Active response load prediction method and system based on hybrid neural network

Abstract

The invention discloses an active response load prediction method and system based on a hybrid neural network, and the method employs the characteristics of real load data and generates time sequence data as input, a load classifier employs a CNN-LSTM neural network model, employs an AD response model to simulate the active response effect of a load, and employs an RNN neural network model for load prediction. Therefore, the method has the advantages of being novel in scheme and accurate in result, and load feature analysis and preprocessing are carried out on the basic data set containing different types of loads; according to the preprocessed data set, a time series data set is created to serve as a reference data set for load classification research, the affiliation labels of the loads can be predicted, and the loads are classified. The model extracts different types of load features through CNN, introduces a batch normalization layer to standardize the output of the CNN layer, stabilizes the training process, and accelerates convergence; and the dependency relationship between the time steps is captured through the LSTM, so that the output result of the classification model has relatively high accuracy and reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of load classification and prediction, and in particular to an active response load prediction method and system based on a hybrid neural network. Background Art

[0002] With the increasing intelligence of power systems, power demand response technology has gradually become an important means of stable operation of power systems. Active response load classification and prediction are key links in demand response. By accurately classifying and predicting active response loads, reasonable load scheduling and optimization can be achieved, and the operation efficiency and stability of the power grid can be improved.

[0003] Traditional load classification methods are divided into unsupervised clustering and supervised classification. The former is mostly based on various clustering methods such as k-means and fuzzy C-means (FCM), while the latter often uses artificial intelligence neural networks as load classification methods. Clustering algorithms are not suitable for data types with many outliers, and often need to specify the number of categories based on experience, which lacks effective basis; although neural networks have good effects, such as BP neural networks, these classification methods only train the mapping relationship between load curves and category labels from a mathematical level, while ignoring the temporal correlation of load data. In addition, when faced with massive data, the classification efficiency of neural networks will be seriously affected.

[0004] As for load forecasting, since the load components of traditional power systems are simple and the forecasting scenarios are often targeted at system-level or bus-level loads, traditional load forecasting methods are mainly based on statistical analysis methods. However, with the rapid development and continuous evolution of power systems, especially the construction of new power systems will largely introduce a large number of new elements such as distributed new energy and electric vehicles, and the load side of power systems has become more flexible and changeable. Faced with a variety of complex load influencing factors, traditional load forecasting methods are difficult to accurately construct load patterns under the background of new power systems. Therefore, it is urgent to further improve the accuracy of load forecasting under the background of new power systems to ensure the safe and stable operation of the power grid and the balance of power supply and demand. Summary of the invention

[0005] The purpose of the present invention is to provide an active response load forecasting method and system based on a hybrid neural network to solve the problem that it is difficult to accurately predict future loads when multiple active response loads are mixed. The present invention aims to improve the accuracy of load classification and forecasting.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is as follows: The active response load forecasting method based on hybrid neural network includes the following steps: Step S1, obtaining a load basic data set and performing load characteristic analysis to obtain load characteristics, and creating a time series data set based on the load characteristics as a benchmark data set for load classification research; Step S2, constructing a CNN-LSTM neural network model for load classification based on the time series data set; Step S3, constructing an AD response model for simulating the active response effect of load; Step S4, constructing an RNN neural network model for load forecasting; classifying the load data to be predicted based on the CNN-LSTM neural network model, simulating and generating the active response results of the current load based on the AD response model, predicting the active response result data based on the RNN model, and obtaining images of various types of loads after classification and curve comparison images of the predicted load and the actual load after prediction.

[0007] Preferably, obtaining a load basic data set and performing load characteristic analysis to obtain load characteristics specifically includes the following steps: S11: Collect different types of load data and add corresponding labels to different types of data; S12: group by label and date, calculate the statistical characteristics of the load, and form a data frame; S13: extract statistical features and labels from the data frame, standardize the features and encode the labels as integers; S14: Create a time series dataset based on the standardized features, encoded labels and specified time steps, and convert the input load data and encoded labels into time series data suitable for model input.

[0008] Preferably, the statistical features include mean, standard deviation, maximum value and minimum value.

