Dynamic frequency modulation control method and system for gas turbine power generation equipment

Abstract

The invention relates to the field of dynamic frequency modulation control, and provides a dynamic frequency modulation control method and system for gas turbine power generation equipment, and the method comprises the steps: obtaining sequential gas turbine power generation equipment operation parameters and power grid frequency parameters, and carrying out the normalization processing, thereby obtaining a normalized frequency modulation data set; constructing a gas turbine dynamic model and a power grid dynamic load model based on the normalized frequency modulation data set; obtaining a gas turbine load set and a power grid frequency set based on the gas turbine dynamic model and the power grid dynamic load model; establishing a dynamic frequency modulation control model based on frequency modulation dynamic characteristics, primary frequency modulation success rate and frequency modulation energy consumption of the gas turbine dynamic model triggered by power grid frequency change; and performing dynamic frequency modulation control on the gas turbine power generation equipment by using the dynamic frequency modulation control model. The method overcomes the problem that the existing model cannot accurately respond to the power grid frequency change by combining the primary frequency modulation success rate and the frequency modulation energy consumption and cannot accurately respond to the primary frequency modulation under a complex condition, so that the primary frequency modulation success rate is reduced.

Description

Technical Field

[0001] This invention relates to the field of dynamic frequency regulation control technology for gas turbine power generation equipment, and particularly to a dynamic frequency regulation control method and system for gas turbine power generation equipment. Background Technology

[0002] Gas turbine power generation equipment is an advanced power machinery device that uses a continuously flowing gaseous working fluid to drive a high-speed rotating impeller, converting the energy of fuel into useful work. It is a type of rotating impeller thermal engine. Gas turbines have advantages such as light weight, small size, quick start-up, convenient maintenance, reliable operation, low pollution, high thermal efficiency, and good peak-shaving performance. The main objective of traditional gas turbine generator set control systems is to implement a constant incremental rate of operation, which limits their role in grid frequency regulation.

[0003] Chinese patent application CN113919249A discloses a dynamic simulation modeling method for heavy-duty single-shaft gas turbine generator sets. This method includes: establishing a static model of the performance parameters of the heavy-duty gas turbine generator set; correcting external disturbances related to environmental parameters; modeling grid frequency disturbances after grid connection; theoretically modeling the process quantities—pressure ratio, compressor temperature ratio, airflow, turbine expansion ratio, and gas turbine combustion temperature—using the unit's physical characteristic model; performing dynamic characteristic correction; and integrating and establishing a dynamic mathematical model of the heavy-duty single-shaft gas turbine generator set. However, the dynamic mathematical model in the above method cannot accurately respond to changes in grid frequency by combining primary frequency regulation success rate and frequency regulation energy consumption, thus failing to accurately respond to primary frequency regulation under complex conditions, resulting in a reduced primary frequency regulation success rate. Summary of the Invention

[0004] The present invention aims to solve at least one of the problems existing in the prior art, and to provide a dynamic frequency regulation control method and system for gas turbine power generation equipment.

[0005] In one aspect, the present invention provides a dynamic frequency regulation control method for a gas turbine power generation device, the dynamic frequency regulation control method comprising:

[0006] The operating parameters of the gas turbine power generation equipment and the grid frequency parameters are obtained in a time sequence. The operating parameters of the gas turbine power generation equipment and the grid frequency parameters are normalized to obtain a normalized frequency regulation dataset.

[0007] Based on the normalized frequency regulation dataset, a dynamic model of the gas turbine and a dynamic load model of the power grid, which are respectively constructed as islands, are constructed.

[0008] Execute the gas turbine dynamic model and the power grid dynamic load model, and obtain the gas turbine load set and the power grid frequency set based on the gas turbine dynamic model and the power grid dynamic load model, respectively;

[0009] Based on the gas turbine load set and the power grid frequency set, and considering the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by power grid frequency changes, a dynamic frequency regulation control model is established that takes into account the success rate of primary frequency regulation and frequency regulation energy consumption.

[0010] The gas turbine load set and the power grid frequency set are used as inputs to the dynamic frequency regulation control model. The dynamic frequency regulation control model is executed, and the gas turbine power generation equipment is dynamically regulated using the dynamic frequency regulation control model.

