Renewable energy intelligent scheduling method and system

Abstract

The invention discloses an intelligent scheduling method and system for renewable energy sources, and the method comprises the steps: collecting power generation data, energy storage data and power utilization data of a power station, and enabling the power utilization data to comprise power grid load power utilization data and charging facility power utilization data; respectively establishing a power generation prediction model and a power consumption prediction model based on the historical power generation data and the historical power consumption data, and performing training; on the basis of the power generation prediction model and the power consumption prediction model, respectively predicting the power generation capacity and the power consumption demand trend of the power grid load and the charging facility in the future set time; and formulating an energy scheduling strategy by adopting an optimization algorithm based on the predicted generating capacity, the power consumption demand, the current real-time energy storage data, the power generation data and the power consumption data. According to the invention, accurate control and optimal scheduling of photovoltaic energy, an energy storage system, a power grid and a load can be realized. The method has remarkable advantages in the aspects of improving energy utilization efficiency, reducing operation cost, enhancing power grid stability and the like.

Description

Technical Field

[0001] The present invention belongs to the field of smart grids, and particularly relates to an intelligent scheduling method and system for renewable energy. Background Art

[0002] With the global emphasis on renewable energy and the rapid development of photovoltaic technology, photovoltaic energy has become one of the important power sources. In the current construction and development of industrial parks, photovoltaic energy supply will also be increased to further promote the realization of carbon-neutral industrial parks. However, the output of photovoltaic power generation is affected by factors such as weather and time, and has intermittency and uncertainty, which poses challenges to the stable operation of the power grid. At the same time, the fluctuations of the power grid load, the charge and discharge strategies of energy storage systems, and the demand response of charging facilities such as electric vehicles also increase the complexity of energy scheduling. Therefore, it is particularly important to develop an intelligent scheduling method and system that can comprehensively consider the above factors.

[0003] The existing source-network-load-storage-charge of photovoltaic energy mainly adopts the following solutions:

[0004] (1) Manually formulate the power generation plan of photovoltaic energy;

[0005] (2) Adjust the output power of photovoltaic energy according to the power grid load demand;

[0006] (3) Adopt a simple energy storage charge and discharge strategy, such as charge and discharge according to the electricity price fluctuation;

[0007] (4) Monitor the operation status of the power grid through manual or simple automation systems.

[0008] The above technical solutions lack sufficient intelligence and automation levels. In the face of large-scale access of photovoltaic energy and complex power grid environments, problems such as inflexible scheduling and slow response speed may occur. Summary of the Invention

[0009] In view of the above problems, the technical solution adopted by the present invention is: an intelligent scheduling method for renewable energy, and the scheduling method includes the following steps:

[0010] Collect power generation data, energy storage data, and power consumption data of power generation stations, where the power consumption data includes power grid load power consumption data and charging facility power consumption data;

[0011] Based on historical power generation data and historical power consumption data, establish a power generation prediction model and a power consumption prediction model respectively and train them;

[0012] Based on the power generation prediction model and the power consumption prediction model, predict the power generation amount and the power grid load and charging facility power consumption demand trends within a future set time respectively;

[0013] Based on the predicted power generation, electricity demand, and current real-time energy storage data, power generation data, and electricity consumption data, an optimization algorithm is used to formulate an energy scheduling strategy.

[0014] Optionally, in the step of establishing and training the power generation prediction model, it includes:

[0015] Obtain historical photovoltaic power generation data, including power generation and influencing factors, and perform preprocessing;

[0016] Select relevant parameters as feature values and perform power generation power correlation analysis;

[0017] Divide the data into a training set and a test set;

[0018] Construct a power generation prediction model, and configure an LSTM layer for capturing long-term dependencies in time series data and a fully connected layer for outputting prediction results;

[0019] Add multiple LSTM layers or convolutional layer CNNs, and based on the Adam optimizer and mean squared error MSE loss function, set the compilation parameters;

[0020] Use the training set data to train the power generation prediction model. During the training process, continuously adjust the parameters according to the loss function to minimize the prediction error, and at the same time use the validation set data to monitor the training process of the model;

[0021] Use the test set data to evaluate the trained model and calculate the prediction error;

[0022] According to the evaluation results, adopt corresponding strategies to optimize the power generation prediction model;

[0023] Repeat the processes of training, evaluation, and optimization until the model reaches the set prediction accuracy.

