Operation control method and system for pumped storage unit

Abstract

The invention discloses an operation control method and system for a pumped storage unit, and the method comprises the steps: screening the output data and future meteorological data of a new energy unit, and obtaining a new energy output prediction result according to a multi-source prediction data set constructed by a screening result; determining each data subset of the load data under different dimensions, and inputting each data subset into a corresponding prediction model to obtain a medium and long term load prediction result; inputting the load data and the future meteorological data into the short-term load prediction model to obtain a short-term load prediction result; determining an optimal reservoir water storage target curve based on the new energy output prediction result and the medium and long term load prediction result, and determining an optimal unit output result based on the new energy output prediction result and the short term load prediction result; and executing an operation control strategy generated by the optimal reservoir water storage target curve and the optimal unit output result. By means of the method, operation optimization control over the pumped storage unit is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of pumped storage, and in particular to an operation control method and system of a pumped storage unit. Background Art

[0002] With the global emphasis on environmental protection and sustainable development, the installed capacity and power generation of new energy sources such as solar energy and wind energy are increasing. However, the uncertainty of new energy sources has brought great challenges to the stable operation of the power system. On this basis, pumped storage units, as an important energy storage facility, have multiple functions in the power system, such as peak regulation, frequency regulation, voltage regulation and standby.

[0003] However, traditional operation and control methods of pumped storage units are often based on manual experience or a single model, without considering the uncertainty and complexity of renewable energy output and load, resulting in low operating efficiency.

[0004] It can be seen that how to reasonably control the operation of pumped storage units to adapt to the complex and changing power system environment after the access of new energy has become a technical problem that needs to be urgently solved by technical personnel in this field. Summary of the invention

[0005] The present invention provides an operation control method and system for a pumped storage unit, so as to solve the technical problem of how to reasonably control the operation of the pumped storage unit.

[0006] In order to solve the above technical problems, an embodiment of the present invention provides an operation control method of a pumped storage unit, comprising: respectively acquiring load data of a target area that matches a target pumped-storage unit, output data of new energy units in the target area, and future meteorological data of the target area; The output data of the new energy units and the future meteorological data are screened based on the mutual information method, and the output data of the target new energy units in the target area are predicted according to the multi-source prediction data set constructed by the screening results to obtain a new energy output prediction result; Determine each data subset of the load data in different dimensions, input each data subset into a prediction model of a corresponding dimension for prediction, and merge each prediction result to obtain a medium- and long-term load prediction result of the target area, wherein the data subset includes at least seasonal dimension data, linear trend dimension data and random data; Inputting the load data and the future meteorological data into a pre-built short-term load forecasting model to obtain a short-term load forecasting result for the target area; Based on the new energy output forecast result and the medium- and long-term load forecast result, determine the optimal reservoir water storage target curve of the target pumped-storage unit, and based on the new energy output forecast result and the short-term load forecast result, determine the optimal unit output result of the target pumped-storage unit; Execute the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result.

[0007] As one of the preferred solutions, the output data of the new energy units and the future meteorological data are screened based on the mutual information method, and a multi-source prediction data set constructed based on the screening results includes: Taking the output data of the new energy unit as the target variable and the future meteorological data as the characteristic variable, a characteristic matrix is ​​constructed; Calculate the mutual information values ​​between different meteorological data and the output data of the new energy unit according to the mutual information calculation method to obtain a mutual information value array; Traversing the mutual information value array, screening out meteorological data features whose mutual information value array is greater than a preset mutual information threshold; The screened future meteorological data features are combined with the output data of the new energy units to obtain the multi-source prediction data.

[0008] As one of the preferred solutions, predicting the output data of the target new energy units in the target area to obtain the new energy output prediction result includes: The pre-acquired historical multi-source prediction data is divided in chronological order to obtain a training data set, a validation data set, and a test data set; Construct an initial new energy output prediction model based on LSTM; The training data set is input into the initial new energy output prediction model for training. During the training process, the parameters of the initial new energy output prediction model are optimized using a particle swarm optimization algorithm and the verification data set, and the accuracy of the initial new energy output prediction model is evaluated according to the test data set to obtain a trained new energy output prediction model; In the actual prediction process, the acquired real-time multi-source prediction data is input into the new energy output prediction model for prediction to obtain the new energy output prediction result.

[0009] As one preferred solution, determining each data subset of the load data in different dimensions includes: Using the time of the load data as an index and the load value as a data column, converting the load data into time series data; The time series data is decomposed based on the STL decomposition method to obtain seasonal dimension data, trend dimension data and random data.

[0010] As one of the preferred solutions, each of the data subsets is input into a prediction model of a corresponding dimension for prediction, and each prediction result is integrated to obtain a medium- and long-term load prediction result of the target area, including: Predicting the seasonal dimension data based on a seasonal autoregressive integrated moving average model to obtain a seasonal prediction result; Predict the trend dimension data according to the linear regression model to obtain a trend prediction result; Predicting the random data based on an autoregressive integrated moving average model to obtain a random prediction result; The seasonal forecast results, the trend forecast results and the random forecast results are integrated to obtain medium- and long-term load forecast results.

