A new energy power distribution strategy decision-making method based on deep learning

Abstract

The present invention discloses a new energy power distribution strategy decision-making method based on deep learning, which belongs to the field of energy management and distribution technology, including: acquiring a new energy power data set through big data technology; performing standardized preprocessing on the new energy power data set through a data preprocessing program; constructing a new energy power transaction prediction model using a deep learning network; determining the new energy power stage prediction result based on the new energy power transaction prediction model; obtaining the power distribution target, and establishing a new energy power time period allocation function based on the new energy power stage prediction result; obtaining a new energy power distribution plan cluster based on the new energy power stage prediction result, and evaluating and iteratively optimizing the new energy power distribution plan cluster based on the new energy power time period allocation function as the target new energy power time period allocation strategy. This application solves the technical problems of insufficient accuracy and optimization of new energy power distribution strategies in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of energy management and distribution technology, and in particular to a new energy power distribution strategy decision-making method based on deep learning. Background Art

[0002] In recent years, with increasing global energy demand and heightened awareness of environmental protection, new energy technologies have rapidly developed and been widely adopted. Renewable energy sources such as wind, solar, and hydropower have gradually gained a significant position in power systems. However, the intermittent and fluctuating nature of these new energy sources poses new challenges to the stable operation of power systems.

[0003] To better meet user needs, promote economic development, and improve energy efficiency, the power market needs to optimize resource allocation and efficient utilization. However, traditional power distribution strategies are often based on static data and empirical judgment, lacking intelligent and dynamic adjustment and optimization capabilities.

[0004] In summary, this application aims to solve the technical problems of insufficient accuracy and optimization of new energy power distribution strategies in the existing technology. Summary of the Invention

[0005] This application provides a new energy power distribution strategy decision-making method based on deep learning, aiming to solve the technical problems of insufficient accuracy and optimization of new energy power distribution strategies in the existing technology.

[0006] In view of the above problems, this application provides a new energy power distribution strategy decision-making method based on deep learning.

[0007] One aspect disclosed in the present application provides a new energy power distribution strategy decision-making method based on deep learning, the method comprising collecting and acquiring a new energy power data set through big data technology, the new energy power data set comprising new energy time-series power generation information, power market basic information, power generation-side transaction information, medium- and long-term and spot transaction time-series data, and grid-side power shortage data; obtaining a data preprocessing program, performing standardized preprocessing on the new energy power data set through the data preprocessing program to obtain a new energy standard power data set; using a deep learning network to train and optimize the new energy standard power data set to construct a new energy power transaction prediction model; performing power prediction for a preset time window based on the new energy power transaction prediction model to determine a new energy power stage prediction result; obtaining a power distribution target, optimizing weight allocation for the power distribution target based on the new energy power stage prediction result, and establishing a new energy power time period allocation function; obtaining a new energy power distribution scheme cluster based on the new energy power stage prediction result, and evaluating and iteratively optimizing the new energy power distribution scheme cluster based on the new energy power time period allocation function to obtain an optimal power distribution scheme as a target new energy power time period allocation strategy.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] Using big data technology, a new energy power dataset is collected, including time-series information on renewable energy generation, basic electricity market information, power generation-side transaction information, mid- and long-term spot transaction time-series data, and grid-side power shortage data. A data preprocessing program is then used to standardize and preprocess the dataset to obtain a new energy standard power dataset. Based on this, a deep learning network is used to train and optimize the new energy standard power dataset to construct a new energy power transaction prediction model. Based on this prediction model, power demand and supply within a preset time window are predicted to determine a new energy power phase forecast result. Next, power allocation targets are obtained and, based on the new energy power phase forecast result, optimized weights are assigned to the power allocation targets to establish a new energy power time allocation function. Based on the forecast results, multiple new energy power allocation plan clusters are generated. These plans are evaluated and iteratively optimized using the new energy power time allocation function to ultimately obtain the optimal power allocation plan, which serves as the target new energy power time allocation strategy. This approach addresses the technical issues of insufficient accuracy and optimization of new energy power allocation strategies in existing technologies, effectively improving the accuracy and optimization efficiency of new energy power allocation, and ensuring the stable operation and economic benefits of the power system.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A flowchart of a new energy power distribution strategy decision-making method based on deep learning is provided for an embodiment of the present application.

[0012] Figure 2 A flowchart of constructing a new energy power transaction prediction model in a new energy power distribution strategy decision-making method based on deep learning is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0013] The overall idea of the technical solution provided by this application is as follows:

[0014] The embodiment of the present application provides a new energy power distribution strategy decision-making method based on deep learning. First, a new energy power data set including new energy time series power generation, power market basic information, power generation side transaction information, medium- and long-term and spot transaction time series data, and grid side power shortage data is collected through big data technology. Then, the data set is standardized through a data preprocessing program to obtain a new energy standard power data set. The data set is trained and optimized using a deep learning network to construct a new energy power transaction prediction model. Based on the prediction model, the power demand and supply situation within a preset time window is predicted to determine the new energy power stage prediction result. The power distribution target is obtained, and the weight distribution is optimized based on the prediction result to establish a new energy power time period allocation function. Finally, the optimal power distribution plan is obtained by evaluating and iteratively optimizing multiple power distribution plans generated as the target new energy power time period allocation strategy.