[0011] Optionally, the normalization process for the operating parameters of the gas turbine power generation equipment and the grid frequency parameters to obtain a normalized frequency regulation dataset includes:

[0012] The operating parameters of the gas turbine power generation equipment and the grid frequency parameters are loaded in real time. Outlier point processing is performed on the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, and abnormal values ​​in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters are deleted.

[0013] The operating parameters of the gas turbine power generation equipment and the grid frequency parameters after removing outliers are standardized. The standardized operating parameters of the gas turbine power generation equipment and the grid frequency parameters are then normalized, and a normalized set is output. The normalization method is Box-Cox transformation.

[0014] Load the normalized set and process it using max-min normalization, wherein the max-min normalization formula is:

[0015]

[0016] Where, x z X represents the normally distributed data output after max-min normalization, and X represents the normally distributed data in the input normalization set. max X min These are the maximum and minimum values ​​of the normally distributed data in the normalized set, respectively.

[0017] Obtain the normalized set after normalization, introduce white noise to iteratively filter the normalized set, and obtain the normalized frequency modulation dataset.

[0018] Optionally, the iterative filtering of the normalized set after normalization by introducing white noise to obtain the normalized frequency modulated dataset includes:

[0019] The normalized set is loaded, and white noise with amplitude ε is added to the normalized set. Integrated empirical mode decomposition is then performed on the normalized set with added white noise, and the decomposed components are output. The decomposed components are represented as follows:

[0020]

[0021] Where EEMD(IMF,t) represents the decomposed component, t represents the sampling time, a represents the number of iterations of white noise ε, and δ(x) represents the decomposed component. z ) represents the output quantity x of the normalized set. z The Dirac function, where e represents the natural constant;

[0022] Define a noise tolerance and an iteration counter. Calculate the frequency domain modes of the decomposed components based on the noise tolerance and the iteration counter. Obtain the mean value of the frequency domain modes as the decomposed mean.

[0023] The decomposition average is expressed as:

[0024]

[0025] Wherein, EEMD(IMF,t) τ U represents the decomposition mean. b (x z ) represents the frequency domain mode of the decomposed component at the b-th iteration, u b+1 (x z ) represents the frequency domain mode of the decomposed component at the (b+1)th iteration, b represents the iteration number recorded by the iteration counter, and χ represents the noise tolerance;

[0026] Perform discrete wavelet transform on the decomposed mean to output the normalized frequency modulated dataset.

[0027] Optionally, the step of constructing a gas turbine dynamic model and a power grid dynamic load model as islands based on the normalized frequency regulation dataset includes:

[0028] Load the normalized frequency regulation dataset and identify the gas turbine type data, start-stop / variable load data, fuel data, and environmental data associated with the gas turbine in the normalized frequency regulation dataset;

[0029] Using gas turbine type data, start-stop / variable load data, fuel data, and environmental data as constraints, a visual component model of the gas turbine is established. The visual component model of the gas turbine includes a regulating valve module, a regenerative cycle module, a compressor module, a combustion chamber module, a turbine module, a regenerator module, and a moment of inertia model.

[0030] The loads in the gas turbine visualization component model are divided into rigid loads and flexible loads. The rigid loads and flexible loads in the gas turbine visualization component model are described using a Markov process. The overall disturbance of the gas turbine visualization component model is calculated based on the weighted calculation of rigid load disturbances and flexible load disturbances. The level changes of the rigid load disturbances and flexible load disturbances both follow the probability matrix of the Markov process.

[0031] Simulation software is used to perform simulation tests on the visualized component model of the gas turbine, and the dynamic model of the gas turbine is output.

[0032] Optionally, the step of constructing a gas turbine dynamic model and a power grid dynamic load model as islands based on the normalized frequency regulation dataset includes:

[0033] Load the normalized frequency regulation dataset and identify the active power, reactive power, power factor, frequency parameters, and response time associated with the power grid in the normalized frequency regulation dataset;

[0034] The active power, reactive power, power factor, frequency parameters, and response time associated with the power grid in the normalized frequency regulation dataset are fused to obtain a modeling fusion set, which is then divided into a training set and a validation set.

[0035] An initial model for a dynamic load model of a power grid based on a combination of long short-term memory neural networks and convolutional neural networks is established, and the loss function, iteration rounds, and dynamic hyperparameters of the initial model are defined.

[0036] The initial model of the power grid dynamic load model is trained based on the training set, the validation set, the loss function, the iteration rounds, and the dynamic hyperparameters to obtain the power grid dynamic load model.