[0024] Optionally, the electricity consumption prediction model includes a power grid load electricity consumption prediction model and a charging facility electricity consumption prediction model; in the step of establishing and training the power grid load electricity consumption prediction model or the charging facility electricity consumption prediction model, it includes:

[0025] Obtain historical electricity consumption data and perform preprocessing, including electricity consumption and corresponding influencing factors of electricity consumption types;

[0026] Extract features and combine them with load data to form an input feature rectangle;

[0027] Divide the data into a training set, a validation set, and a test set;

[0028] Design a CNN model according to the dimensions of the input feature matrix and output requirements, including convolutional layers, pooling layers, and fully connected layers. Add activation functions to enhance the non-linear expression ability of the model, and select the corresponding loss function according to the task type;

[0029] Feed the input feature matrix into the CNN model to output predicted values;

[0030] Calculate the loss based on the output predicted values and the true values, calculate the gradient according to the loss function, and update the model parameters through the backpropagation algorithm. Repeat the steps until the preset number of iterations is reached or the loss converges;

[0031] During the training process, use the validation set data to evaluate the model performance, and adjust the model parameters or structure according to the evaluation results.

[0032] Optionally, in the step of formulating the energy scheduling strategy using the optimization algorithm, it includes:

[0033] Regard the charging and discharging strategy of the energy storage device as a multi-dimensional vector, and each dimension represents the state of a charging and discharging period;

[0034] Adopt the particle swarm optimization algorithm to randomly generate a group of particles. Each particle represents a charging and discharging strategy, and each particle has its own position and velocity information;

[0035] Initialize the individual best position and the global best position;

[0036] Calculate the fitness value of each particle according to the objective function. If the current fitness of the particle is better than the individual best fitness, update the individual best position. If it is better than the global best fitness, update the global best position. Repeat the calculation and update steps until the maximum number of iterations is reached or the stop condition is satisfied; where the objective function is the weighted sum between the peak-valley load demand of the power grid and the operating cost of the energy storage system;

[0037] Return the global best position and the fitness as the solution to the optimization problem.

[0038] Optionally, the scheduling method further includes the following steps:

[0039] Determine the power supply strategy of the charging facility according to the charging and discharging strategy of the energy storage device, the predicted power generation, and the power consumption of the power grid load in the energy scheduling strategy;

[0040] Judge whether the determined charging facility power supply strategy meets the power demand trend of the charging facility, and calculate the supply-demand ratio;

[0041] Formulate a power price adjustment strategy based on the judgment result and the supply-demand ratio value.

[0042] Optionally, the scheduling method further includes the following steps:

[0043] Analyze and evaluate the energy utilization rate based on the power generation, grid load, charge and discharge of energy storage devices, and electricity consumption data of charging facilities; and analyze and evaluate the grid stability based on the real-time grid load fluctuation and photovoltaic power generation change data.

[0044] Optimize the energy scheduling strategy according to the data analysis results.

[0045] Optionally, in the step of collecting energy storage data, it includes:

[0046] Set the acquisition nodes of the energy storage device and the corresponding type of data collector, where the data collector includes a temperature sensor, an electricity sensor, and a health status monitoring sensor;

[0047] Configure the parameters of the energy storage data collector, including the acquisition period and data storage format;

[0048] Send the data collected by the data collector to the data processing center based on the communication network;

[0049] Extract the temperature data, conduct temperature trend analysis and temperature anomaly detection and write them into the database; extract the health status data and write them into the database for health assessment and fault prediction.