[0015] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically introduced in conjunction with the drawings in the specification.

[0016] Example

[0017] like Figure 1 As shown, an embodiment of the present application provides a new energy power distribution strategy decision-making method based on deep learning, the method comprising:

[0018] Step S100: acquiring a new energy power data set through big data technology, wherein the new energy power data set includes new energy time-series power generation information, power market basic information, power generation side transaction information, medium- and long-term and spot transaction time-series data, and grid-side power shortage data.

[0019] Specifically, big data technology is used to collect new energy power data sets from various sources (such as power monitoring systems, market trading platforms, etc.).

[0020] Renewable energy generation information refers to time-series data recording the temporal changes in renewable energy generation, such as wind and photovoltaic power generation. This data comes from monitoring systems and sensors at power generation facilities like wind farms, photovoltaic power plants, and hydropower stations. This data exhibits time-series characteristics, changing over time and reflecting the dynamics of renewable energy generation.

[0021] Basic electricity market information includes basic information such as electricity market prices, supply and demand, etc. Data sources include power trading centers and power market operation platforms. This provides market context and serves as a reference for power allocation decisions.

[0022] Power generation transaction information includes transaction data from various trading markets. It originates from power generation companies' transaction records and power trading platforms. It reflects market behavior and trading activity on the power generation side and is crucial data for developing power generation strategies.

[0023] The medium- and long-term and spot transaction time series data includes transaction time series data for medium- and long-term contracts and the spot market. It originates from power trading markets and contract record systems. It can provide long- and short-term transaction information, helping to predict future market behavior and price fluctuations.

[0024] Grid-side power shortage data refers to the data on power shortages at the user end. It is sourced from the power company's customer service system, smart meter data, and user feedback. It reflects demand and emergency situations at the user end and serves as a crucial basis for adjusting power distribution.

[0025] Based on this step, a comprehensive, accurate, and real-time updated new energy power data set can be constructed, providing a solid data foundation for subsequent power distribution strategy decisions.

[0026] Step S200: obtaining a data preprocessing program, and performing standardization preprocessing on the new energy power data set through the data preprocessing program to obtain a new energy standard power data set.

[0027] Specifically, data preprocessing is a key step in data analysis and machine learning. Its purpose is to convert raw data into a format suitable for model training and analysis. For renewable energy power datasets, preprocessing steps include data cleaning, normalization, feature extraction, and other operations.

[0028] The new energy standard power dataset processed in this step can effectively support the training and optimization of deep learning models, improving the model's predictive capabilities and decision-making effects.

[0029] Step S300: Using a deep learning network to train and optimize the new energy standard power data set to build a new energy power transaction prediction model.

[0030] Specifically, after completing data preprocessing, the standardized new energy power data set is trained and optimized using a deep learning network to build a model that can effectively predict new energy power transactions.

[0031] First, select an appropriate deep learning model to build a prediction model. Common deep learning models include convolutional neural networks (CNNs), recurrent neural networks (RNNs), long short-term memory networks (LSTMs), and transformer models. This application uses long short-term memory networks (LSTMs), an improved version of RNNs that can effectively handle long-term dependencies.

[0032] Next, we designed a deep learning model architecture suitable for the characteristics of renewable energy power data. The following is an example of a typical LSTM model architecture: The input layer accepts a standardized renewable energy power dataset. The LSTM layer is used to capture long-term dependencies in time series data. The fully connected layer (Dense Layer) connects the output of the LSTM layer to the fully connected layer for feature combination and mapping. The output layer predicts electricity trading volume in the future time window.

[0033] Next, the new energy standard electricity dataset is divided into training, validation, and test sets for model training and evaluation. An appropriate loss function (such as mean squared error) and optimizer (such as Adam) are selected for model compilation. The model is trained using the training data, and hyperparameters are tuned on the validation set to prevent overfitting. Model performance is evaluated on the test set to ensure good generalization.

[0034] Finally, adjust the model's hyperparameters (such as learning rate, batch size, number of layers, and number of units) to optimize model performance. Use regularization techniques (such as L2 regularization and dropout) to prevent overfitting and improve the model's generalization ability. Combine the predictions of multiple models (such as model ensembles and voting methods) to improve prediction accuracy and stability.

[0035] Based on this step, a deep learning network can be used to train and optimize the new energy standard power dataset, building an effective new energy power trading forecasting model. This model can accurately predict future power trading volumes, providing a scientific basis for power system scheduling and resource optimization, and improving system efficiency and stability.

[0036] Step S400: performing power forecasting for a preset time window based on the new energy power transaction forecasting model to determine a new energy power stage forecasting result.