[0009] Preferably, constructing a CNN-LSTM neural network model for load classification based on a time series data set specifically includes the following steps: S21: Dimension conversion, converting the load time series data into the input shape expected by the convolutional layer; S22: Set CNN layer 1 and batch normalization layer 1; CNN layer 1 is used to extract local features of input sequence data and generate feature maps; batch normalization layer 1 is used to standardize the output of the convolution layer; select the ReLU activation function and introduce nonlinear activation; S23: Set the maximum pooling layer to downsample the feature map, reduce the data size, and retain the main features; S24: Set CNN layer 2 and batch normalization layer 2; further extract the output features of CNN layer 1 to generate a higher-level feature map; standardize the output of CNN layer 2, and also select the ReLU activation function to improve the model's expressiveness; S25: Set up the LSTM layer to process the time series data, capture the dependencies between time steps, and take the output of the last time step of the LSTM as a summary of the entire sequence; S26: Set up a fully connected layer to map the output of LSTM to category scores for the final classification decision.

[0010] Preferably, constructing an AD response model for simulating the active response effect of loads specifically includes the following steps: Assuming the consumer response model is linear and time-invariant, a simple finite impulse response model is appropriate:

[0011] Where: k is the discrete time index. ad(k) is the standard AD response load, is an independent and identically distributed distribution with zero mean The error term, used to model the random component, is the actual AD response load. is the user's responsiveness at a certain moment, and is the backward shift operator The polynomial is specifically expressed as:

[0012] in, Items represent the AD profile and user responsiveness based on validation to estimate the actual AD response.

[0013] Preferably, constructing the RNN neural network model for load forecasting specifically includes the following steps: S41: screening and preprocessing active response load data; S42: Set RNN layer 1 and Dropout layer 1. RNN layer 1 is used to extract time series features and return the output of the entire sequence. Dropout layer 1 prevents overfitting. S43: Set RNN layer 2 and Dropout layer 2. RNN layer 2 is used to further extract time series features and return the output of the last time step. Dropout layer 2 prevents overfitting. S44: Set up a fully connected layer to map the output of the RNN to the target space and generate a predicted value; S45: Train the RNN model and monitor the loss on the validation set; use the trained model to predict the test set and denormalize the results; finally generate load forecast values ​​and compare them with the true values.

[0014] Preferably, screening and preprocessing the active response load data include normalization, generating a load time series data set, and splitting a training set and a test set.

[0015] The active response load forecasting system based on hybrid neural network includes data collection and sorting module, load classification module, load active response module and load forecasting module: The data collection and sorting module obtains the basic load data set and performs load characteristic analysis to obtain the load characteristics. Based on the load characteristics, a time series data set is created as a benchmark data set for load classification research. Load classification module, which builds a CNN-LSTM neural network model for load classification based on time series data sets; Active load response module, building an AD response model for simulating the active load response effect; The load forecasting module builds an RNN neural network model for load forecasting; classifies the load data to be predicted based on the CNN-LSTM neural network model, simulates and generates the active response results of the current load based on the AD response model, and predicts the active response result data based on the RNN model to obtain images of various loads after classification and curve comparison images of the predicted load and the actual load after prediction.

[0016] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for monitoring wide-band oscillation of a power grid when executing the computer program.

[0017] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for monitoring broadband oscillation of a power grid are implemented.

[0018] Compared with the prior art, the present invention has the following beneficial technical effects: The active response load forecasting method based on the hybrid neural network of the present invention adopts the characteristics of real load data and generates time series data as input. The load classifier adopts the CNN-LSTM neural network model, and uses the AD response model to simulate the active response effect of the load. The load forecasting adopts the RNN neural network model. Therefore, the present invention has the advantages of novel scheme and accurate results.

[0019] In the present invention, load characteristic analysis and preprocessing are performed on a basic data set containing different types of loads; a time series data set is created based on the preprocessed data set as a benchmark data set for load classification research.

[0020] Based on the CNN-LSTM network model, the present invention can systematically predict the attribution label of the load and classify the load. The model extracts the characteristics of different types of loads through CNN, introduces a batch normalization layer to standardize the output of the CNN layer, stabilizes the training process, and accelerates convergence; and captures the dependency between time steps through LSTM, so that the output results of the classification model have higher accuracy and reliability.

[0021] The present invention introduces an AD response model to simulate the active response effect of the load. Determine the generation method of AD information; generate an AD profile; simulate the AD effect; add the AD effect to the original data set, and add the simulated AD effect to the corresponding time point according to the date and load information of the data set, so as to obtain a data set containing AD information.

[0022] The present invention predicts the classified active response load based on the RNN neural network model. The RNN model is trained and the loss is monitored on the validation set; MAPE is used as the standard for the prediction accuracy of the validation model.