[0050] And, an intelligent scheduling system for renewable energy, the scheduling system includes:

[0051] Data acquisition module: including various data collectors, used to collect power generation data, energy storage data, and electricity consumption data of power stations, and the electricity consumption data includes grid load electricity consumption data and charging facility electricity consumption data;

[0052] Model training module: used to establish and train a power generation prediction model and an electricity consumption prediction model respectively based on historical power generation data and historical electricity consumption data;

[0053] Prediction module: used to predict the power generation volume and the trends of grid load and charging facility electricity consumption demand within a set future time respectively based on the power generation prediction model and the electricity consumption prediction model;

[0054] Scheduling decision module: used to formulate an energy scheduling strategy by using an optimization algorithm based on the predicted power generation volume, electricity consumption demand, and current real-time energy storage data, power generation data, and electricity consumption data;

[0055] Energy scheduling module, used to control the charge and discharge of the energy storage system according to the strategy formulated by the scheduling decision module.

[0056] Optionally, when formulating the energy scheduling strategy, the scheduling decision module specifically includes the following steps:

[0057] Regard the charge and discharge strategy of the energy storage device as a multi-dimensional vector, where each dimension represents the state of a charge and discharge period;

[0058] Adopt the particle swarm optimization algorithm to randomly generate a group of particles. Each particle represents a charge and discharge strategy, and each particle has its own position and velocity information;

[0059] Initialize the individual best position and the global best position;

[0060] Calculate the fitness value of each particle according to the objective function. If the current fitness of the particle is better than the individual best fitness, update the individual best position. If it is better than the global best fitness, update the global best position. Repeat the calculation and update steps until the maximum number of iterations is reached or the stop condition is satisfied. Among them, the objective function is the weighted sum between the peak-valley load demand of the power grid and the operating cost of the energy storage system;

[0061] Return the global best position and the fitness as the solution to the optimization problem.

[0062] Optionally, the scheduling system further includes a electricity price regulation module for regulating the electricity price of the charging facilities according to the energy scheduling strategy, specifically including:

[0063] Determine the power supply strategy of the charging facilities according to the charge and discharge strategy of the energy storage device, the predicted power generation, and the power consumption of the power grid load in the energy scheduling strategy;

[0064] Judge whether the determined power supply strategy of the charging facilities meets the power consumption demand trend of the charging facilities, and calculate the supply-demand ratio;

[0065] Formulate an electricity price regulation strategy based on the judgment result and the supply-demand ratio value.

[0066] Due to the adoption of the above technical solutions, the present invention has the following beneficial effects: It can achieve precise control and optimal scheduling of photovoltaic energy, energy storage systems, power grids, and loads. It has significant advantages in improving energy utilization efficiency, reducing operating costs, and enhancing power grid stability.

[0067] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures pointed out in the specification and claims. Detailed implementation manners

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0069] Embodiment 1

[0070] The intelligent scheduling method for renewable energy in the embodiments of the present invention includes the following steps:

[0071] S1: Collect power generation data, energy storage data, and power consumption data of the power station. The power consumption data includes power grid load power consumption data and charging facility power consumption data.

[0072] In the park of this embodiment, a photovoltaic power station is set up. In the step of collecting power generation data of the power station, it includes:

[0073] S1101: Configure the parameters of the data collector according to the actual situation of the photovoltaic power station in the park, such as the collection frequency, data type, etc., to ensure that the power generation data of the photovoltaic panels can be accurately collected;

[0074] S1102: Build an RS485, RS232, or JS45 network port communication network to ensure that the data collector can transmit data to the data processing center in real time. The data processing center is used to receive, store, and analyze the data from the data collector;

[0075] S1103: The data processing center receives the data from the data collector and performs verification to ensure the accuracy and integrity of the data; parse and convert the received data and convert it into a format that can be used for analysis and display;

[0076] S1104: Store the processed data in the database of the data processing center for subsequent analysis and model training.

[0077] In the step of collecting energy storage data, it includes:

[0078] S1201: Set the collection nodes of the energy storage device and the corresponding type of data collector. The data collector includes a temperature sensor, a power sensor, and a health status monitoring sensor. The power sensor is connected to the electronic control system of the energy storage device. The temperature sensor is placed at key positions inside or outside the energy storage device. The health status monitoring sensor is connected to the control system or key components of the energy storage device;

[0079] S1202: Configure the parameters of the energy storage data collector, including the collection period and data storage format;

[0080] S1203: Set up a communication network and send the data collected by the data collector to the data processing center based on the communication network;

[0081] S1204: The data processing center receives the data from the data collector and conducts a preliminary verification to ensure the integrity, accuracy, and timeliness of the data;

[0082] S1205: Extract the temperature data, conduct temperature trend analysis and temperature anomaly detection, and write them into the database; extract the health status data and write them into the database for health assessment and fault prediction.