[0037] Optionally, training the initial model of the power grid dynamic load model based on the training set, the validation set, the loss function, the number of iterations, and the dynamic hyperparameters to obtain the power grid dynamic load model includes:

[0038] Using the training set as input, the initial model of the power grid dynamic load model is executed. Based on the loss function, the number of iterations, and the dynamic hyperparameters, the initial model of the power grid dynamic load model is iteratively trained to obtain a converged power grid dynamic load model.

[0039] Using the validation set as input, the converged power grid dynamic load model is executed, and the validation result is output. The validity of the validation result is determined based on the model accuracy threshold. If the validation result meets the model accuracy threshold, the converged power grid dynamic load model is output.

[0040] Optionally, the step of establishing a dynamic frequency regulation control model that considers the primary frequency regulation success rate and frequency regulation energy consumption based on the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by changes in the grid frequency, according to the gas turbine load set and the grid frequency set, includes:

[0041] The gas turbine load set and the power grid frequency set are obtained, and the gas turbine load set and the power grid frequency set are standardized. The standardized gas turbine load set and the power grid frequency set are processed based on principal component analysis to extract the feature factors of the gas turbine load set and the power grid frequency set.

[0042] The feature factors are loaded, and the dynamic frequency modulation control model is constructed based on the partial least squares method of the neural network combined with the feature factors. The internal relationship of the dynamic frequency modulation control model is solved by the neural network.

[0043] The dynamic frequency modulation control model is loaded, and the input-output mapping relationship of the feature factors is solved using a three-layer BP network. The number of neural network nodes of the dynamic frequency modulation control model is determined based on the cross-validation method, and the converged dynamic frequency modulation control model is output.

[0044] The first frequency modulation success rate and the frequency modulation energy consumption are introduced into the dynamic frequency modulation control model as constraints, and the first frequency modulation success rate and the frequency modulation energy consumption residual are calculated.

[0045] The final dynamic frequency modulation control model is determined based on the calculation results of the first frequency modulation success rate and the frequency modulation energy consumption residual.

[0046] Optionally, the objective function of the dynamic frequency modulation control model is defined as:

[0047]

[0048] Where f(·) represents the objective function of the dynamic frequency modulation control model, P s f s Let be the input representations of the gas turbine load set and the power grid frequency set, respectively, min(ys ) represents the success rate of a single frequency modulation (FM) y. s Minimization is a constraint condition for Q. s Indicates frequency modulation energy consumption, C s T s u s These are the specific heat capacity influence coefficient, temperature, and fuel flow rate, respectively. R and L represent the gas constant and equipment load torque, respectively, V s Let υ represent fuel velocity, η represent reduced velocity, and η represent gas turbine efficiency.

[0049] In another aspect, the present invention provides a dynamic frequency regulation control system for a gas turbine power generation device, the dynamic frequency regulation control system comprising:

[0050] The parameter acquisition module is used to acquire time-series operating parameters of the gas turbine power generation equipment and grid frequency parameters, and to normalize the operating parameters of the gas turbine power generation equipment and the grid frequency parameters to obtain a normalized frequency regulation dataset.

[0051] The model building module is used to construct a gas turbine dynamic model and a power grid dynamic load model as islands based on the normalized frequency regulation dataset; execute the gas turbine dynamic model and the power grid dynamic load model, and obtain the gas turbine load set and the power grid frequency set based on the gas turbine dynamic model and the power grid dynamic load model, respectively;

[0052] The dynamic frequency regulation module is used to establish a dynamic frequency regulation control model that considers the success rate of primary frequency regulation and frequency regulation energy consumption based on the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by changes in the grid frequency, according to the gas turbine load set and the grid frequency set; the dynamic frequency regulation control model is executed by using the gas turbine load set and the grid frequency set as inputs to the dynamic frequency regulation control model, and the dynamic frequency regulation control model is used to perform dynamic frequency regulation control on the gas turbine power generation equipment.

[0053] Optionally, the parameter acquisition module specifically includes:

[0054] An outlier processing unit is used to load the operating parameters of the gas turbine power generation equipment and the grid frequency parameters in real time, perform outlier processing on the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, and delete outliers in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters.