[0023] The present invention will help solve the problem of difficult to accurately predict future loads when multiple active response loads are mixed, improve the accuracy of load prediction in the context of demand response, and ensure safe and stable operation of the power grid and balance of power supply and demand. These and other aspects of the present application will be more concise and easy to understand in the description of the following embodiments. It should be understood that the above general description and the detailed description below are only exemplary and explanatory and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0025] Figure 1 This is a flow chart of an active response load classification and prediction method based on a hybrid neural network according to an embodiment of the present invention.

[0026] Figure 2 This is a model diagram of the CNN-LSTM hybrid neural network of an embodiment of the present invention.

[0027] Figure 3 It is a CNN feature extractor according to an embodiment of the present invention.

[0028] Figure 4 This is the LSTM unit structure of an embodiment of the present invention.

[0029] Figure 5This is a graph showing the experimental results of active response load classification based on a hybrid neural network according to an embodiment of the present invention.

[0030] Figure 6 This is a comparison chart of experimental results of active response load prediction based on hybrid neural network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] The purpose of the present invention is to solve the shortcomings in the above background and propose an active response load classification and prediction method based on a hybrid neural network. In order to make the above objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0034] The present invention provides an active response load forecasting method based on a hybrid neural network, the method comprising the following steps: Step S1, constructing a basic data set for load classification prediction research.

[0035] This step may specifically include: Step S11, obtaining basic load data from the OSF public database; The load data set is obtained from the OSF public database. The data set contains the power load data of different types of buildings in an industrial park with a time resolution of 30 minutes for one year. These data come from smart meters. There are four load types: Public, Office, Residential, and Commercial. Each load type has 365 days of data. The daily data contains 49 data at a time resolution of 30 minutes. Each data has two characteristics: time and corresponding load amount.

[0036] Step S12, preprocessing the load data obtained in step S11 to complete the construction of a benchmark data set for load classification research; Add corresponding labels to different types of data; group by label and date, and calculate the statistical characteristics of the load (mean, standard deviation, maximum, minimum) to form a data frame; extract the statistical features and labels in the data frame, standardize the features and encode the labels as integers. Feature standardization will perform the following transformations on the feature data:

[0037] in, is the mean of the feature data, is the standard deviation of the feature data. After this processing, the distribution of the data is more concentrated in the range of mean 0 and standard deviation 1.

[0038] Create a time series dataset based on the standardized features, encoded labels, and specified time steps. The specific process is as follows:

[0039] Where X is the feature data, Y is the corresponding label, load is the input load data, labels_encoded is the encoded label data, time_steps is the specified time step, and the present invention specifies the time step as 10.

[0040] The active response load classification and prediction method based on hybrid neural network of the present invention further includes the following steps: Step S2, constructing a CNN-LSTM neural network model for load classification.

[0041] Figure 2 This is a system framework diagram used by the present invention for load classification based on the CNN-LSTM model.

[0042] like Figure 2 As shown in the figure, the CNN-LSTM model consists of five parts: CNN layer 1, maximum pooling layer, CNN layer 2, LSTM layer and fully connected layer. Figure 3 Shown is the convolutional neural network (CNN) architecture used in the feature extractor.

[0043] Step S2 may specifically include: Step S21: Dimension conversion, converting the load time series data obtained in step S1 into the input shape expected by the convolutional layer; Step S22: Set CNN layer 1 and batch normalization layer 1; convolution layer 1 is used to extract local features of input sequence data and generate feature maps; batch normalization layer 1 is used to standardize the output of the convolution layer, stabilize the training process, and accelerate convergence; select ReLU activation function, introduce nonlinear activation, and improve the model expression ability; The BN (batch normalization) algorithm standardizes the intermediate calculation data, accelerates the network convergence speed, and reduces the network's sensitivity to weight initialization. The algorithm is described as follows:

[0044] In the above algorithm, batch-size is the size of each batch of sample data; is a minimum value (such as ) avoid The denominator of the expression is 0; and are the mean and variance of the data samples respectively; is the standard normalized intermediate variable; are the parameters that the network needs to learn, which are the scale factor and the translation factor.