[0083] In the step of collecting the power consumption data of the power grid load, it includes:

[0084] S1301: Set the types, collection frequencies, and progress of the power grid load data to be collected;

[0085] S1302: Install intelligent meter collection devices at the key nodes of the power grid to ensure that the power grid load data can be comprehensively and accurately collected;

[0086] S1303: Set up a communication network to ensure that the data collector can transmit the data to the data processing center in real time;

[0087] S1304: After receiving the data, the data processing center conducts verification to ensure the accuracy and integrity of the data, stores the verified data in the database, and provides necessary data support for subsequent model analysis.

[0088] In the step of collecting the power consumption data of the charging facilities, it includes:

[0089] S1401: The data collected by the charging facilities includes charging requests, charging rates, charging times, and charging amounts, etc.; configure sensors on the charging piles to monitor the operating status and various parameters of the charging piles, such as current, voltage, temperature, etc., so as to indirectly obtain the charging rate, charging time, and charging amount data;

[0090] S1402: Develop a unified data interface standard to ensure that charging facilities of different brands can be connected to the same data platform;

[0091] S1403: Configure a data collector through the RS485 / RS232 / JS45 network port to collect the data information of the charging piles in real time and transmit the collected data to the data processing center;

[0092] S1404: The data processing center conducts preprocessing operations such as cleaning and format conversion on the collected original data to improve the data quality and usability;

[0093] S1405: Map the preprocessed data into the corresponding data model and perform parsing for subsequent analysis and application;

[0094] S1406: Set up a real-time monitoring mechanism to monitor the operating status of the charging facilities in real time, and give an alarm in a timely manner when abnormalities occur. And write the relevant data into the above database.

[0095] S2: Based on historical power generation data and historical power consumption data, establish and train a power generation prediction model and a power consumption prediction model respectively;

[0096] Among them, in the step of establishing and training the power generation prediction model, it includes:

[0097] S2101: Obtain historical photovoltaic power generation data, including power generation amount, date and time, power generation equipment identification, etc., and also include influencing factors, such as weather data during the same period, irradiance, temperature, wind speed, cloud cover, etc.; and perform preprocessing, including format conversion, necessary resampling and interpolation, and handling missing values and outliers;

[0098] S2102: Select relevant parameters as feature values and perform power generation power correlation analysis;

[0099] S2103: Divide the data into a training set and a test set, and normalize the data set to improve the convergence speed and prediction accuracy of the model; yuce

[0100] S2104: Initialize a Sequential Model, add an LSTM layer, set parameters such as the number of LSTM units and activation functions. The LSTM layer is used to capture the long-term dependencies of time series data, and add a DenseLayer for outputting prediction results;

[0101] S2105: Add multiple LSTM layers or convolutional layer CNNs to build a more complex model, and based on the Adam optimizer and the mean square error MSE loss function, set other compilation parameters such as the learning rate;

[0102] Among them, the LSTM unit controls the flow and update of information through the forget gate, input gate and output gate. The following are the key formulas for LSTM unit state update:

[0103] Forget gate:

[0104] f t =σ(W f ·[h t-1 , x t +b f )

[0105] Input gate:

[0106] i t = σ(W i · [h t-1 , x t + b i )

[0107] Candidate memory unit:

[0108]

[0109] Memory unit state update:

[0110]

[0111] Output gate:

[0112] o t = σ(W o · [h t-1 , x t + b o )

[0113] Hidden layer state update:

[0114] h t = o t * tanh(C t )

[0115] where σ represents the sigmoid function, tanh represents the hyperbolic tangent function, * represents element-wise multiplication, W f , W i , W c , W o represent the weight matrices of the forget gate, input gate, candidate memory unit, and output gate respectively, and b f , b i , b c , b o represent the corresponding bias terms. h (t-1) represents the hidden layer state at the previous time step, and x t represents the input at the current time step.