[0037] Specifically, after building and optimizing a new energy power trading forecasting model, the model is used to predict future power demand and supply. The preset time window is determined based on specific needs, such as the next hour, day, or week. The choice of time window depends on the specific application scenario and decision-making requirements.

[0038] Acquire real-time data from the power system, including the latest renewable energy generation, meteorological data, power market data, etc. Integrate real-time data with historical data to form a complete input data set for the model to make predictions.

[0039] Normalize the live data to meet the model input requirements. Use the same normalization methods (e.g., mean and standard deviation) used when training the model. Extract the required features from the live data, such as temporal features (e.g., hour, day of the week, season, etc.) and statistical features.

[0040] The preprocessed data is fed into a trained renewable energy power trading forecasting model. The model calculates the input data and outputs forecasts. These forecasts include power demand and supply within a pre-set time window.

[0041] Based on the model's forecast results, determine the phased trends and specific values of renewable energy power. For example, forecast hourly power generation and demand for the next day. Evaluate the uncertainty of the forecast results, for example, using confidence intervals or prediction intervals to indicate the reliability of the forecast results.

[0042] Use forecast results to guide power dispatch decisions. For example, adjust power generation plans and load management strategies based on forecasted power demand and supply. Optimize the allocation and use of new energy resources to improve the operational efficiency and economic benefits of the power system. Use forecast results to proactively identify potential power shortages or surpluses and take appropriate preventative measures.

[0043] Based on this step, it is possible to make power forecasts for the preset time window based on the new energy power trading forecast model, determine the new energy power stage forecast results, and provide a scientific basis for the scheduling and optimization of the power system.

[0044] Step S500: obtaining a power allocation target, optimizing weight allocation for the power allocation target based on the new energy power stage prediction result, and establishing a new energy power time period allocation function.

[0045] Specifically, after completing the power forecast and obtaining the new energy power stage forecast results, the power allocation target is obtained, and the weight distribution is optimized based on the forecast results to establish the new energy power period allocation function.

[0046] Among them, the power distribution targets include power demand targets, power generation targets, system stability targets, and economic targets. The power demand target refers to determining the demand targets of the power system in various time periods, including basic demand, peak demand, and peak-shaving demand. The power generation target refers to determining the power generation targets of each new energy power generation unit, including power generation, output rate, power generation cost, etc. The system stability target refers to ensuring the stable operation of the power system, including frequency stability, voltage stability, and supply and demand balance. The economic target refers to economic targets such as optimizing power generation costs, reducing wind and solar power curtailment, and improving the utilization rate of new energy.

[0047] Furthermore, based on the predicted power demand data, the power demand situation in each time period is analyzed, including peak period and valley period, etc. Based on the predicted renewable energy power generation data, the power generation capacity in each time period is analyzed, including power generation fluctuation and intermittency, etc.

[0048] Optimization weight assignment requires a clear definition of the specific optimization objective, such as increasing renewable energy utilization, reducing power generation costs, and ensuring system stability. Different weights are assigned based on the importance of each optimization objective. For example, different weights can be assigned to economic efficiency, stability, and environmental protection to reflect their priority within the overall optimization. For example, system stability could have a weight of 0.4, renewable energy utilization 0.3, economic efficiency 0.2, and environmental protection 0.1. Based on real-time operational status and forecast results, the weights of each objective can be dynamically adjusted to accommodate changes in power demand and generation.

[0049] Finally, based on the optimized weighted allocation results, a new energy power time allocation function is established. This function takes the predicted value of new energy power as input and calculates the power allocation for each time period based on the weighted allocation results. The allocation function can consider various factors, such as power demand forecasts, power network transmission capacity, and power market pricing mechanisms, to ensure the rationality and effectiveness of power allocation.

[0050] The new energy power time allocation function can be expressed as:

[0051] Objective function = ∑ i w i ·f i (x);

[0052] Among them, w i is the weight of the i-th target, f i (x) is a function of the i-th objective, and x is the decision variable. Determine the constraints that must be met during the allocation process, such as power balance constraints, generation capacity constraints, and line transmission capacity constraints. Use an optimization algorithm (such as linear programming, nonlinear programming, or genetic algorithms) to solve the objective function and obtain the optimal power allocation solution.

[0053] Based on this step, the prediction results can be effectively used to optimize the weight distribution of power distribution targets, establish a new energy power time allocation function, and achieve the optimal power distribution strategy.

[0054] Step S600: Obtain a new energy power allocation scheme cluster based on the new energy power stage prediction result, evaluate and iteratively optimize the new energy power allocation scheme cluster based on the new energy power period allocation function, and obtain the optimal power allocation scheme as the target new energy power period allocation strategy.

[0055] Specifically, based on the renewable energy power phase forecast results, a preliminary set of power allocation scenarios is generated. These scenarios consider different generation and load management strategies to meet the forecasted power demand and supply. A scenario cluster is constructed, which is a collection of multiple candidate power allocation scenarios. Each scenario has different power allocation strategies for different time periods to maximize renewable energy utilization, minimize costs, and ensure system stability.