[0055] The normalization processing unit is used to standardize the operating parameters of the gas turbine power generation equipment and the grid frequency parameters after removing outliers, and to normalize the standardized operating parameters of the gas turbine power generation equipment and the grid frequency parameters, outputting a normalized set; wherein, the normalization processing method is Box-Cox transform;

[0056] A normalization unit is used to load the normalized set, process the normalized set using max-min normalization, obtain the normalized set, introduce white noise to iteratively filter the normalized set, and obtain the normalized frequency modulation dataset.

[0057] Compared with the prior art, this invention overcomes the problem that the dynamic mathematical model in existing methods cannot accurately respond to changes in grid frequency by combining the primary frequency regulation success rate and frequency regulation energy consumption, thus failing to accurately respond to primary frequency regulation under complex conditions, resulting in a reduced primary frequency regulation success rate. It also has the following beneficial effects:

[0058] 1. By establishing a dynamic frequency regulation control model that considers the success rate of primary frequency regulation and frequency regulation energy consumption, the success rate of primary frequency regulation of the dynamic frequency regulation control model can be improved by combining the gas turbine load and the grid frequency, and the frequency regulation energy consumption can be reduced. This allows for more precise control of the gas turbine's output power, more accurate response to changes in grid frequency, improved success rate of primary frequency regulation, and faster response to changes in grid frequency, which helps maintain grid stability.

[0059] 2. By performing outlier processing, normalization, and iterative filtering on the operating parameters of gas turbine power generation equipment and the grid frequency parameters, the operating status of gas turbine power generation equipment and the stability of grid frequency can be assessed more accurately. This results in smoothed data with better stability, reducing misjudgments and overreactions caused by excessive data fluctuations. It also helps to promptly identify potential safety hazards and take corresponding measures to address them.

[0060] 3. This invention introduces white noise to perform iterative filtering on the normalized set, and outputs a normalized frequency modulation dataset after performing discrete wavelet transform on the decomposed mean. The iterative filtering process smooths the random fluctuations and noise in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters by applying the filter multiple times. It can retain useful trend information, which helps to more accurately grasp the operating status of the gas turbine power generation equipment and the changing trend of the grid frequency, providing a more reliable basis for prediction and decision-making. The data after iterative filtering is more stable and regular, which helps to improve the accuracy and reliability of the dynamic frequency modulation control model.

[0061] 4. A dynamic load model for power grids and its construction method are provided. The dynamic load model for power grids is constructed by combining long short-term memory neural networks and convolutional neural networks. By combining convolutional neural networks and long short-term memory neural networks, both spatial and temporal features in the time series can be considered simultaneously, thereby improving the accuracy and stability of the dynamic load model prediction. The combination of convolutional neural networks and long short-term memory neural networks can also capture the complex patterns and trends in power load data more comprehensively, enhancing the model's adaptability and data processing capabilities, and providing strong support for the efficient operation of smart grids. Attached Figure Description

[0062] One or more embodiments are illustrated by way of example with the corresponding pictures in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0063] Figure 1 A flowchart illustrating a dynamic frequency regulation control method for a gas turbine power generation device according to an embodiment of the present invention;

[0064] Figure 2 A flowchart illustrating the specific steps of obtaining the normalized frequency modulation dataset in step S10, as provided in another embodiment of the present invention;

[0065] Figure 3 A flowchart illustrating the specific steps of obtaining the normalized frequency modulation dataset in step S104, as provided in another embodiment of the present invention;

[0066] Figure 4 A flowchart illustrating the specific steps of step S20 in constructing the dynamic model of the gas turbine, as provided in another embodiment of the present invention;

[0067] Figure 5 A flowchart illustrating the specific steps of step S20 in constructing a dynamic load model of the power grid, as provided in another embodiment of the present invention;

[0068] Figure 6 A flowchart illustrating the specific steps included in step S304, as provided in another embodiment of the present invention;

[0069] Figure 7 A flowchart illustrating the specific steps included in step S40, as provided in another embodiment of the present invention;

[0070] Figure 8 This is a schematic diagram of the structure of a dynamic frequency regulation control system for a gas turbine power generation device, provided as another embodiment of the present invention. Detailed Implementation

[0071] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details are presented in the various embodiments of the present invention to facilitate a better understanding of the invention. However, the technical solutions claimed in the present invention can be implemented even without these technical details and with various variations and modifications based on the following embodiments. The division of the various embodiments below is for ease of description and should not constitute any limitation on the specific implementation of the present invention. The various embodiments can be combined with and referenced by each other without contradiction.