[0116] When training an LSTM model, the mean squared error (MSE) is usually used as the loss function to measure the difference between the predicted value and the true value. The formula for MSE is as follows:

[0117]

[0118] where n is the number of samples, yi is the true value of the i-th sample, is the predicted value of the i-th sample, ∑ represents the summation operation, that is, summing the squared prediction errors of all samples, and (1 / n) is a normalization factor used to make the unit of MSE consistent with the units of the true value and the predicted value, facilitating the comparison of the performance of different models.

[0119] S2106: Use the training set data to train the model. During the training process, continuously adjust the parameters according to the loss function to minimize the prediction error, and at the same time use the validation set data to monitor the training process of the model to prevent overfitting;

[0120] S2107: Use the test set data to evaluate the trained model and calculate the prediction error, such as the absolute error, relative error, etc.;

[0121] S2108: According to the evaluation results, adopt corresponding strategies to optimize the power generation prediction model, such as adjusting the number of LSTM units, adding regularization terms, using dropout and other strategies;

[0122] S2109: Repeat the process of training, evaluation and optimization until the model reaches the set prediction accuracy.

[0123] The electricity consumption prediction model includes a power grid load electricity consumption prediction model and a charging facility electricity consumption prediction model; in the steps of establishing and training the power grid load electricity consumption prediction model or the charging facility electricity consumption prediction model, it includes:

[0124] S2201: Obtain historical power grid load electricity consumption data and perform preprocessing, including electricity consumption and corresponding influencing factors of electricity consumption types, including marked data such as time and date, holiday information, and economic activities; the preprocessing includes handling missing values, outliers, etc. to ensure data quality;

[0125] S2202: Extract features according to factors such as holidays and economic activities and combine them with the load data to form an input feature matrix;

[0126] S2203: Normalize or standardize the input features to improve the model training efficiency; for data types such as graphics, increase data diversity through rotation, scaling, etc.; for power grid load data, enhance it through time window slicing, sequence recombination, etc. Then divide the data into a training set, a validation set and a test set;

[0127] S2204: Design the network structure of the CNN model according to the dimensions of the input feature matrix and the output requirements, including convolutional layers, pooling layers, fully connected layers, etc., add activation functions for enhancing the nonlinear expression ability of the model, such as ReLU, Sigmoid and other activation functions, and select appropriate loss functions according to the task type (such as regression or classification), such as mean square error (MSE) or cross-entropy loss (Cross-Entropy Loss);

[0128] S2205: Feed the input feature matrix into the CNN model, and obtain the output prediction value through operations such as convolution, pooling, and fully connected layers;

[0129] Among them, the relevant algorithm formulas of CNN are as follows:

[0130] Formula for the output size of the convolutional layer:

[0131]

[0132] Formula for the output size of the pooling layer:

[0133]

[0134] In the formula, N: the size of the output feature map, W: the size of the input feature map, F: the size of the convolutional kernel, P: the padding size, S: the stride size, denotes rounding down.

[0135] For regression tasks, the mean squared error (MSE) is commonly used as the loss function:

[0136]

[0137] In the formula, y i denotes the true value, denotes the predicted value, and n denotes the number of samples.

[0138] For classification tasks, the cross-entropy loss is commonly used as the loss function:

[0139]

[0140] In the formula: y i denotes the true label of the i-th sample denotes the predicted probability.

[0141] S2206: Calculate the loss based on the output prediction value and the true value, calculate the gradient according to the loss function, and update the model parameters through the backpropagation algorithm. Repeat the steps until the preset number of iterations or the loss converges;

[0142] S2207: Evaluate the model performance using the validation set data during training, such as accuracy, recall, etc., and adjust the model parameters or structure according to the evaluation results.

[0143] Similarly, the electricity consumption prediction model for charging facilities is obtained according to the steps of S2201 to S2207 above. The difference lies in the different influencing factors. Weather data, etc. can also be added to the electricity consumption prediction of charging facilities.