[0011] As one preferred solution, the step of inputting the load data and the future meteorological data into a pre-built short-term load forecasting model to obtain a short-term load forecasting result for the target area includes: Extracting features from the load data and the future meteorological data to obtain a training data set; Construct an initial short-term load forecasting model based on LSTM; The training data set is input into the initial short-term load forecasting model for training. During the training process, the parameters of the initial short-term load forecasting model are optimized by a back propagation algorithm to obtain a trained short-term load forecasting model. In the actual prediction process, the real-time multi-source prediction data and real-time load data after feature extraction are input into the initial short-term load prediction model for prediction to obtain the short-term load prediction result.

[0012] As one of the preferred solutions, the optimal reservoir water storage target curve of the target pumped storage unit is determined based on the new energy output forecast result and the medium- and long-term load forecast result, including: The new energy output forecast result and the medium- and long-term load forecast result are input into a pre-constructed medium- and long-term optimization objective function for solution to obtain the optimal reservoir water storage target curve of the target pumped storage unit; wherein the medium- and long-term optimization objective function is expressed as: in, For the The new energy output predicted for each time period, For the The medium- and long-term load forecast value for each time period, For pumped storage units The power generation in the time period, It is The pumping cost for each period of time, It is The maintenance cost for a period of time, , , is the corresponding weight coefficient, It is the total number of time periods for medium- and long-term planning.

[0013] As one of the preferred solutions, the determining of the optimal unit output result of the target pumped-storage unit based on the new energy output forecast result and the short-term load forecast result includes: The new energy output forecast result and the short-term load forecast result are input into a pre-constructed short-term optimization objective function for solution to obtain the optimal unit output result of the target pumped storage unit; wherein the short-term optimization objective function is expressed as: The constraints of the short-term optimization objective function are expressed as: in, For the The power generation of the pumped storage unit in a certain period of time, For the The pumping power of the pumped storage unit in a certain period of time, Indicates that the pumped storage unit is Output per period, Indicates The forecast value of new energy output for each period, Indicates The short-term load forecast value for a period of time.

[0014] As one of the preferred solutions, after executing the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result, it also includes: Obtain the output data, load data and future meteorological data of new energy units in the area affected by the target pumped storage unit in real time; Update the new energy output forecast result, the medium- and long-term load forecast result, and the load forecast result according to the data obtained in real time; Obtaining an optimization result based on the optimal reservoir water storage target curve and the optimal unit output result according to the updated prediction result; According to the optimization result, the control strategy of the pumped storage unit is dynamically adjusted.

[0015] Another embodiment of the present invention provides an operation control system of a pumped storage unit, comprising: An acquisition module, used to respectively acquire load data of a target area matching a target pumped-storage unit, output data of new energy units in the target area, and future meteorological data of the target area; An output prediction module is used to screen the output data of the new energy unit and the future meteorological data based on the mutual information method, and predict the output data of the target new energy unit in the target area according to the multi-source prediction data set constructed by the screening results to obtain a new energy output prediction result; A first load forecasting module is used to determine each data subset of the load data in different dimensions, input each data subset and the future meteorological data into a prediction model of a corresponding dimension for prediction, and merge each prediction result to obtain a medium- and long-term load forecast result of the target area, wherein the data subset includes at least seasonal dimension data, linear trend dimension data and random data; A second load forecasting module, used for inputting the load data and the future meteorological data into a pre-built short-term load forecasting model to obtain a short-term load forecasting result for the target area; An analysis module, configured to determine an optimal reservoir water storage target curve of the target pumped-storage unit based on the new energy output forecast result and the medium- and long-term load forecast result, and to determine an optimal unit output result of the target pumped-storage unit based on the new energy output forecast result and the short-term load forecast result; An execution module is used to execute the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result.

[0016] Compared with the prior art, the embodiments of the present invention have the following advantages: (1) This application determines the optimal reservoir water storage target curve based on the new energy output forecast results and the medium- and long-term load forecast results, which can consider the changing trends of new energy power generation and load demand from a long-term perspective and reasonably plan the reservoir water storage capacity. More water can be stored during the period of large-scale new energy production, and more water can be released for power generation during peak loads and when the new energy output is insufficient, so as to optimize the utilization of water resources, improve the energy storage efficiency and regulation capacity of pumped storage power stations, and ensure the stable power supply of the power system.

[0017] (2) The scheme of this application enables pumped storage units to better match with new energy units and power loads, enhances the ability of the power system to cope with the uncertainty of new energy and load changes, improves the flexibility and adaptability of the power system, and helps to build a new power system with new energy as the main body, and promote the green and low-carbon transformation of energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flow chart of an operation control method of a pumped storage unit in one embodiment of the present invention; Figure 2 It is a schematic diagram of an operation control system of a pumped storage unit in one embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] In the description of this application, the terms "first", "second", "third", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", "third", etc. may explicitly or implicitly include one or more of the feature. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0021] In the description of the present application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" 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 it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. The terms "vertical", "horizontal", "left", "right", "upper", "lower" and similar expressions used herein are only for illustrative purposes, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0022] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those commonly understood by those skilled in the art. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by specific circumstances.