[0072] One embodiment of the present invention relates to a dynamic frequency regulation control method for a gas turbine power generation device, the process of which is as follows: Figure 1 As shown, it includes steps S10 to S50.

[0073] Step S10: Obtain the time-series operating parameters of the gas turbine power generation equipment and the grid frequency parameters, and normalize the operating parameters of the gas turbine power generation equipment and the grid frequency parameters to obtain a normalized frequency regulation dataset. The operating parameters of the gas turbine power generation equipment include the shaft power, frequency, voltage, speed, temperature, pressure, and flow rate of the power generation equipment. The grid frequency parameters include the system frequency, speed inequality, speed dead zone, and limiting.

[0074] Step S20: Based on the normalized frequency regulation dataset, construct the dynamic model of the gas turbine as an island and the dynamic load model of the power grid.

[0075] Step S30: Execute the gas turbine dynamic model and the power grid dynamic load model, and obtain the gas turbine load set and the power grid frequency set based on the gas turbine dynamic model and the power grid dynamic load model, respectively. Specifically, input the gas turbine power generation equipment operating parameters from the normalized frequency regulation dataset into the gas turbine dynamic model to obtain the gas turbine load set output by the gas turbine dynamic model, and input the power grid frequency parameters from the normalized frequency regulation dataset into the power grid dynamic load model to obtain the power grid frequency set output by the power grid dynamic load model.

[0076] Step S40: Based on the gas turbine load set and the power grid frequency set, and based on the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by power grid frequency changes, establish a dynamic frequency regulation control model that considers the success rate of primary frequency regulation and frequency regulation energy consumption.

[0077] Step S50: The gas turbine load set and the grid frequency set are used as inputs to the dynamic frequency regulation control model. The dynamic frequency regulation control model is executed, and the gas turbine power generation equipment is dynamically regulated using the dynamic frequency regulation control model.

[0078] Specifically, when implementing the dynamic frequency regulation control model, the established dynamic frequency regulation control model of the gas turbine is used to control the dynamic processes of gas turbine startup, shutdown and frequency regulation, and a reasonable primary frequency regulation strategy is formulated in combination with the grid dispatch requirements to ensure that the gas turbine can provide timely and effective support when the grid frequency fluctuates.

[0079] The dynamic frequency regulation control method for gas turbine power generation equipment provided by this invention overcomes the problem that existing methods cannot accurately respond to changes in grid frequency by combining the primary frequency regulation success rate and frequency regulation energy consumption, thus failing to accurately respond to primary frequency regulation under complex conditions and reducing the primary frequency regulation success rate. By establishing a dynamic frequency regulation control model that considers the primary frequency regulation success rate and frequency regulation energy consumption, the primary frequency regulation success rate of the dynamic frequency regulation control model can be improved by combining the gas turbine load and grid frequency, and the frequency regulation energy consumption can be reduced. This allows for more precise control of the gas turbine output power, more accurate response to changes in grid frequency, improved primary frequency regulation success rate, and faster response to changes in grid frequency, thus helping to maintain grid stability.

[0080] For example, in step S10, the operating parameters of the gas turbine power generation equipment and the grid frequency parameters are normalized to obtain a normalized frequency regulation dataset, including steps S101 to S104. The following is in conjunction with... Figure 2 Steps S101 to S104 will be explained in detail.

[0081] Step S101: Load the operating parameters of the gas turbine power generation equipment and the grid frequency parameters in real time, perform outlier processing on the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, and delete the abnormal values ​​in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters.

[0082] It should be noted that in step S101, when processing outliers in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, outliers that are significantly different from the overall data distribution can first be identified using statistical analysis methods such as calculating the mean and standard deviation. Then, outliers can be removed using a deletion method, and the causes of their occurrence can be further analyzed. For example, outliers may be caused by equipment failure, operational errors, or external interference. Understanding the causes of outliers helps in taking appropriate measures to prevent similar situations from recurring.

[0083] Step S102 involves standardizing the operating parameters of the gas turbine generator and the grid frequency parameters after removing outliers, and then normalizing the standardized operating parameters and grid frequency parameters to output a normalized set. The normalization method is Box-Cox transform.