[0144] S3: Based on the power generation prediction model and the power consumption prediction model, respectively predict the power generation within a set future time, as well as the trends of grid load and charging facility power consumption demand;

[0145] S4: Based on the predicted power generation, power consumption demand, as well as the current real-time energy storage data, power generation data, and power consumption data, use an optimization algorithm to formulate an energy scheduling strategy.

[0146] In the step of formulating the energy scheduling strategy using the optimization algorithm, it includes:

[0147] S4101: Regard the charge and discharge strategy of the energy storage device as a multi-dimensional vector, and each dimension represents the state of a charge and discharge period;

[0148] S4102: Use the particle swarm optimization algorithm to randomly generate a group of particles, each particle represents a charge and discharge strategy, and each particle has its own position and velocity information;

[0149] S4103: Initialize the individual best position and the global best position;

[0150] S4104: Calculate the fitness value of each particle according to the objective function. If the current fitness of the particle is better than the individual best fitness, update the individual best position. If the current fitness of the particle is better than the global best fitness, update the global best position; repeat the calculation and update steps until the maximum number of iterations is reached or the stop condition is satisfied: Among them, the objective function is the weighted sum between the peak-valley load demand of the power grid and the operating cost of the energy storage system;

[0151] S4105: Return the global best position and fitness as the solution to the optimization problem.

[0152] S5: Send the formulated scheduling strategy to the power station, power grid scheduling system, energy storage device, and charging station to achieve the matching of photovoltaic power generation and grid load, optimize the charge and discharge strategy of the energy storage device, and the demand response of the charging facility.

[0153] S6: Analyze and evaluate the energy utilization rate based on the power generation, grid load, charge and discharge of the energy storage device, and the power consumption data of the charging facility, and analyze and evaluate the grid stability based on the real-time grid load fluctuation and photovoltaic power generation change data; according to the data analysis results, optimize the energy scheduling strategy.

[0154] In this embodiment, by differentiating the power consumption of the grid load and the charging facility, the power consumption is predicted separately to improve the accuracy of the power consumption demand prediction. At the same time, the formulation of the energy scheduling strategy is carried out separately, with the peak-valley load demand of the power grid as one of the objective functions, to ensure that the power consumption demand of the grid load is met and then the power supply demand of the charging equipment is met to ensure the normal operation of the conventional electrical facilities in the park.

[0155] Embodiment 2

[0156] Compared with the above Embodiment 1, this embodiment further includes the following steps:

[0157] S01: Determine the power supply strategy of the charging facilities according to the charge and discharge strategy of the energy storage device, the predicted power generation, and the power consumption of the power grid load in the energy dispatching strategy;

[0158] S02: Judge whether the determined power supply strategy of the charging facilities meets the power consumption demand trend of the charging facilities, and calculate the supply-demand ratio;

[0159] S03: Formulate a power price adjustment strategy based on the judgment result and the supply-demand ratio value.

[0160] Specifically, the basic power price M and adjustment parameters a, b, c, d can be set first, and a < b < 1 < c < d. If the charging facility demand is met and the supply-demand ratio is very high (for example, greater than 2), then the adjustment parameter a is selected to lower the power price to M×a to promote the charging demand; if the charging facility demand is met and the supply-demand ratio is relatively high (for example, between 1.5 and 2), then the adjustment parameter b is selected to lower the power price to M×b to promote the charging demand; if the charging facility demand is not met and the supply-demand ratio is low, then the adjustment parameter c is selected to increase the power price to M×c to reduce the charging demand; if the charging facility demand is not met and the supply-demand ratio is very low, then the adjustment parameter d is selected to increase the power price to M×d to further reduce the charging demand. In addition, the park management software can be combined to send charging information to the personnel in the park to achieve the purpose of promoting or reducing the charging demand.