[0023] An embodiment of the present invention provides an operation control method for a pumped storage unit. For details, see Figure 1 , Figure 1 The flowchart of the operation control method of the pumped storage unit in one embodiment of the present invention is shown, including steps S1-S6: S1: respectively obtain the load data of the target area that matches the target pumped storage unit, the output data of the new energy unit in the target area, and the future meteorological data of the target area; Among them, load data can directly obtain real-time load data and historical load data of the target area through the data interface with the power system, such as the power grid energy management system (EMS), distribution management system (DMS), etc., which can provide load data information of different voltage levels and different regions, covering the power load conditions of various types of users such as industry, commerce, and residents.

[0024] Wind power, photovoltaic and other new energy power stations are equipped with their own monitoring systems. For example, the monitoring and data acquisition system (SCADA) of the wind farm can monitor the operating status, wind speed, wind direction, power generation and other data of each wind turbine in real time; the monitoring system of the photovoltaic power station can monitor the power generation, light intensity, temperature and other parameters of the photovoltaic panels. By establishing a communication connection with the monitoring systems of these new energy power stations, the real-time and historical output data of the new energy units can be directly obtained.

[0025] Future meteorological data can be obtained by establishing cooperation with local meteorological departments or professional meteorological service agencies and obtaining future meteorological data through the data interfaces provided by them.

[0026] S2: Screen the output data of new energy units and future meteorological data based on the mutual information method, and predict the output data of target new energy units in the target area according to the multi-source prediction data set constructed by the screening results to obtain the new energy output prediction result; Among them, the output data of new energy units reflects the actual power generation of new energy equipment such as wind power generation and photovoltaic power generation at a specific time point, and is the target of subsequent predictions. Future meteorological data contains a variety of meteorological elements, such as temperature, humidity, wind speed, wind direction, light intensity, etc. These meteorological conditions have a direct or indirect impact on the output of new energy units.

[0027] Preferably, in one embodiment of the present invention, the output data of new energy units and future meteorological data are screened based on the mutual information method, and a multi-source prediction data set constructed based on the screening results includes: The output data of new energy units is used as the target variable and the future meteorological data is used as the characteristic variable to construct the characteristic matrix; According to the mutual information calculation method, the mutual information values ​​between different meteorological data and the output data of new energy units are calculated to obtain a mutual information value array; Traversing the mutual information value array, filtering out meteorological data features whose mutual information value array is greater than a preset mutual information threshold; The screened future meteorological data characteristics are spliced ​​with the output data of new energy units to obtain multi-source prediction data.

[0028] Among them, mutual information is a concept in information theory, which is used to measure the correlation between two random variables. This application uses the mutual information calculation method to measure the degree of correlation between different meteorological data (characteristic variables) and the output data of new energy units (target variables). For each type of meteorological data, the mutual information value between it and the output data of the new energy unit is calculated. Assuming that there is a kind of meteorological data, a mutual information value array of length will be obtained in the end, and each element in the array represents the mutual information value between a kind of meteorological data and the output data of the new energy unit.

[0029] A mutual information threshold is preset, which is determined based on experience, experiments or data analysis to determine whether the correlation between meteorological data and the output data of new energy units is strong enough. Traverse the mutual information value array and check whether each mutual information value is greater than the preset mutual information threshold one by one. Filter out meteorological data features with mutual information values ​​greater than the threshold. These filtered meteorological data have a strong correlation with the output data of new energy units.

[0030] The screened future meteorological data features are spliced ​​with the output data of the new energy unit. In one embodiment of the present application, if the two meteorological data features of wind speed and light intensity are screened, then the two meteorological data and the output data of the new energy unit are arranged in chronological order to form a new data set, i.e., multi-source prediction data.

[0031] Preferably, in one embodiment of the present invention, the output data of the target new energy units in the target area are predicted to obtain the new energy output prediction result, including: The pre-acquired historical multi-source prediction data is divided in chronological order to obtain a training data set, a validation data set, and a test data set; Construct an initial new energy output prediction model based on LSTM; The training data set is input into the initial new energy output prediction model for training. During the training process, the particle swarm optimization algorithm and the validation data set are used to optimize the parameters of the initial new energy output prediction model, and the accuracy of the initial new energy output prediction model is evaluated based on the test data set to obtain a trained new energy output prediction model. In the actual prediction process, the acquired real-time multi-source prediction data is input into the new energy output prediction model for prediction to obtain the new energy output prediction result.

[0032] The historical multi-source forecast data within a certain period of time is obtained in advance. These data are constructed based on the above method and include key meteorological data related to the output of new energy units and corresponding output data. The historical multi-source forecast data is divided in chronological order.

[0033] We chose to build the initial new energy output prediction model based on the long short-term memory network (LSTM). LSTM is a special recurrent neural network (RNN) that can effectively handle long-term dependency problems in time series data and is suitable for predicting data that changes over time, such as the output of new energy units.