[0056] Then, based on the renewable energy power time allocation function, the fitness of each allocation scheme is calculated. The fitness function comprehensively considers the weights of various allocation objectives and evaluates each scheme's performance against these different objectives. Evaluation criteria include system stability, renewable energy utilization, economic efficiency, and environmental benefits. These evaluation criteria are used to judge the merits of each scheme.

[0057] The initially generated cluster of solutions is evaluated, and the fitness value of each solution is calculated. Based on the fitness values, the optimal solutions are initially screened. These selected solutions are then optimized and adjusted. Adjustments may be made to improve renewable energy utilization, reduce power generation costs, and enhance system stability. The solutions are then improved using optimization algorithms (such as genetic algorithms and particle swarm optimization). The improved solutions are re-evaluated and recalculated to obtain new fitness values. Continuous iterative optimization and evaluation are performed to gradually approach the optimal solution. Termination conditions for the iterative optimization are set, such as reaching a preset fitness threshold or convergence after a certain number of iterations. Based on the final evaluation results, the solution with the highest fitness value is selected as the optimal power allocation solution. This solution, after comprehensively considering various objectives, achieves the best balance among all aspects. The optimal power allocation solution is used as the target renewable energy power time allocation strategy. This strategy will be used in actual power allocation decisions for different time windows, guiding the scheduling and optimization of the power system.

[0058] Based on this step, the optimal power distribution scheme can be evaluated and optimized from the generated multiple power distribution schemes, realizing an efficient, stable, economical and environmentally friendly new energy power distribution strategy.

[0059] Furthermore, the method described in this application also includes:

[0060] Abnormal data identification is performed on the new energy power data set to obtain an abnormal power data set, where the abnormal power data set includes inconsistent data, missing data, and out-of-bounds data; the abnormal power data set is mapped and matched with the data preprocessing program to determine an abnormal data preprocessing matching program; based on the abnormal data preprocessing matching program, the inconsistent data, missing data, and out-of-bounds data are preprocessed respectively to obtain consistent power data, complete power data, and standardized power data; and the new energy standard power data set is obtained based on the consistent power data, complete power data, and standardized power data.

[0061] Specifically, identify inconsistent data within a dataset. This data may be caused by different formats or recording standards across different data sources. For example, inconsistent timestamp formats or numerical units. Also, identify missing data within a dataset. This may be due to sensor failure, data transmission errors, or other reasons. Also, identify data points outside of a reasonable range. This data may be caused by collection errors or unusual events. For example, a wind speed sensor reading may be unusually high or low.

[0062] Map and match the identified abnormal data with the preprocessing procedures to determine the preprocessing methods for different types of abnormal data. Design specific preprocessing procedures for different types of abnormal data to ensure the consistency, integrity, and standardization of the preprocessed data.

[0063] Preprocess inconsistent data. For example, unify the timestamp format and convert data in different units to a unified unit to ensure data consistency. Fill in missing data. Common methods include: using the mean of the data to fill missing values; using the median of the data to fill missing values; using linear interpolation or other interpolation methods to fill missing values; using regression models or other machine learning methods to predict and fill missing values. Process out-of-bounds data. Common methods include: directly deleting out-of-bounds data and correcting out-of-bounds data to values within a reasonable range. For example, correct an abnormally high wind speed reading to a reasonable maximum wind speed value. Replace out-of-bounds data with normal data from adjacent time points.

[0064] Finally, consistent, complete, and standardized power data are integrated to form a standardized new energy power dataset. This effectively handles abnormal data within the new energy power dataset, resulting in a high-quality, standardized new energy power dataset, providing a reliable data foundation for subsequent deep learning model training and optimization.

[0065] Further, such as Figure 2 As shown, the method described in this application also includes:

[0066] The new energy standard power data set is labeled with power supply and demand conditions, power generation and transaction electricity prices to obtain a power supply and demand sample set, a power generation sample set and a power transaction electricity price sample set; based on a deep learning network, the power supply and demand sample set, the power generation sample set and the power transaction electricity price sample set are subjected to cyclic supervised training to generate a power supply and demand prediction model, a power generation prediction model and a power transaction electricity price prediction model; the power supply and demand prediction model, the power generation prediction model and the power transaction electricity price prediction model are merged to generate an initial power transaction prediction model; the initial power transaction prediction model is verified and optimized to construct the new energy power transaction prediction model.

[0067] Specifically, based on the historical and real-time data in the new energy standard power data set, the power supply and demand situation is identified. This includes power demand and load changes. A power supply and demand situation sample set is generated, which contains power demand and supply data for different time periods. The historical power generation data of each new energy power generation unit, including wind power and photovoltaic power, is identified. A power generation sample set is generated, which contains power generation data for each new energy power generation unit in different time periods. Transaction electricity price information in the power market is identified, including spot market electricity prices and medium- and long-term transaction electricity prices. A power transaction price sample set is generated, which contains power transaction price data for different time periods.