[0161] Embodiment 3

[0162] Based on the above Embodiments 1 and 2, the embodiment of the present invention provides an intelligent dispatching system for renewable energy, including:

[0163] Data acquisition module: including various data collectors, used to collect power generation data, energy storage data, and power consumption data of the power station, and the power consumption data includes power grid load power consumption data and charging facility power consumption data;

[0164] Model training module: used to establish and train a power generation prediction model and a power consumption prediction model respectively based on historical power generation data and historical power consumption data;

[0165] Prediction module: used to predict the power generation amount and the power consumption demand trends of the power grid load and charging facilities within a set future time respectively based on the power generation prediction model and the power consumption prediction model;

[0166] Dispatch decision module: used to formulate an energy dispatching strategy by using an optimization algorithm based on the predicted power generation amount, power consumption demand, and current real-time energy storage data, power generation data, and power consumption data;

[0167] An energy scheduling module for regulating the charging and discharging of an energy storage system according to the strategy formulated by the scheduling decision-making module;

[0168] A electricity price adjustment module for adjusting the electricity price of charging facilities according to the energy scheduling strategy;

[0169] A scheduling decision optimization module for optimizing the energy scheduling strategy.

[0170] It should be noted that the implementation processes of the functions and roles of each device, module, unit, etc. in this system are specifically described in the corresponding steps of the above method, and will not be elaborated here.

[0171] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments by using the disclosed technical content within the scope of the technical solution of the present invention. However, as long as it does not depart from the content of the technical solution of the present invention, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A method for intelligent dispatching of renewable energy, characterized in that: The scheduling method comprises the following steps: Collecting power generation data, energy storage data and power consumption data of power stations, wherein the power consumption data includes power grid load power consumption data and charging facility power consumption data; Based on historical power generation data and historical power consumption data, a power generation prediction model and a power consumption prediction model are established and trained respectively; Based on the power generation prediction model and the power consumption prediction model, respectively predict the power generation within a set time in the future and the power demand trend of the grid load and the charging facility; Based on the predicted power generation, power demand and current real-time energy storage data, power generation data and power consumption data, an optimization algorithm is used to formulate an energy scheduling strategy.

2. The intelligent dispatching method of renewable energy according to claim 1, characterized in that: The steps of establishing and training the power generation prediction model include: Obtain historical photovoltaic power generation data, including power generation and influencing factors, and perform preprocessing; Select relevant parameters as characteristic values ​​and conduct power generation correlation analysis; Divide the data into training and testing sets; Build a power generation forecasting model and configure the LSTM layer for capturing the long-term dependencies of time series data and the fully connected layer for outputting forecast results; Add multiple LSTM layers or convolutional CNN layers, and set compilation parameters based on the Adam optimizer and the mean square error MSE loss function; The power generation prediction model is trained using the training set data, and during the training process, parameters are continuously adjusted according to the loss function to minimize the prediction error, and the training process of the model is monitored using the validation set data; Use the test set data to evaluate the trained model and calculate the prediction error; According to the evaluation results, adopting corresponding strategies to optimize the power generation prediction model; The process of training, evaluation, and optimization is repeated until the model reaches the set prediction accuracy.

3. The intelligent dispatching method of renewable energy according to claim 1, characterized in that: The power consumption prediction model includes a power grid load power consumption prediction model and a charging facility power consumption prediction model; The steps of establishing and training a power grid load power consumption prediction model or a charging facility power consumption prediction model include: Obtain historical electricity consumption data and pre-process it, including electricity consumption and the corresponding factors affecting electricity consumption types; Extract features and combine them with load data to form input feature rectangles; Divide the data into training, validation and test sets; Design a CNN model based on the dimensions of the input feature matrix and output requirements, including convolutional layers, pooling layers, and fully connected layers, add activation functions to enhance the nonlinear expression capabilities of the model, and select the corresponding loss function based on the task type; Feed the input feature matrix into the CNN model to output the predicted value; Calculate the loss based on the output prediction value and the true value, calculate the gradient based on the loss function, and update the model parameters through the back propagation algorithm, repeating the steps until the preset number of iterations is reached or the loss converges; During the training process, validation set data is used to evaluate model performance and model parameters or structure are adjusted based on the evaluation results.