[0084] Specifically, by standardizing the operating parameters of gas turbine power generation equipment and grid frequency parameters after removing outliers, the problem of data from different data sources potentially having different dimensions and distribution characteristics can be addressed. Normalization transforms data from different data sources to the same scale, facilitating data fusion and comprehensive analysis. This is of great significance for achieving comprehensive monitoring and intelligent management of power systems.

[0085] Step S103: Load the normalized set and process the normalized set using max-min normalization.

[0086] The maximum-min normalization formula is as follows: Where, x z Let X represent the normally distributed data output after max-min normalization, and let X represent the normally distributed data in the input normalization set. max X min These are the maximum and minimum values ​​of the normally distributed data in the normalized set, respectively.

[0087] Step S104: Obtain the normalized set after normalization, introduce white noise to iteratively filter the normalized set, and obtain the normalized frequency modulation dataset.

[0088] This implementation method, by performing outlier processing, normalization, and iterative filtering on the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, can more accurately assess the operating status of the gas turbine power generation equipment and the stability of the grid frequency. This results in smoothed data with better stability, reduces misjudgments and overreactions caused by excessive data fluctuations, and helps to promptly identify potential safety hazards and take corresponding measures to address them.

[0089] For example, in step S104, white noise is introduced to iteratively filter the normalized set after normalization processing to obtain a normalized frequency-modulated dataset, including steps S1041 to S1043. The following is in conjunction with... Figure 3 Steps S1041 to S1043 will be explained in detail.

[0090] Step S1041: Load the normalized set after normalization, add white noise with amplitude ε to the normalized set after normalization, use integrated empirical mode decomposition to add white noise to the normalized set, and output the decomposed components.

[0091] The decomposed components are represented as follows:

[0092]

[0093] Where EEMD(IMF,t) represents the decomposition components, t represents the sampling time, a represents the number of iterations of white noise ε, and δ(x) represents the decomposition components. z ) represents the output quantity x of the normalized set. z The Dirac function, where e represents the natural constant.

[0094] Step S1042: Define noise tolerance and iteration counter, calculate the frequency domain modes of the decomposed components based on noise tolerance and iteration counter, and obtain the mean of the frequency domain modes as the decomposed mean.

[0095] The mean of the decomposition is expressed as:

[0096]

[0097] Wherein, EEMD(IMF,t) τ U represents the decomposition mean. b (x z ) represents the frequency domain mode of the decomposed component at the b-th iteration, u b+1 (x z ) represents the frequency domain mode of the decomposed component at the (b+1)th iteration, b represents the iteration number recorded by the iteration counter, and χ represents the noise tolerance.

[0098] Step S1043: Perform discrete wavelet transform on the decomposed mean and output a normalized frequency modulated dataset.

[0099] This implementation introduces white noise for iterative filtering of the normalized set, and performs discrete wavelet transform on the decomposed mean to output a normalized frequency modulation dataset. The iterative filtering process smooths out random fluctuations and noise in the operating parameters of the gas turbine generator and the grid frequency parameters by applying filters multiple times. This preserves useful trend information, helping to more accurately grasp the operating status of the gas turbine generator and the changing trends of the grid frequency, providing a more reliable basis for prediction and decision-making. The data after iterative filtering is more stable and regular, contributing to improving the accuracy and reliability of the dynamic frequency modulation control model.

[0100] For example, step S20 involves constructing a dynamic model of the gas turbine as an island and a dynamic load model of the power grid based on the normalized frequency regulation dataset, including steps S201 to S205. Steps S201 to S205 are used to construct the dynamic model of the gas turbine based on the normalized frequency regulation dataset, which will be discussed below. Figure 4 Steps S201 to S205 will be explained in detail.

[0101] Step S201: Load the normalized frequency regulation dataset and identify the gas turbine type data, start-stop / variable load data, fuel data, and environmental data associated with the gas turbine in the normalized frequency regulation dataset.

[0102] Step S202: Using gas turbine type data, start-up / load variation data, fuel data, and environmental data as constraints, establish a visual component model of the gas turbine. This visual component model includes a regulating valve module, a regenerative cycle module, a compressor module, a combustion chamber module, a turbine module, a regenerator module, and a moment of inertia model.

[0103] Step S203: Divide the loads in the gas turbine visualization component model into rigid loads and flexible loads, and use a Markov process to describe the rigid loads and flexible loads in the gas turbine visualization component model.