[0068] Then, a long short-term memory (LSTM) network was trained on a sample set of electricity supply and demand to develop a power supply and demand forecasting model. During the training process, historical power demand data was input to predict power demand for future time periods. A similar method was used to train a sample set of power generation to develop a power generation forecasting model. By inputting historical power generation data and meteorological data, renewable energy power generation for future time periods was forecast. A sample set of power trading prices was trained to develop a power trading price forecasting model. By inputting historical power price data and market supply and demand conditions, power trading prices for future time periods were forecast.

[0069] The output results of the three models are spliced together to form a comprehensive input feature matrix. For example, assuming that the output of each model is the predicted value for the next 24 hours, the final feature matrix will contain three 24-dimensional vectors (power demand, power generation, and power trading price). A new deep learning model is designed to process the spliced feature matrix. Common fusion model structures include: fully connected neural networks (DNNs), which perform further nonlinear transformations and combinations on the spliced features. Time series models, such as LSTM or Transformer, can capture the temporal dependencies between features. The fusion model is trained using historical data with the goal of minimizing the prediction error. During the training process, the input is the spliced feature matrix, and the output is the actual power trading data (such as the actual supply and demand balance, power prices, etc.).

[0070] The initial power trading forecast model was validated using a validation set to evaluate its prediction accuracy and generalization capabilities. The model's predictions were compared with actual data to calculate the model's error and accuracy. Based on the validation results, the initial power trading forecast model was optimized. Optimization methods included adjusting model parameters, improving the training algorithm, and increasing the amount of training data. The optimized model was used to re-predict and re-validate to ensure higher prediction accuracy and reliability.

[0071] Based on this step, the power supply and demand forecast model, power generation forecast model and power transaction price forecast model can be effectively integrated to generate a high-precision new energy power transaction forecast model, providing strong support for the optimization and scheduling of the power system.

[0072] Furthermore, the method described in this application also includes:

[0073] A recurrent neural network structure is selected according to a deep learning network, and an initial hidden layer value and a model weight matrix of the recurrent neural network structure are obtained; based on the initial hidden layer value and the model weight matrix, hidden layer calculations are performed on the power supply and demand sample set, the power generation sample set, and the power trading price sample set, respectively, to obtain current hidden layer values; based on the current hidden layer value and the model weight matrix, iterative calculation training is performed on the power supply and demand sample set, the power generation sample set, and the power trading price sample set, respectively, to generate the power supply and demand prediction model, the power generation prediction model, and the power trading price prediction model.

[0074] Specifically, select a suitable RNN structure, such as a long short-term memory network (LSTM), and initialize the RNN's hidden layer values (usually zero) and model weight matrix before training begins. The weight matrix can be initialized using a random initialization method (such as Xavier initialization). The power supply and demand sample set, power generation sample set, and power transaction price sample set are used as input data and input into the RNN model in sequence. At each time step, the data xt of the current time step is input and combined with the hidden layer value h of the previous time step. t-1 , calculate the hidden layer value ht and output value of the current time step. The update formula is:

[0075] ht=σ(W h ·h t-1 +W x ·X t +b h );

[0076] Among them, W h is the hidden layer weight matrix, W x is the input weight matrix, b h is the bias term, and σ is the activation function (such as tanh or ReLU). Calculate the data for each time step to obtain the hidden layer value h of the current time step t .

[0077] Iterative training is performed using the backpropagation through time (BPTT) algorithm, performing forward computation and backward propagation on the data at each time step to update the model's weight matrix. At each iteration, the loss function (such as mean squared error (MSE)) is calculated, and the gradient is computed. The weight matrix and bias terms are updated to gradually approach the optimal prediction performance. After multiple iterative training, models for predicting power supply and demand, power generation, and power trading prices are generated. Each model accurately predicts relevant data for future time periods based on the input time series data.

[0078] This step uses deep learning networks to generate models for predicting power supply and demand, power generation, and power trading prices. These models can accurately predict relevant data for future time periods, providing a scientific basis for decision-making on renewable energy power allocation strategies.

[0079] Furthermore, the method described in this application also includes:

[0080] The initial power trading prediction model is evaluated and verified using a model loss function to determine a model performance loss parameter; a gradient calculation is performed on the cyclic model parameters of the initial power trading prediction model based on the model performance loss parameter to obtain prediction model parameter gradient information; the cyclic model parameters are iteratively updated based on the prediction model parameter gradient information using a back propagation algorithm to obtain cyclic model update parameters; the initial power trading prediction model is optimized and configured based on the cyclic model update parameters to obtain the new energy power trading prediction model.

[0081] Specifically, the initial power trading forecast model is evaluated using a loss function. Common loss functions include mean squared error (MSE) and mean absolute error (MAE). By calculating the model's prediction error on the validation set, the loss parameter is determined to reflect the model's performance.

[0082] Based on the model's performance loss parameters, the gradient information of the initial power trading prediction model is calculated. The gradient information is used to guide the update of model parameters to ensure that the model is optimized in the direction of reducing losses.