[0034] Determine the basic structure of the LSTM model, including the number of neurons in the input layer, hidden layer, and output layer, as well as parameters such as the number of layers in the network. For example, the number of neurons in the input layer can be determined based on the number of features of the multi-source prediction data, the hidden layer can be set with multiple LSTM units, and the number of neurons in the output layer is 1, which is used to output the predicted output value of the new energy unit.

[0035] The training data set is input into the initial renewable energy output prediction model for training. During the training process, the model will continuously adjust its parameters according to the input data to minimize the error between the predicted value and the actual value.

[0036] The particle swarm optimization algorithm (PSO) is used to optimize the parameters of the initial new energy output prediction model. PSO is an optimization algorithm based on swarm intelligence, which searches for the optimal solution through continuous search of particles in the solution space. In model training, the PSO algorithm can help find a set of optimal model parameters to make the model's prediction performance better.

[0037] During the training process, the validation data set is also used to evaluate the model. The validation data set does not participate in the training of the model, but is used to check the generalization ability of the model during the training process. If the error of the model on the validation data set begins to increase, it means that overfitting may have occurred. At this time, it is necessary to adjust the model parameters or take some measures to prevent overfitting, such as adding regularization terms.

[0038] The accuracy of the optimized initial new energy output prediction model is evaluated based on the test data set. The prediction results are obtained by inputting the data in the test data set into the model, and compared with the actual new energy unit output data, and various evaluation indicators such as mean square error (MSE) and mean absolute error (MAE) are calculated to determine the prediction accuracy of the model. After multiple adjustments and optimizations, a trained new energy output prediction model is obtained, which has high accuracy and generalization ability in predicting the output of new energy units.

[0039] In the actual prediction process, real-time multi-source prediction data are obtained. These data are also screened and constructed according to the above-mentioned mutual information method, and include key meteorological data at the prediction time and corresponding new energy unit output data (if there is real-time output data).

[0040] The acquired real-time multi-source forecast data is input into the trained new energy output forecast model for forecasting. The model processes the real-time data according to the learned patterns and rules, and finally outputs the predicted new energy output results. This result can provide an important reference for the dispatching and energy management of the power system, help to reasonably arrange power resources, and improve the stability and reliability of the power system.

[0041] S3: Determine each data subset of the load data in different dimensions, input each data subset into the prediction model of the corresponding dimension for prediction, and merge each prediction result to obtain the medium- and long-term load prediction result of the target area, wherein the data subset includes at least seasonal dimension data, linear trend dimension data and random data; Preferably, in one embodiment of the present invention, determining each data subset of the load data in different dimensions includes: Using the time of load data as the index and the load value as the data column, the load data is converted into time series data; Based on the STL decomposition method, the time series data is decomposed to obtain seasonal dimension data, trend dimension data and random data.

[0042] Among them, STL decomposition (Seasonal-Trend decomposition using Loess) is a commonly used time series decomposition method, which is based on the local weighted regression (Loess) algorithm. Decompose the converted time series data. Among them, seasonal dimension data reflects the repeated change pattern of load data within a fixed period. For example, in the power load, the daily peak and trough of electricity consumption show a certain periodicity, and this periodic change is reflected in the seasonal dimension data. Through the STL decomposition method, this periodic change feature can be accurately extracted.

[0043] Trend dimension data represents the overall change trend of load data over a long period of time. For example, with economic development, population growth or technological progress, the power load may show a gradual upward or downward trend, and this part of the trend information is included in the trend dimension data.

[0044] Random data is the unpredictable fluctuation part after removing seasonality and trend. This part of data may be affected by some unexpected factors, such as the surge in temporary electricity demand under extreme weather conditions, abnormal electricity consumption behavior of individual large industrial users, etc. The load fluctuations caused by these random factors are reflected in the random data. Through STL decomposition, the load data is completely decomposed into these three different dimensions of data subsets, providing a basis for more accurate subsequent predictions.

[0045] Preferably, in one embodiment of the present invention, each data subset is input into a prediction model of a corresponding dimension for prediction, and each prediction result is fused to obtain a medium- and long-term load prediction result of the target area, including: The seasonal dimension data is predicted based on the seasonal autoregressive integrated moving average model to obtain the seasonal prediction results; Predict the trend dimension data according to the linear regression model to obtain the trend prediction result; Predict random data based on the autoregressive integrated moving average model to obtain random prediction results; The seasonal forecast results, trend forecast results and random forecast results are integrated to obtain the medium- and long-term load forecast results.

[0046] Among them, the SARIMA model is a prediction model specifically used to process time series data with seasonal characteristics. The extracted seasonal dimension data is input into the SARIMA model. The model will learn the change rules of load in different seasons and time periods based on the historical seasonal data patterns. For example, for seasonal characteristics such as peak power consumption periods in summer and winter, and differences in power consumption between weekdays and weekends, the SARIMA model can predict the load changes in the corresponding time period in the future by learning from historical data, thereby obtaining seasonal prediction results.