[0045] In order to avoid gradient disappearance or gradient explosion, the ReLU activation function is used to process the data after Batch-Normalization. The ReLU activation function can be expressed as

[0046] Step S23: Setting the maximum pooling layer to downsample the feature map, reduce the data size, and retain the main features; Step S24: Set CNN layer 2 and batch normalization layer 2; further extract the output features of convolution layer 1 to generate a higher-level feature map; standardize the output of convolution layer 2 to further stabilize the training process and accelerate convergence; also select the ReLU activation function to improve the model's expressiveness; Step S25: Set up the LSTM layer to process the sequence data, capture the dependencies between time steps, and only take the output of the last time step of the LSTM as a summary of the entire sequence; When ordinary recurrent neural networks are used to process sequence data, they cannot solve the long-term dependency problem, and the gradient diffusion problem makes the model difficult to train. Therefore, the LSTM network is proposed. The LSTM neural network can solve the problem that RNN cannot handle long-distance dependencies, make full use of historical information, and have stronger adaptability in time series data analysis. The unit structure of LSTM is as follows: Figure 4 shown.

[0047] Figure 4 middle: and Represent the input vector and hidden layer state value at time t respectively. The memory unit is the memory of the neuron state, which is used to record the current state. The input gate and output gate are used to read, output and correct parameters. The forget gate is used to selectively forget the unit state at the previous moment and correct the parameters. The calculation formula is as follows:

[0048] Where: f, i, g, C, o represent the forget gate, input gate, alternative cell state for updating, updated cell state and output gate respectively; W and b are the corresponding weight coefficient matrix and bias term respectively; σ and tanh represent the sigmoid activation function and the hyperbolic tangent activation function, respectively.

[0049] Step S26: Set up a fully connected layer to map the output of the LSTM to the category score for the final classification decision.

[0050] The active response load classification and prediction method based on hybrid neural network of the present invention further includes the following steps: Step S3, simulating the active response effect of the load data according to the constructed AD response model, specifically includes: Step S31, assuming that the consumer response model is linear and time-invariant, a simple finite impulse response model can be used to describe The relationship between ad and ; Step S32, generating a specified amount of AD information, including (AD volume), (AD duration) and (Time interval between the start of two continuous AD products); randomly select AD volume in the interval [30, 100] kW and duration In the interval [1, 4]h, Randomly generated in different ranges, [1,24]h and [12, 72]h are taken respectively. AD products mean that in a certain period of time, consumers may reduce or increase power consumption according to the signal, which forms an AD product. Each load data corresponds to a set of information; Step S33, simulate the effect of the AD response model and calculate the impact of AD on the load curve. Generate the basic part of the AD effect, in [ , ] is evenly distributed within the range; the energy recovery range is calculated to be equal to 30% of the total, and its duration is assumed to be (1 / 2) , based on the amplitude and duration of energy recovery, an energy recovery sequence is generated to simulate the energy recovery process after AD in practice. The energy recovery process is the energy recovery process after the demand modification (i.e., AD effect), which may have a modification with the opposite sign to the previous demand modification; the energy recovery sequence is merged with the basic part of the AD effect; considering The current AD effect sequence is padded with a length of (Tint * 2). When the padded AD effect sequence is convolved with the polynomial B (q), the time interval between AD products can be more accurately considered. , during the experiment, take , use convolution to calculate this part; randomly generate the noise part, the noise is a simulated zero-mean independent and identically distributed random number sequence, and add the noise to get the final AD effect; Step S34: for each load data, the original load data is added to the AD effect to obtain the load data after considering the AD effect, and the load data is added to the data set.

[0051] The active response load classification and prediction method based on hybrid neural network of the present invention further includes the following steps: Step S4, constructing a RNN neural network model for load forecasting, specifically including: Step S41: screening and preprocessing the active response load data, including normalization, generating a load time series data set, and splitting a training set and a test set; Step S42: set two layers of RNN and Dropout layer, the RNN layer is used to extract time series features and return the output of the entire sequence; the Dropout layer prevents overfitting; the number of output units of the first layer RNN is 128, the number of output units of the second layer is 64, and ReLU is used as the activation function; the dropout layer discard rate is 0.2; Step S43: Setting a fully connected layer to map the output of the RNN to the target space and generate a predicted value; Step S44: Train the RNN model with 30 training rounds and a batch size of 8. Improve the performance and generalization ability of the model by adjusting the learning rate and using dropout, and monitor the loss on the validation set. Use the trained model to predict the test set and denormalize the results. Finally, generate the load forecast value and compare it with the true value.

[0052] Recurrent Neural Network (RNN) is a neural network used to process sequence data. The basic structure of RNN includes input layer, hidden layer and output layer. In the hidden layer, there are recurrent connections between neurons, so that information can be transmitted in time. Specifically, for the input at time t , the output of the hidden layer Not only depends on the current input , but also depends on the output of the hidden layer at the previous moment This recurrent connection enables RNN to capture the temporal dependencies in sequence data. The mathematical expression of RNN is as follows:

[0053] in, , , They are the weight matrices from input to hidden layer, from hidden layer to hidden layer, and from hidden layer to output layer. , are the bias vectors of the hidden layer and the output layer respectively, and f and g are the activation functions.