[0083] The backpropagation algorithm is used to iteratively update the model parameters based on the calculated gradient information. The backpropagation algorithm calculates the contribution of each parameter to the loss using the chain rule and updates the parameters based on the learning rate.

[0084] The hidden layer parameters and weight matrix of the recurrent neural network are updated to obtain new recurrent model parameters. In each iteration, the updated parameters gradually approach the optimal solution, which continuously improves the prediction accuracy of the model.

[0085] Using the updated recurrent model parameters, the initial power trading forecasting model was optimized. Through multiple iterative updates and optimizations, the model's performance on the validation set gradually improved. After sufficient training and optimization, the final renewable energy power trading forecasting model was constructed. This model comprehensively predicts power supply and demand, power generation, and trading prices for future time periods, providing a scientific basis for power allocation strategy decisions.

[0086] This step uses the model loss function to evaluate and validate the initial power trading forecast model, calculates gradient information, and iteratively updates model parameters through the backpropagation algorithm. Ultimately, the optimized configuration results in a highly accurate new energy power trading forecast model. This model can accurately predict future power supply and demand, power generation, and trading prices, providing reliable decision support for new energy power allocation strategies.

[0087] Furthermore, the method described in this application also includes:

[0088] The power distribution target is parsed as a sub-target to determine a set of power distribution sub-targets; a new energy power distribution strategy space is constructed, and functions are fitted to the new energy power distribution strategy space based on the power distribution sub-target set to obtain a set of sub-target fitness functions; a decision-making evaluation is performed on the power distribution sub-target set based on the new energy power stage prediction results to obtain a set of sub-target weight factors; the sub-target fitness function set is weighted and fused based on the sub-target weight factor set to establish the new energy power time period allocation function.

[0089] Specifically, the overall power distribution goal is analyzed in detail and broken down into multiple specific sub-goals. Examples of sub-goals include: maximizing renewable energy utilization: increasing the proportion of renewable energy sources such as wind and solar power. minimizing power generation costs: reducing overall power generation costs and improving economic efficiency. ensuring power system stability: ensuring grid frequency and voltage are stable to meet load demand. reducing carbon emissions: optimizing the power generation structure to reduce emissions of harmful substances such as carbon dioxide. The analyzed sub-goals are aggregated to form a set of power distribution sub-goals.

[0090] Define a renewable energy power allocation strategy space, encompassing all possible power allocation scenarios. This strategy space considers factors such as power supply and demand, generation capacity, and market prices over different time periods. Based on a set of power allocation sub-goals, perform function fitting on the renewable energy power allocation strategy space. The fitting process uses historical and forecast data to establish sub-goal fitness functions. The fitness function represents the effectiveness of each power allocation scenario in achieving each sub-goal.

[0091] Use the output results of the power trading forecasting model to evaluate the performance of each sub-goal in different time periods. The evaluation process includes analyzing the changing trends of power demand, power generation and market prices. Based on the decision-making evaluation results, determine the importance and priority of each sub-goal. Assign weight factors to form a set of sub-goal weight factors. The decision-making evaluation refers to the systematic evaluation of the realization of each power distribution sub-goal in different time periods during the formulation of the new energy power distribution strategy, and determine the importance and priority of each sub-goal based on the evaluation results. The purpose of this evaluation process is to provide a scientific basis for the final power distribution decision, to ensure that the optimized power distribution plan can balance the requirements of multiple sub-goals and achieve the best overall effect.

[0092] Based on a set of sub-goal weight factors, the fitness functions of each sub-goal are weighted and fused. This fused fitness function forms the final renewable energy power time allocation function. This function comprehensively considers the achievement of each sub-goal and provides an optimized solution for power allocation in different time periods.

[0093] Through the above steps, a new energy power time allocation function that comprehensively considers multiple sub-goals can be established to optimize the power allocation strategy in different time periods, improve the utilization rate of new energy, reduce power generation costs, ensure system stability, and reduce carbon emissions.

[0094] Furthermore, the method described in this application also includes:

[0095] Parameter optimization is performed within the new energy power distribution scheme cluster by using a preset parameter step size to obtain multiple power distribution schemes; the multiple power distribution schemes are evaluated and calculated based on the new energy power time period allocation function to obtain the fitness of multiple power distribution schemes; based on the fitness of the multiple power distribution schemes, the new energy power distribution scheme cluster is compared and optimized to determine the optimal power distribution scheme.

[0096] Specifically, a preset parameter step size—the increment by which parameters are adjusted in each iteration—is set to optimize within a cluster of renewable energy power distribution schemes. Parameter optimization is performed within the cluster of renewable energy power distribution schemes based on the preset parameter step size. By systematically adjusting power distribution parameters, multiple different power distribution schemes are generated.