[0047] The linear regression model is a simple and effective prediction model that is suitable for data with linear trends. The trend dimension data reflects the long-term change trend of the load, which can usually be approximated by a linear relationship. Input the trend dimension data into the linear regression model. The model will fit an optimal linear regression line based on the historical trend data, and predict future trend changes by extending this line. For example, if the power load has shown a trend of increasing at a certain rate every year in the past few years, the linear regression model can predict the load growth in the next few years based on this growth trend and obtain trend prediction results.

[0048] The ARIMA model is mainly used to process non-stationary time series data. It stabilizes the data by performing operations such as differentiation, and then uses autoregression and sliding average methods to model and predict. Although random data is unpredictable, it also has some statistical laws to a certain extent.

[0049] Input random data into the ARIMA model. The model will analyze the autocorrelation and partial autocorrelation characteristics of the random data, determine the appropriate model parameters, and predict the future fluctuations of the random data. Although it is impossible to accurately predict every random fluctuation, the ARIMA model can give a prediction value within a certain probability range to obtain a random prediction result.

[0050] The seasonal forecast results, trend forecast results and random forecast results obtained previously are fused. A common fusion method is simple weighted summation, which assigns a weight to each forecast result according to the reliability and importance of different forecast results. For example, if seasonal factors have a greater impact on the load, a higher weight can be assigned to the seasonal forecast result; if the trend forecast is relatively stable and reliable, its weight can also be appropriately increased.

[0051] The three forecast results are combined into a final medium- and long-term load forecast result by weighted summation. This result comprehensively considers the changes in load data in three dimensions: seasonality, trend, and randomness. Compared with a single model or a single-dimensional forecasting method, it can more comprehensively and accurately reflect the future medium- and long-term load change trend, and provide a more reliable decision-making basis for the planning, dispatching, and resource allocation of the power system.

[0052] S4: input the load data and future meteorological data into the pre-built short-term load forecasting model to obtain the short-term load forecasting result of the target area; Preferably, in one embodiment of the present invention, the load data and future meteorological data are input into a pre-built short-term load forecasting model to obtain a short-term load forecasting result for the target area, including: Extract features from load data and future meteorological data to obtain a training data set; Construct an initial short-term load forecasting model based on LSTM; The training data set is input into the initial short-term load forecasting model for training. During the training process, the parameters of the initial short-term load forecasting model are optimized by the back propagation algorithm to obtain a trained short-term load forecasting model. In the actual prediction process, the real-time multi-source prediction data and real-time load data after feature extraction are input into the initial short-term load prediction model for prediction to obtain the short-term load prediction results.

[0053] Specifically, the extracted load data features and future meteorological data features are integrated and organized according to certain formats and rules to form a training data set. The samples in the training data set are input into the initial short-term load forecasting model in sequence. The data enters from the input layer and is processed by the LSTM layer. The LSTM unit processes and memorizes the long-term dependencies in the data through the gating mechanism (input gate, forget gate, output gate), extracts features from the input data and models them, and finally obtains the prediction results through the output layer.

[0054] The prediction results of the model are compared with the actual load data labels, and the loss function is used to calculate the difference between the prediction results and the true value. Common loss functions include mean square error, mean absolute error, mean absolute percentage error, etc. According to the results of the loss function, the gradient of each parameter is calculated through the back propagation algorithm. The gradient represents the degree of influence of the parameter on the loss function. The parameters of the model are updated in the opposite direction of the gradient so that the value of the loss function gradually decreases. This process will continuously adjust the weights and biases in the LSTM network to optimize the performance of the model1. Repeat the process of forward propagation, loss calculation and back propagation, and train the model for multiple iterations. As the training progresses, the parameters of the model are continuously optimized, the value of the loss function gradually decreases, and the degree of fit of the model to the training data continues to improve until the preset stop conditions are met, such as reaching the maximum number of training rounds, the loss function value converges to a certain threshold, etc., to obtain a trained short-term load forecasting model.

[0055] Collect real-time load data and multi-source forecast data (including real-time meteorological data, etc.), and process these data according to the feature extraction method to extract the corresponding features. Input the real-time data after feature extraction into the trained short-term load forecasting model. The model processes and analyzes the input data based on the learned relationship between load data and meteorological data, and outputs the short-term load forecast results for the target area through the memory and prediction capabilities of the LSTM network.

[0056] S5: Based on the new energy output forecast results and the medium- and long-term load forecast results, determine the optimal reservoir water storage target curve of the target pumped storage unit, and based on the new energy output forecast results and the short-term load forecast results, determine the optimal unit output result of the target pumped storage unit; Preferably, in one embodiment of the present invention, based on the new energy output forecast result and the medium- and long-term load forecast result, determining the optimal reservoir water storage target curve of the target pumped storage unit includes: The new energy output forecast results and the medium- and long-term load forecast results are input into the pre-built medium- and long-term optimization objective function for solution, and the optimal reservoir water storage target curve of the target pumped storage unit is obtained; wherein, the medium- and long-term optimization objective function is expressed as: in, For the The new energy output predicted for each time period, For the The medium- and long-term load forecast value for each time period, For pumped storage units The power generation in the time period, It is The pumping cost for each period of time, It is The maintenance cost for a period of time, , , is the corresponding weight coefficient, It is the total number of time periods for medium- and long-term planning.