[0054] The active response load classification and prediction method based on hybrid neural network of the present invention further includes the following steps: Step S5, combining the CNN-LSTM classification model and the RNN prediction model for testing and verification, also includes the following steps: Step S51, select a verification test data set. This experiment selects the power load data of a building of type Residential in an industrial park in Suzhou, China for one year (from January 1, 2021 to December 31, 2021) with a time resolution of 30 minutes; Step S52, preprocess the data, resample the test data, calculate the statistical characteristics of the load such as the mean, standard deviation, maximum and minimum values ​​on a daily basis; fill missing values ​​with the mean; standardize the characteristics of the test data using the normalizer of the previous training data; create a time series data set using the method used for the training data and specify a time step of 10; Step S53, based on the CNN-LSTM model, the load data to be predicted is predicted, the load labels are obtained, and the average load curves corresponding to different labels are visualized to observe and analyze the classification results more intuitively. The classified images are as follows: Figure 5 As shown; Step S54, simulating and generating the active response effect of the classified load according to the method of step S3, and obtaining load data with active response information; Step S55: predict the active response load data based on the RNN model. Obtain a curve comparison image of the predicted load and the actual load and a prediction error. The predicted curve comparison image is as follows: Figure 6 shown.

[0055] The active response load forecasting system based on hybrid neural network includes data collection and sorting module, load classification module, load active response module and load forecasting module: The data collection and sorting module obtains the basic load data set and performs load characteristic analysis to obtain the load characteristics. Based on the load characteristics, a time series data set is created as a benchmark data set for load classification research. Load classification module, which builds a CNN-LSTM neural network model for load classification based on time series data sets; Active load response module, building an AD response model for simulating the active load response effect; The load forecasting module builds an RNN neural network model for load forecasting; classifies the load data to be predicted based on the CNN-LSTM neural network model, simulates and generates the active response results of the current load based on the AD response model, and predicts the active response result data based on the RNN model to obtain images of various loads after classification and curve comparison images of the predicted load and the actual load after prediction.

[0056] In another embodiment of the present invention, a terminal device is provided, the terminal device includes a processor and a memory, the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGAs), or other processors. GateArray, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, which are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding functions; the processor described in the embodiment of the present invention can be used for the operation of the active response load prediction method based on the hybrid neural network, including the following steps: obtaining a load basic data set and performing load characteristic analysis to obtain load characteristics, creating a time series data set based on the load characteristics as a benchmark data set for load classification research; constructing a CNN-LSTM neural network model for load classification based on the time series data set; constructing an AD response model for simulating the active response effect of the load; constructing an RNN neural network model for load prediction; classifying the load data to be predicted based on the CNN-LSTM neural network model, simulating and generating the active response result of the current load based on the AD response model, predicting the active response result data based on the RNN model, and obtaining images of various types of loads after classification and curve comparison images of the predicted load and the actual load after prediction.

[0057] In another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It is understandable that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory, or a non-volatile memory, such as at least one disk memory.

[0058] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the active response load prediction method based on a hybrid neural network in the above-mentioned embodiment; one or more instructions in the computer-readable storage medium are loaded by the processor and execute the following steps: obtaining a load basic data set and performing load characteristic analysis to obtain load characteristics, creating a time series data set based on the load characteristics as a benchmark data set for load classification research; constructing a CNN-LSTM neural network model for load classification based on the time series data set; constructing an AD response model for simulating the active response effect of the load; constructing an RNN neural network model for load prediction; classifying the load data to be predicted based on the CNN-LSTM neural network model, simulating and generating the active response result of the current load based on the AD response model, predicting the active response result data based on the RNN model, and obtaining images of various types of loads after classification and curve comparison images of the predicted load and the actual load after prediction.

[0059] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0060] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0061] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0062] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0063] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. Active response load forecasting method based on hybrid neural network, characterized in that: The following steps are involved: S1, obtain the load basic data set and perform load characteristic analysis to obtain the load characteristics, and create a time series data set based on the load characteristics as the benchmark data set for load classification research; S2, building a CNN-LSTM neural network model for load classification based on the time series dataset; S3, constructing an AD response model for simulating the active response effect of load; S4, construct an RNN neural network model for load forecasting; classify the load data to be predicted based on the CNN-LSTM neural network model, simulate and generate the active response results of the current load based on the AD response model, predict the active response result data based on the RNN model, and obtain images of various types of loads after classification and curve comparison images of the predicted load and the actual load after prediction.

