this study introduced the attention mechanism, which can automatically adjust the hydrological simulator to pay attention to information in different time dimensions according to the characteristics of the upstream hydropower station.

[0039] Secondly, considering that the long short-term memory network itself has more gates and more complex operations, the long short-term memory network has been improved. The hydrological simulator recorded in this manual uses residuals instead of forget gates to improve the model running speed while retaining the long memory characteristics of the model.

[0040] Figure 2 is a schematic diagram of the calculation process of the hydrological simulator according to some embodiments of this specification, Figure 5 is a schematic diagram of the structure of a hydrological simulator according to some embodiments of this specification, Figure 5 middle are all input data, All are output data. are weights, such as Figure 2 and Figure 5 As shown, the hydrological simulator at least includes a residual structure, an input gate, a cell state update layer, an output gate and a linear layer. The residual structure is used to introduce the input cell state and the output cell state of the previous time step. The input gate is used to calculate the degree to which new information is accepted and to generate a new candidate memory state. The cell state update layer is used to update the cell state according to the output of the residual structure and the output of the input gate. The output gate is used to determine the information in the output cell state of the current time step output by the cell state update layer that needs to be output to the hidden state. The linear layer is used to predict the water level according to the output of the output gate.

[0041] Preferably, the residual structure introduces the input cell state and output cell state of the previous time step based on the following formula:

[0042] in, is the output of the residual structure, is the output cell state at the previous time step, is the input cell state at the previous time step.

[0043] The input gate determines which new information needs to be added to the cell state. It is divided into two parts: one is the sigmoid layer that determines which information needs to be updated, and the other is the tanh layer that generates new candidate memory states. The formula is:

[0044] in, is the output of the input gate, a value between 0 and 1, indicating the degree to which the new information is accepted. is the weight matrix of the input gate. is the new candidate memory state. is the weight matrix of the candidate memory state. and are the bias terms for the input gate and candidate memory state, respectively. is the sigmoid function, which maps the input to between 0 and 1. tanh is the tangent function. Represents the feature matrix. Represents the output hidden layer state at the previous time step.

[0045] Preferably, the cell state update layer updates the cell state according to the output of the residual structure and the output of the input gate based on the following formula:

[0046] in, is the output cell state at the current time step, It is a value between 0 and 1 output by the input gate, indicating the degree to which the new information is accepted. is the new candidate memory state output by the input gate, is element-wise multiplication.

[0047] Preferably, the output gate determines which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state based on the following formula:

[0048] in, It is a value between 0 and 1 for the output of the output gate, which is used to determine which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state. is the weight matrix of the output gate, is the hidden state of the previous time step, is the feature matrix, is the bias term of the output gate, is the sigmoid function, is the tangent value of the output cell state at the current time step, is the tangent function, is the hidden state at the current time step.

[0049] Preferably, the linear layer predicts the water level based on the output of the output gate based on the following formula:

[0050] in, is the predicted water level output by the linear layer, is the weight of the linear layer, is the bias of the linear layer.

[0051] Step 140: Create multiple training samples.

[0052] Step 150: training the hydrological simulator using a plurality of training samples to generate a trained hydrological simulator.

[0053] The loss function for training the hydrological simulator is:

[0054] in, is the mean square error, is the total number of training samples, is the prediction result of the i-th training sample, is the label of the i-th training sample.

[0055] The weight matrix in the hydrological simulator (such as ) and bias terms (such as ) are all acquired through learning.

[0056] Step 160, based on multiple inflow influencing factors and the optimal inflow influence time lags corresponding to the multiple inflow influencing factors and multiple outflow influencing factors and the optimal outflow influence time lags corresponding to the multiple outflow influencing factors, obtain current input data of the hydropower station to be simulated.

[0057] As an example only, the current input data of the hydropower station to be simulated at least includes the upstream power station characteristic data determined based on multiple inflow influencing factors and multiple inflow influencing factors as shown in Table 1.

[0058] Table 1

[0059] Step 170, predicting the water level of the hydropower station to be simulated according to the current input data of the hydropower station to be simulated by the trained hydrological simulator.

[0060] The following is an explanation of the technical effects of the hydrological simulation method used to simulate the hydrological processes of the water body of a hydropower station in conjunction with experiments.