[0057] Specifically, the previously obtained new energy output forecast results and medium- and long-term load forecast results are substituted into the medium- and long-term optimization objective function. These data reflect the power generation situation of new energy and the changing trend of power load in the future.

[0058] Use a suitable optimization algorithm to solve the function. Common optimization algorithms include linear programming algorithms (such as the simplex method) and nonlinear programming algorithms (such as the gradient descent method, the quasi-Newton method, etc.). According to the power generation principle of the pumped storage unit and the relationship between the reservoir water volume and the power generation, the reservoir water storage corresponding to each time period is further determined. For example, if the power generation efficiency of the unit, the head height of the reservoir and other parameters are known, the change in reservoir water storage required to generate the corresponding power generation can be calculated through a physical model or empirical formula, thereby obtaining the optimal reservoir water storage target curve. This curve describes the optimal value that the reservoir water storage should reach in each time period within the medium and long-term planning time to achieve economic and stable operation of the power system.

[0059] Preferably, in one embodiment of the present invention, based on the new energy output forecast result and the short-term load forecast result, determining the optimal unit output result of the target pumped storage unit includes: The new energy output forecast results and short-term load forecast results are input into the pre-built short-term optimization objective function for solution to obtain the optimal unit output result of the target pumped storage unit; the short-term optimization objective function is expressed as: The constraints of the short-term optimization objective function are expressed as: in, For the The power generation of the pumped storage unit in a certain period of time, For the The pumping power of the pumped storage unit in a certain period of time, Indicates that the pumped storage unit is Output per period, Indicates The forecast value of new energy output for each period, Indicates The short-term load forecast value for a period of time.

[0060] Specifically, the new energy output forecast results and short-term load forecast results are substituted into the short-term optimization objective function and constraints. These real-time or recent forecast data can reflect the current and short-term new energy power generation and load demand of the power system. By solving the objective function to find a set of optimal pumped storage unit power generation and pumping power, the optimal unit output results of the target pumped storage unit in each time period are determined. According to these results, the power generation and pumping operations of the pumped storage unit in the short term can be directly controlled to achieve real-time balance and stable operation of the power system. For example, when the new energy output is insufficient and the load is high, the unit power generation is increased; when the new energy is large and the load is low, the pumping power is increased for energy storage.

[0061] S6: Execute the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result.

[0062] Based on the determined optimal reservoir water storage target curve and the optimal unit output results, the operation control strategy of the pumped storage unit is generated. This strategy will specify in detail how the pumped storage unit should adjust its power generation and pumping volume in each time period to follow the optimal reservoir water storage target curve and achieve the optimal unit output. For example, in a certain time period, according to the optimal unit output results, if the unit needs to generate electricity to meet the load demand, the strategy will specify the power output of the unit; at the same time, combined with the optimal reservoir water storage target curve, if the reservoir water storage is higher than the target value at this time, the strategy may arrange appropriate pumping operations to store excess water for subsequent power generation or to adjust the power balance.

[0063] The control system of the pumped storage power station will receive this operation control strategy and control the unit in real time according to the requirements of the strategy. The automation system of the power station will adjust the guide vane opening of the unit (for power generation conditions) or the speed of the pump (for pumping conditions) according to the strategy instructions to accurately control the power generation and pumping of the unit. At the same time, the relevant monitoring system will monitor the operating status of the unit in real time, including the power output of the unit, the change of reservoir water level, the operating parameters of the equipment, etc., to ensure that the unit operates stably according to the strategy requirements.

[0064] Preferably, in one embodiment of the present invention, after executing the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result, it also includes: Obtain the output data, load data and future meteorological data of new energy units in the area affected by the target pumped storage unit in real time; Update the new energy output forecast results, medium- and long-term load forecast results, and load forecast results based on the data obtained in real time; According to the updated prediction results, the optimal reservoir water storage target curve and the optimal unit output result are optimized to obtain the optimization result.

[0065] Specifically, the updated new energy output forecast results and medium- and long-term load forecast results are input into the medium- and long-term optimization objective function and solved again. Since the forecast results have changed, the optimal reservoir water storage target curve may also change accordingly. For example, if the updated new energy output forecast results show that the new energy power generation will increase significantly in the future, in order to better absorb the new energy power, it may be necessary to adjust the reservoir water storage capacity and increase pumped storage during the period of large new energy generation, so as to obtain the optimized optimal reservoir water storage target curve.