4. The intelligent dispatching method of renewable energy according to claim 1, characterized in that: The step of using the optimization algorithm to formulate an energy dispatch strategy includes: The charging and discharging strategy of the energy storage device is regarded as a multi-dimensional vector, where each dimension represents the state of a charging and discharging period; The particle swarm optimization algorithm is used to randomly generate a group of particles, each particle represents a charging and discharging strategy, and each particle has its own position and speed information; Initialize individual best position and global best position; The fitness value of each particle is calculated according to the objective function. If the current fitness of the particle is better than the individual best fitness, the individual best position is updated. If it is better than the global best fitness, the global best position is updated. Repeat the calculation and update steps until the maximum number of iterations is reached or the stop condition is met. The objective function is the weighted sum of the peak and valley load demand of the power grid and the operating cost of the energy storage system. Returns the global best position and fitness as a solution to the optimization problem.

5. The intelligent dispatching method of renewable energy according to claim 4, characterized in that: The scheduling method further comprises the following steps: Determine the power supply strategy for charging facilities based on the energy storage device charging and discharging strategy, predicted power generation, and grid load power consumption in the energy dispatch strategy; Determine whether the charging facility power supply strategy meets the power demand trend of the charging facility, and calculate the supply-demand ratio; Formulate electricity price adjustment strategies based on the judgment results and supply-demand ratio.

6. The intelligent dispatching method for renewable energy according to any one of claims 1 to 5, characterized in that: The scheduling method further comprises the following steps: Evaluate energy utilization based on analysis of power generation, grid load, energy storage device charging and discharging, and charging facility power consumption data; and analyze and evaluate grid stability based on real-time grid load fluctuations and photovoltaic power generation change data; According to the data analysis results, the energy scheduling strategy is optimized.

7. The intelligent dispatching method of renewable energy according to claim 1, characterized in that: The step of collecting energy storage data includes: Setting the collection nodes of the energy storage device and the corresponding types of data collectors, wherein the data collectors include temperature sensors, power sensors and health status monitoring sensors; Configure the parameters of the energy storage data collector, including the collection cycle and data storage format; Sending the data collected by the data collector to a data processing center based on a communication network; Extract temperature data, perform temperature trend analysis, temperature anomaly detection and write to the database; extract health status data and write to the database for health assessment and fault prediction.

8. An intelligent dispatching system for renewable energy, characterized in that: The dispatching system comprises: Data acquisition module: including various data collectors, used to collect power station data, energy storage data and power consumption data, including power grid load power consumption data and charging facility power consumption data; Model training module: used to establish and train power generation prediction models and power consumption prediction models based on historical power generation data and historical power consumption data respectively; Prediction module: used to predict the power generation and power demand trends of the grid load and charging facilities within a set time in the future based on the power generation prediction model and the power consumption prediction model; Scheduling decision module: used to formulate energy scheduling strategies using optimization algorithms based on the predicted power generation, power demand, and current real-time energy storage data, power generation data, and power consumption data; The energy scheduling module is used to regulate the charging and discharging of the energy storage system according to the strategy formulated by the scheduling decision module.

9. The intelligent dispatching system for renewable energy according to claim 8, characterized in that: When formulating the energy scheduling strategy, the scheduling decision module specifically includes the following steps: The charging and discharging strategy of the energy storage device is regarded as a multi-dimensional vector, where each dimension represents the state of a charging and discharging period; The particle swarm optimization algorithm is used to randomly generate a group of particles, each particle represents a charging and discharging strategy, and each particle has its own position and speed information; Initialize individual best position and global best position; The fitness value of each particle is calculated according to the objective function. If the current fitness of the particle is better than the individual best fitness, the individual best position is updated. If it is better than the global best fitness, the global best position is updated. Repeat the calculation and update steps until the maximum number of iterations is reached or the stop condition is met. The objective function is the weighted sum of the peak and valley load demand of the power grid and the operating cost of the energy storage system. Returns the global best position and fitness as a solution to the optimization problem.

10. The intelligent dispatching system for renewable energy according to claim 9, characterized in that: The dispatching system also includes an electricity price adjustment module, which is used to adjust the electricity price of the charging facility according to the energy dispatching strategy, specifically including: Determine the power supply strategy for charging facilities based on the energy storage device charging and discharging strategy, predicted power generation, and grid load power consumption in the energy dispatch strategy; Determine whether the charging facility power supply strategy meets the power demand trend of the charging facility, and calculate the supply-demand ratio; Formulate electricity price adjustment strategies based on the judgment results and supply-demand ratio.

Images