[0097] Using the established renewable energy power time allocation function, multiple generated power allocation schemes were evaluated. The evaluation process involved calculating each scheme's performance in achieving each sub-goal, generating a fitness value. The fitness function was used to calculate the overall effectiveness of each scheme in achieving each sub-goal. The fitness value was used to compare the strengths and weaknesses of different schemes.

[0098] The fitness of all generated power allocation plans is compared to identify the one with the highest fitness. This plan is the one that performs best across all sub-goals. By comparing fitness, the optimal power allocation plan is determined. This plan strikes the best balance between maximizing renewable energy utilization, minimizing power generation costs, ensuring system stability, and reducing carbon emissions.

[0099] This step utilizes parameter optimization, evaluation calculations, and fitness comparisons to identify the optimal power distribution solution within a cluster of renewable energy power distribution solutions. This solution achieves an optimal balance among multiple sub-objectives, providing reliable decision support for the optimized operation of renewable energy power systems.

[0100] Furthermore, the method described in this application also includes:

[0101] Based on the adaptability of the multiple power distribution schemes, the multiple power distribution schemes are compared and analyzed to determine the parameter optimization direction; according to the parameter optimization direction, they are matched and divided with the new energy power distribution scheme cluster to obtain a local power distribution scheme set; a local search step is set, and based on the local search step, the local power distribution scheme set is optimized and analyzed to determine the optimal power distribution scheme.

[0102] Specifically, the fitness of all generated power distribution schemes is calculated and compared. By comparing these fitness values, each scheme is evaluated for its effectiveness in achieving each sub-goal. Based on the fitness comparison results, further optimization directions are determined. The parameter optimization direction indicates how to adjust the power distribution parameters to improve the overall fitness of the scheme.

[0103] Based on the determined parameter optimization direction, the new energy power distribution plan clusters are matched and divided. A local power distribution plan set is generated, focusing on the most promising parameter areas for further optimization.

[0104] Set the local search step size, which is the increment by which parameters are adjusted in each iteration when optimizing within the set of local power allocation solutions. This local search step size is typically smaller than the initial parameter step size to allow for fine-grained optimization. Optimization analysis is performed within the set of local power allocation solutions using the set local search step size. Power allocation parameters are iteratively adjusted to gradually approach the optimal solution. Through fine-grained optimization within the set of local solutions, the power allocation solution with the highest fitness is identified. This solution, which performs best in achieving multiple sub-goals, becomes the final optimal power allocation solution.

[0105] This step, based on comparative analysis, matching partitioning, and local search optimization, allows for detailed optimization within a cluster of renewable energy power distribution solutions to identify the optimal power distribution solution. This solution achieves an optimal balance among multiple sub-goals, providing reliable decision support for the optimized operation of renewable energy power systems.

[0106] In summary, the deep learning-based new energy power distribution strategy decision-making method provided in the embodiments of the present application has the following technical effects:

[0107] 1. Using big data technology to collect and standardize preprocessed renewable energy power datasets, and employing deep learning networks for training and optimization, a renewable energy power trading forecasting model was constructed. Based on the forecast results, power distribution targets were optimized and a power time allocation function was established. Ultimately, the optimal power distribution plan was obtained through evaluation and iterative optimization. This approach significantly improved the accuracy and optimization efficiency of renewable energy power distribution.

[0108] 2. Detailed labeling of new energy standard power data sets was performed to generate power supply and demand sample sets, power generation sample sets, and power transaction price sample sets. Using deep learning networks, recurrent supervised training was performed to generate multiple prediction models. Through model fusion and validation optimization, a comprehensive new energy power transaction prediction model was constructed, improving the accuracy of forecasts for future power supply and demand, power generation, and transaction prices.

[0109] 3. The initial power trading forecast model was evaluated and verified using the model loss function, gradient information was calculated, and model parameters were iteratively updated using the backpropagation algorithm. Ultimately, the optimized configuration resulted in a highly accurate new energy power trading forecast model. This process improved the model's prediction accuracy and generalization capabilities.

[0110] 4. Generate multiple power distribution plans through parameter optimization. These plans are evaluated and calculated based on the renewable energy power time allocation function to determine the optimal power distribution plan with the highest adaptability. This step improves the scientific and rationality of the power distribution plan and ensures the optimal plan's comprehensive performance across multiple sub-goals.

[0111] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0112] Furthermore, the terms "first" or "second" as described above may not only represent an order relationship but may also represent a specific concept and / or refer to the selection of multiple elements individually or collectively. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, if such modifications and variations fall within the scope of this application and its equivalents, this application is intended to include such modifications and variations.