[0061] The water level simulator was used to simulate the water level of hydropower station A during the dry season. Simulation experiments were performed for 90 minutes at a time for 5 minutes in the future. Under the same boundary conditions, the simulation model based on the NHQ curve was compared. The experimental results are shown in Table 2 and Figure 6 As shown in the figure, from the experimental results, the water level simulator has achieved high accuracy. The simulation error relative to the actual water level in the next 90 minutes is only 0.094m, which is much lower than the simulation error based on the NHQ curve.

[0062] Table 2

[0063] Figure 7is a schematic diagram of a module of a hydrological simulation system for simulating the hydrological process of a water body in a hydropower station according to some embodiments of this specification, such as Figure 7 As shown, the hydrological simulation system for simulating the hydrological process of the water body of a hydropower station may include a parameter determination module, a model establishment module, a sample establishment module, a model training module, a data acquisition module and a hydrological simulation module.

[0064] The parameter determination module can be used to determine multiple inbound flow influencing factors and the optimal inbound impact time lag corresponding to the multiple inbound flow influencing factors; it can also be used to determine multiple outbound flow influencing factors and the optimal outbound impact time lag corresponding to the multiple outbound flow influencing factors.

[0065] The model building module can be used to establish a hydrological simulator based on multiple inflow influencing factors and the optimal inflow impact time lags corresponding to the multiple inflow influencing factors and multiple outflow influencing factors and the optimal outflow impact time lags corresponding to the multiple outflow influencing factors, wherein the hydrological simulator at least includes a residual structure, an input gate, a cell state update layer, an output gate and a linear layer, the residual structure is used to introduce the input cell state and the output cell state of the previous time step, the input gate is used to calculate the degree to which new information is accepted and to generate new candidate memory states, the cell state update layer is used to update the cell state according to the output of the residual structure and the output of the input gate, the output gate is used to determine which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state, and the linear layer is used to predict the water level according to the output of the output gate.

[0066] The sample building module can be used to build multiple training samples.

[0067] The model training module can be used to train the hydrological simulator through multiple training samples to generate a trained hydrological simulator.

[0068] The data acquisition module can be used to obtain current input data of the hydropower station to be simulated based on multiple inflow influencing factors and the optimal inflow influence time lags corresponding to the multiple inflow influencing factors and multiple outflow influencing factors and the optimal outflow influence time lags corresponding to the multiple outflow influencing factors.

[0069] The hydrological simulation module can be used to predict the water level of the hydropower station to be simulated based on the current input data of the hydropower station to be simulated through a trained hydrological simulator.

[0070] The hydrological simulation system for simulating the hydrological processes of the water body of a hydropower station can be used to execute the hydrological simulation method for simulating the hydrological processes of the water body of a hydropower station. For more descriptions of the hydrological simulation system for simulating the hydrological processes of the water body of a hydropower station, please refer to the relevant descriptions of the hydrological simulation method for simulating the hydrological processes of the water body of a hydropower station, which will not be repeated here.

[0071] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A hydrological simulation method for simulating the hydrological process of a water body in a hydropower station, characterized in that: include: Determine a plurality of inbound flow influencing factors and optimal inbound flow influencing time lags corresponding to the plurality of inbound flow influencing factors; Determine a plurality of outbound flow influencing factors and optimal outbound flow influencing time lags corresponding to the plurality of outbound flow influencing factors; Based on the multiple inflow influencing factors and the optimal inflow influence time lags corresponding to the multiple inflow influencing factors and the multiple outflow influencing factors and the optimal outflow influence time lags corresponding to the multiple outflow influencing factors, a hydrological simulator is established, wherein the hydrological simulator at least includes a residual structure, an input gate, a cell state update layer, an output gate and a linear layer, the residual structure is used to introduce the input cell state and the output cell state of the previous time step, the input gate is used to calculate the degree to which new information is accepted and generate a new candidate memory state, the cell state update layer is used to update the cell state according to the output of the residual structure and the output of the input gate, the output gate is used to determine the information in the output cell state of the current time step output by the cell state update layer that needs to be output to the hidden state, and the linear layer is used to predict the water level according to the output of the output gate; Create multiple training samples; Training the hydrological simulator using the multiple training samples to generate a trained hydrological simulator; Based on the multiple inflow influencing factors and the optimal inflow impact time lags corresponding to the multiple inflow influencing factors and the multiple outflow influencing factors and the optimal outflow impact time lags corresponding to the multiple outflow influencing factors, current input data of the hydropower station to be simulated is obtained; The trained hydrological simulator predicts the water level of the hydropower station to be simulated according to the current input data of the hydropower station to be simulated.

2. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to claim 1, characterized in that: The multiple factors affecting the inflow flow rate at least include the unit load of the upstream hydropower station, the gate opening, the water level in front of the dam, the water level behind the dam and multiple environmental factors.

3. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to claim 1, characterized in that: Determining the optimal time lag of the inbound impact corresponding to the multiple inbound flow influencing factors includes: Determine first correlation coefficients between the plurality of inflow influencing factors and the water level corresponding to the plurality of first candidate time lags; Based on the first correlation coefficients between the multiple inflow influencing factors and the water level corresponding to the multiple first candidate time lags, optimal inflow influencing time lags corresponding to the multiple inflow influencing factors are selected from the multiple first candidate time lags.

4. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to claim 1, characterized in that: The multiple outflow flow influencing factors at least include the current power generation load plan and gate opening plan of the hydropower station.

5. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to claim 1, characterized in that: Determining the optimal time lag of the outbound impact corresponding to the multiple outbound flow influencing factors includes: Determine second correlation coefficients between the plurality of outflow flow influencing factors and the water level corresponding to the plurality of second candidate time lags; Based on the second correlation coefficients between the multiple outbound flow rate influencing factors and the water level corresponding to the multiple second candidate time lags, optimal outbound flow rate influencing time lags corresponding to the multiple outbound flow rate influencing factors are selected from the multiple second candidate time lags.

6. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to any one of claims 1 to 5, characterized in that: The residual structure introduces the input cell state and output cell state of the previous time step based on the following formula: in, is the output of the residual structure, is the output cell state at the previous time step, is the input cell state at the previous time step.

7. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to claim 6, characterized in that: The cell state update layer updates the cell state according to the output of the residual structure and the output of the input gate based on the following formula: in, is the output cell state at the current time step, It is a value between 0 and 1 output by the input gate, indicating the degree to which the new information is accepted. is the new candidate memory state output by the input gate, is element-wise multiplication.

8. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to claim 7, characterized in that: The output gate determines which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state based on the following formula: in, It is a value between 0 and 1 for the output of the output gate, which is used to determine which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state. is the weight matrix of the output gate, is the hidden state of the previous time step, is the feature matrix, is the bias term of the output gate, is the sigmoid function, is the tangent value of the output cell state at the current time step, is the tangent function, is the hidden state at the current time step.

9. The hydrological simulation method for simulating the hydrological process of a water body in a hydropower station according to claim 8, characterized in that: The linear layer predicts the water level according to the output of the output gate based on the following formula: in, is the predicted water level output by the linear layer, is the weight of the linear layer, is the bias of the linear layer.

10. A hydrological simulation system for simulating the hydrological process of a water body in a hydropower station, characterized in that: include: A parameter determination module, used to determine a plurality of inbound flow influencing factors and optimal inbound flow influencing time lags corresponding to the plurality of inbound flow influencing factors; It is also used to determine multiple outbound flow influencing factors and optimal outbound flow influencing time lags corresponding to the multiple outbound flow influencing factors; A model building module, used to build a hydrological simulator based on the multiple inflow influencing factors and the optimal inflow influence time lags corresponding to the multiple inflow influencing factors and the multiple outflow influencing factors and the optimal outflow influence time lags corresponding to the multiple outflow influencing factors, wherein the hydrological simulator at least includes a residual structure, an input gate, a cell state update layer, an output gate and a linear layer, the residual structure is used to introduce the input cell state and the output cell state of the previous time step, the input gate is used to calculate the degree to which new information is accepted and generate a new candidate memory state, the cell state update layer is used to update the cell state according to the output of the residual structure and the output of the input gate, the output gate is used to determine which information in the output cell state of the current time step output by the cell state update layer needs to be output to the hidden state, and the linear layer is used to predict the water level according to the output of the output gate; A sample establishment module, used to establish multiple training samples; A model training module, used for training the hydrological simulator through the multiple training samples to generate a trained hydrological simulator; A data acquisition module, used for acquiring current input data of the hydropower station to be simulated based on the multiple inflow influencing factors and the optimal inflow influence time lags corresponding to the multiple inflow influencing factors and the multiple outflow influencing factors and the optimal outflow influence time lags corresponding to the multiple outflow influencing factors; The hydrological simulation module is used to predict the water level of the hydroelectric power station to be simulated according to the current input data of the hydroelectric power station to be simulated through a trained hydrological simulator.

Images