[0066] Substitute the updated renewable energy output forecast results and short-term load forecast results into the short-term optimization objective function and constraints, and solve them again. The changes in the power system reflected by real-time data may make the original optimal unit output result no longer optimal. For example, if the real-time load suddenly increases, and the renewable energy output does not change accordingly, it is necessary to readjust the unit's power generation and pumping volume to meet the balance of power supply and demand and achieve the optimal operating state, thereby obtaining the optimized optimal unit output result. Through continuous real-time data acquisition, forecast result update and optimization calculation, the operation control strategy of the pumped storage unit can always adapt to the dynamic changes of the power system, improving the stability, reliability and economy of the power system.

[0067] Another embodiment of the present invention provides an operation control system for a pumped storage unit. For details, see Figure 2 , Figure 2 The schematic diagram of the operation control system of the pumped storage unit in one embodiment of the present invention is shown, comprising: An acquisition module 11 is used to respectively acquire load data of a target area that matches a target pumped-storage unit, output data of a new energy unit in the target area, and future meteorological data of the target area; The output prediction module 12 is used to screen the output data of the new energy units and the future meteorological data based on the mutual information method, and predict the output data of the target new energy units in the target area according to the multi-source prediction data set constructed by the screening results to obtain the new energy output prediction result; The first load forecasting module 13 is used to determine each data subset of the load data in different dimensions, input each data subset and future meteorological data into the prediction model of the corresponding dimension for prediction, and merge each prediction result to obtain the medium- and long-term load prediction result of the target area, wherein the data subset includes at least seasonal dimension data, linear trend dimension data and random data; The second load forecasting module 14 is used to input the load data, the output data of the new energy units and the future meteorological data into the pre-built short-term load forecasting model to obtain the short-term load forecasting result of the target area; The analysis module 15 is used to determine the optimal reservoir water storage target curve of the target pumped storage unit based on the new energy output forecast result and the medium- and long-term load forecast result, and to determine the optimal unit output result of the target pumped storage unit based on the new energy output forecast result and the short-term load forecast result; The execution module 16 is used to execute the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result.

[0068] Compared with the prior art, the embodiments of the present invention have the following advantages: (1) This application determines the optimal reservoir water storage target curve based on the new energy output forecast results and the medium- and long-term load forecast results, which can consider the changing trends of new energy power generation and load demand from a long-term perspective and reasonably plan the reservoir water storage capacity. More water can be stored during the period of large-scale new energy production, and more water can be released for power generation during peak loads and when the new energy output is insufficient, so as to optimize the utilization of water resources, improve the energy storage efficiency and regulation capacity of pumped storage power stations, and ensure the stable power supply of the power system.

[0069] (2) The scheme of this application enables pumped storage units to better match with new energy units and power loads, enhances the ability of the power system to cope with the uncertainty of new energy and load changes, improves the flexibility and adaptability of the power system, and helps to build a new power system with new energy as the main body, and promote the green and low-carbon transformation of energy.

[0070] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for controlling the operation of a pumped storage unit, characterized in that: include: respectively acquiring load data of a target area that matches a target pumped-storage unit, output data of new energy units in the target area, and future meteorological data of the target area; The output data of the new energy units and the future meteorological data are screened based on the mutual information method, and the output data of the target new energy units in the target area are predicted according to the multi-source prediction data set constructed by the screening results to obtain a new energy output prediction result; Determine each data subset of the load data in different dimensions, input each data subset into a prediction model of a corresponding dimension for prediction, and merge each prediction result to obtain a medium- and long-term load prediction result of the target area, wherein the data subset includes at least seasonal dimension data, linear trend dimension data and random data; Inputting the load data and the future meteorological data into a pre-built short-term load forecasting model to obtain a short-term load forecasting result for the target area; Based on the new energy output forecast result and the medium- and long-term load forecast result, determine the optimal reservoir water storage target curve of the target pumped-storage unit, and based on the new energy output forecast result and the short-term load forecast result, determine the optimal unit output result of the target pumped-storage unit; Execute the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result.

2. The operation control method of the pumped storage unit according to claim 1, characterized in that: The output data of the new energy generating units and the future meteorological data are screened based on the mutual information method, and a multi-source prediction data set is constructed based on the screening results, including: Taking the output data of the new energy unit as the target variable and the future meteorological data as the characteristic variable, a characteristic matrix is ​​constructed; Calculate the mutual information values ​​between different meteorological data and the output data of the new energy unit according to the mutual information calculation method to obtain a mutual information value array; Traversing the mutual information value array, screening out meteorological data features whose mutual information value array is greater than a preset mutual information threshold; The screened future meteorological data features are combined with the output data of the new energy units to obtain the multi-source prediction data.

3. The operation control method of the pumped storage unit according to claim 1, characterized in that: The step of predicting the output data of the target new energy units in the target area to obtain the new energy output prediction result includes: The pre-acquired historical multi-source prediction data is divided in chronological order to obtain a training data set, a validation data set, and a test data set; Construct an initial new energy output prediction model based on LSTM; The training data set is input into the initial new energy output prediction model for training. During the training process, the parameters of the initial new energy output prediction model are optimized using a particle swarm optimization algorithm and the verification data set, and the accuracy of the initial new energy output prediction model is evaluated according to the test data set to obtain a trained new energy output prediction model; In the actual prediction process, the acquired real-time multi-source prediction data is input into the new energy output prediction model for prediction to obtain the new energy output prediction result.