Claims

1. A new energy power distribution strategy decision-making method based on deep learning, characterized in that: The method comprises: Acquire new energy power data sets through big data technology, including renewable energy time-series power generation information, basic power market information, power generation-side transaction information, mid- and long-term and spot transaction time-series data, and grid-side power shortage data; Obtaining a data preprocessing program, and performing standardized preprocessing on the new energy power data set by the data preprocessing program to obtain a new energy standard power data set; Using a deep learning network to train and optimize the new energy standard power data set to build a new energy power transaction prediction model; Performing power forecasting for a preset time window based on the new energy power transaction forecasting model to determine a new energy power stage forecasting result; Obtaining a power allocation target, optimizing weight allocation for the power allocation target based on the new energy power stage prediction result, and establishing a new energy power time period allocation function; Obtaining a new energy power allocation plan cluster based on the new energy power stage prediction result, evaluating and iteratively optimizing the new energy power allocation plan cluster based on the new energy power time period allocation function, and obtaining an optimal power allocation plan as a target new energy power time period allocation strategy; The step of establishing a new energy power time allocation function includes: Performing sub-goal analysis on the power distribution goal to determine a power distribution sub-goal set; Constructing a new energy power allocation strategy space, performing function fitting on each of the new energy power allocation strategy spaces based on the power allocation sub-goal set, and obtaining a sub-goal fitness function set; Performing a decision-making evaluation on the power allocation sub-goal set based on the new energy power stage prediction result to obtain a sub-goal weight factor set; The sub-objective fitness function set is weighted and fused based on the sub-objective weight factor set to establish the new energy power time period allocation function.

2. A new energy power distribution strategy decision-making method based on deep learning according to claim 1, characterized in that: The obtaining of the new energy standard power data set includes: Performing abnormal data identification on the new energy power data set to obtain an abnormal power data set, wherein the abnormal power data set includes inconsistent data, missing data, and out-of-bounds data; Mapping and matching the power abnormality data set with the data preprocessing program to determine an abnormal data preprocessing matching program; Preprocessing the inconsistent data, missing data and out-of-bounds data based on the abnormal data preprocessing and matching program to obtain consistent power data, complete power data and standardized power data; The new energy standard power data set is obtained according to the consistent power data, the complete power data and the normative power data.

3. The method for decision-making on new energy power distribution strategy based on deep learning according to claim 1, characterized in that: The construction of the new energy power transaction prediction model includes: Marking the power supply and demand situation, power generation, and transaction price of the new energy standard power data set to obtain a power supply and demand situation sample set, a power generation sample set, and a power transaction price sample set; Based on the deep learning network, the power supply and demand sample set, the power generation sample set, and the power transaction price sample set are subjected to cyclic supervised training to generate a power supply and demand prediction model, a power generation prediction model, and a power transaction price prediction model; Merging the power supply and demand prediction model, the power generation prediction model, and the power transaction price prediction model to generate an initial power transaction prediction model; The initial power transaction prediction model is verified and optimized to construct the new energy power transaction prediction model.

4. The method for decision-making on new energy power distribution strategy based on deep learning according to claim 3, characterized in that: The generating of the power supply and demand prediction model, the power generation prediction model and the power transaction price prediction model includes: Selecting a recurrent neural network structure according to a deep learning network, and obtaining an initial hidden layer value and a model weight matrix of the recurrent neural network structure; Based on the initial hidden layer value and the model weight matrix, performing hidden layer calculations on the power supply and demand sample set, the power generation sample set, and the power transaction price sample set to obtain current hidden layer values; Based on the current hidden layer value and the model weight matrix, the electricity supply and demand sample set, the electricity generation sample set and the electricity trading price sample set are respectively subjected to iterative calculation training to generate the electricity supply and demand prediction model, the electricity generation prediction model and the electricity trading price prediction model.

5. The method for decision-making on new energy power distribution strategy based on deep learning according to claim 3, characterized in that: The constructing of the new energy power transaction prediction model includes: Using a model loss function to evaluate and verify the initial power transaction prediction model, and determine the model performance loss parameter; Performing gradient calculation on the cyclic model parameters of the initial power transaction prediction model based on the model performance loss parameter to obtain prediction model parameter gradient information; Iteratively updating the cyclic model parameters based on the prediction model parameter gradient information by a back propagation algorithm to obtain cyclic model update parameters; The initial power transaction prediction model is optimized and configured based on the cycle model update parameters to obtain the new energy power transaction prediction model.

6. The method for decision-making on new energy power distribution strategy based on deep learning according to claim 1, characterized in that: The obtaining of the optimal power distribution plan includes: Performing parameter optimization within the new energy power distribution scheme cluster by using a preset parameter step size to obtain multiple power distribution schemes; Evaluating and calculating the plurality of power distribution schemes based on the new energy power time period allocation function to obtain the adaptability of the plurality of power distribution schemes; The new energy power distribution scheme cluster is compared and optimized based on the adaptability of the multiple power distribution schemes to determine the optimal power distribution scheme.

7. The method for decision-making on new energy power distribution strategy based on deep learning according to claim 6, characterized in that: Determining the optimal power distribution plan includes: Comparing and analyzing the multiple power distribution schemes based on their adaptability, and determining a parameter optimization direction; Matching and dividing the new energy power distribution plan cluster according to the parameter optimization direction to obtain a local power distribution plan set; A local search step size is set, and optimization analysis is performed on the local power allocation solution set based on the local search step size to determine the optimal power allocation solution.

Images