2. The active response load forecasting method based on hybrid neural network according to claim 1 is characterized in that: Obtaining the load basic data set and performing load characteristic analysis to obtain the load characteristics specifically includes the following steps: S11: Collect different types of load data and add corresponding labels to different types of data; S12: group by label and date, calculate the statistical characteristics of the load, and form a data frame; S13: extract statistical features and labels from the data frame, standardize the features and encode the labels as integers; S14: Create a time series dataset based on the standardized features, encoded labels and specified time steps, and convert the input load data and encoded labels into time series data suitable for model input.

3. The active response load forecasting method based on hybrid neural network according to claim 2 is characterized in that: The statistical characteristics include mean, standard deviation, maximum value and minimum value.

4. The active response load forecasting method based on hybrid neural network according to claim 1 is characterized in that: The construction of a CNN-LSTM neural network model for load classification based on a time series dataset includes the following steps: S21: Dimension conversion, converting the load time series data into the input shape expected by the convolutional layer; S22: Set CNN layer 1 and batch normalization layer 1; CNN layer 1 is used to extract local features of input sequence data and generate feature maps; batch normalization layer 1 is used to standardize the output of the convolution layer; select the ReLU activation function and introduce nonlinear activation; S23: Set the maximum pooling layer to downsample the feature map, reduce the data size, and retain the main features; S24: Set CNN layer 2 and batch normalization layer 2; further extract the output features of CNN layer 1 to generate a higher-level feature map; standardize the output of CNN layer 2, and also select the ReLU activation function to improve the model's expressiveness; S25: Set up the LSTM layer to process the time series data, capture the dependencies between time steps, and take the output of the last time step of the LSTM as a summary of the entire sequence; S26: Set up a fully connected layer to map the output of LSTM to category scores for the final classification decision.

5. The active response load forecasting method based on hybrid neural network according to claim 1 is characterized in that: The construction of the AD response model for simulating the active response effect of loads specifically includes the following steps: Assuming the consumer response model is linear and time-invariant, a simple finite impulse response model is appropriate: Where: k is the discrete time index; ad(k) is the standard AD response load, is an independent and identically distributed distribution with zero mean The error term, used to model the random component, is the actual AD response load; is the user's responsiveness at a certain moment, and is the backward shift operator The polynomial is specifically expressed as: in, Items represent the AD profile and user responsiveness based on validation to estimate the actual AD response.

6. The active response load forecasting method based on hybrid neural network according to claim 1 is characterized in that: The construction of the RNN neural network model for load forecasting specifically includes the following steps: S41: screening and preprocessing active response load data; S42: Set RNN layer 1 and Dropout layer 1. RNN layer 1 is used to extract time series features and return the output of the entire sequence. Dropout layer 1 prevents overfitting. S43: Set RNN layer 2 and Dropout layer 2. RNN layer 2 is used to further extract time series features and return the output of the last time step. Dropout layer 2 prevents overfitting. S44: Set up a fully connected layer to map the output of the RNN to the target space and generate a predicted value; S45: Train the RNN model and monitor the loss on the validation set; use the trained model to predict the test set and denormalize the results; finally generate load forecast values ​​and compare them with the true values.

7. The active response load forecasting method based on hybrid neural network according to claim 6 is characterized in that: The screening and preprocessing of active response load data include normalization, generation of load time series datasets, and splitting of training and testing sets.

8. Active response load forecasting system based on hybrid neural network, characterized in that: Including data collection and sorting module, load classification module, load active response module and load forecasting module: The data collection and sorting module obtains the basic load data set and performs load characteristic analysis to obtain the load characteristics. Based on the load characteristics, a time series data set is created as a benchmark data set for load classification research. Load classification module, which builds a CNN-LSTM neural network model for load classification based on time series data sets; Active load response module, building an AD response model for simulating the active load response effect; The load forecasting module builds an RNN neural network model for load forecasting; classifies the load data to be predicted based on the CNN-LSTM neural network model, simulates and generates the active response results of the current load based on the AD response model, and predicts the active response result data based on the RNN model to obtain images of various loads after classification and curve comparison images of the predicted load and the actual load after prediction.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method for monitoring wide-band oscillation of a power grid as claimed in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for monitoring wide-band oscillation of a power grid as claimed in any one of claims 1 to 7 are implemented.

Images