4. The operation control method of the pumped storage unit according to claim 1, characterized in that: The determining of each data subset of the load data in different dimensions includes: Using the time of the load data as an index and the load value as a data column, converting the load data into time series data; The time series data is decomposed based on the STL decomposition method to obtain seasonal dimension data, trend dimension data and random data.

5. The operation control method of the pumped storage unit according to claim 1, characterized in that: The step of inputting each of the data subsets into a prediction model of a corresponding dimension for prediction, and fusing the prediction results to obtain a medium- and long-term load prediction result for the target area includes: Predicting the seasonal dimension data based on a seasonal autoregressive integrated moving average model to obtain a seasonal prediction result; Predict the trend dimension data according to the linear regression model to obtain a trend prediction result; Predicting the random data based on an autoregressive integrated moving average model to obtain a random prediction result; The seasonal forecast results, the trend forecast results and the random forecast results are integrated to obtain medium- and long-term load forecast results.

6. The operation control method of the pumped storage unit according to claim 1, characterized in that: The step of inputting the load data and the future meteorological data into a pre-built short-term load forecasting model to obtain a short-term load forecasting result for the target area includes: Extracting features from the load data and the future meteorological data to obtain a training data set; Construct an initial short-term load forecasting model based on LSTM; The training data set is input into the initial short-term load forecasting model for training. During the training process, the parameters of the initial short-term load forecasting model are optimized by a back propagation algorithm to obtain a trained short-term load forecasting model. In the actual prediction process, the real-time multi-source prediction data and real-time load data after feature extraction are input into the initial short-term load prediction model for prediction to obtain the short-term load prediction result.

7. The operation control method of the pumped storage unit according to claim 1, characterized in that: The determining of the optimal reservoir water storage target curve of the target pumped-storage unit based on the new energy output forecast result and the medium- and long-term load forecast result includes: The new energy output forecast result and the medium- and long-term load forecast result are input into a pre-constructed medium- and long-term optimization objective function for solution to obtain the optimal reservoir water storage target curve of the target pumped storage unit; wherein the medium- and long-term optimization objective function is expressed as: in, For the The new energy output predicted for each time period, For the The medium- and long-term load forecast value for each time period, For pumped storage units The power generation in the time period, It is The pumping cost for each period of time, It is The maintenance cost for a period of time, , , is the corresponding weight coefficient, It is the total number of time periods for medium- and long-term planning.

8. The operation control method of the pumped storage unit according to claim 1, characterized in that: The determining the optimal unit output result of the target pumped-storage unit based on the new energy output forecast result and the short-term load forecast result includes: The new energy output forecast result and the short-term load forecast result are input into a pre-constructed short-term optimization objective function for solution to obtain the optimal unit output result of the target pumped storage unit; wherein the short-term optimization objective function is expressed as: The constraints of the short-term optimization objective function are expressed as: in, For the The power generation of the pumped storage unit in a certain period of time, For the The pumping power of the pumped storage unit in a certain period of time, Indicates that the pumped storage unit is The output of each period, Indicates The forecast value of new energy output for each period, Indicates The short-term load forecast value for a period of time.

9. The operation control method of the pumped storage unit according to claim 1, characterized in that: After executing the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result, the method further includes: Obtain the output data, load data and future meteorological data of new energy units in the area affected by the target pumped storage unit in real time; Update the new energy output forecast result, the medium- and long-term load forecast result, and the load forecast result according to the data obtained in real time; Obtaining an optimization result based on the optimal reservoir water storage target curve and the optimal unit output result according to the updated prediction result; According to the optimization result, the control strategy of the pumped storage unit is dynamically adjusted.

10. An operation control system of a pumped storage unit, characterized in that: include: An acquisition module, used to respectively acquire load data of a target area matching a target pumped-storage unit, output data of new energy units in the target area, and future meteorological data of the target area; An output prediction module is used to screen the output data of the new energy unit and the future meteorological data based on the mutual information method, and predict the output data of the target new energy unit in the target area according to the multi-source prediction data set constructed by the screening results to obtain a new energy output prediction result; A first load forecasting module is used to determine each data subset of the load data in different dimensions, input each data subset and the future meteorological data into a prediction model of a corresponding dimension for prediction, and merge each prediction result to obtain a medium- and long-term load forecast result of the target area, wherein the data subset includes at least seasonal dimension data, linear trend dimension data and random data; A second load forecasting module, used for inputting the load data and the future meteorological data into a pre-built short-term load forecasting model to obtain a short-term load forecasting result for the target area; An analysis module, configured to determine an optimal reservoir water storage target curve of the target pumped-storage unit based on the new energy output forecast result and the medium- and long-term load forecast result, and to determine an optimal unit output result of the target pumped-storage unit based on the new energy output forecast result and the short-term load forecast result; An execution module is used to execute the pumped storage unit operation control strategy generated based on the optimal reservoir water storage target curve and the optimal unit output result.

Images