[0104] Step S204: Calculate the overall disturbance of the gas turbine visualization component model based on a weighted average of rigid load disturbances and flexible load disturbances. The level changes of both rigid and flexible load disturbances follow a Markov probability matrix.

[0105] It should be noted that in step S204, when calculating the overall disturbance of the gas turbine visualization component model based on the weighted calculation of rigid load disturbance and flexible load disturbance, appropriate weights can first be assigned to the rigid and flexible loads according to the system requirements and load characteristics. Then, the assigned weights are used to perform a weighted calculation of the rigid and flexible load disturbances to obtain an estimate of the overall disturbance. The weights assigned to the rigid and flexible loads reflect the importance of different types of loads to the system's stability and responsiveness.

[0106] Step S205: Using simulation software, perform simulation tests on the visualized component model of the gas turbine and output the dynamic model of the gas turbine.

[0107] Specifically, simulation software such as Matlab and Simulink can be selected based on actual needs. By using simulation software such as Matlab and Simulink to simulate and test the visualized component models of gas turbines, the dynamic models of gas turbines can intuitively demonstrate the internal structure and working principles of the gas turbines, helping users to better understand and analyze the models.

[0108] For example, step S20 involves constructing a dynamic model of the gas turbine as an island and a dynamic load model of the power grid based on the normalized frequency regulation dataset, and also includes steps S301 to S304. Steps S301 to S304 are used to construct the dynamic load model of the power grid based on the normalized frequency regulation dataset, which will be discussed below. Figure 5 Steps S301 to S304 will be explained in detail.

[0109] Step S301: Load the normalized frequency regulation dataset and identify the active power, reactive power, power factor, frequency parameters, and response time associated with the power grid in the normalized frequency regulation dataset.

[0110] Step S302: Perform feature fusion on the active power, reactive power, power factor, frequency parameters, and response time associated with the power grid in the normalized frequency regulation dataset to obtain a modeling fusion set. Divide the modeling fusion set into a training set and a validation set. For example, the ratio of the training set to the validation set can be set to 3-4:1.

[0111] It should be noted that in step S302, when performing feature fusion on active power, reactive power, power factor, frequency parameters, and response time associated with the power grid, feature selection algorithms (such as recursive feature elimination, model-based feature selection, etc.) can first be used to select the most representative and discriminative feature subset from the extracted features to reduce feature dimensionality and improve the efficiency and accuracy of subsequent modeling. Then, these feature subsets are fused using simple methods such as concatenation, weighted summation, and principal component analysis (PCA). The purpose of feature fusion is to integrate the information from multiple features to form a more comprehensive feature representation.

[0112] Step S303: Establish an initial model of the power grid dynamic load model based on the combination of long short-term memory neural network and convolutional neural network, and define the loss function, iteration rounds and dynamic hyperparameters of the initial model of the power grid dynamic load model.

[0113] Step S304: Train the initial model of the power grid dynamic load model based on the training set, validation set, loss function, iteration rounds, and dynamic hyperparameters to obtain the power grid dynamic load model.

[0114] For example, step S304 involves training the initial model of the power grid dynamic load model based on the training set, validation set, loss function, iteration rounds, and dynamic hyperparameters to obtain the power grid dynamic load model, including steps S3041 to S3043. The following is in conjunction with... Figure 6 Steps S3041 to S3043 will be explained in detail.

[0115] Step S3041: Using the training set as input, execute the initial model of the power grid dynamic load model. Based on the loss function, iteration rounds, and dynamic hyperparameters, iteratively train the initial model of the power grid dynamic load model to obtain a converged power grid dynamic load model.

[0116] Step S3042: Using the validation set as input, execute the converged power grid dynamic load model and output the validation results.

[0117] Step S3043: Determine whether the verification result is qualified based on the model accuracy threshold. If the verification result meets the model accuracy threshold, output the converged power grid dynamic load model. If the verification result does not meet the model accuracy threshold, return to step S3041 to continue iterative training.

[0118] This embodiment provides a dynamic load model for power grids and its construction method. The dynamic load model is constructed by combining a long short-term memory neural network (LSTM) and a convolutional neural network (CNN). By combining CNNs and LSTMs, both spatial and temporal features in the time series can be considered simultaneously, thereby improving the accuracy and stability of the dynamic load model's predictions. Furthermore, the combination of CNNs and LSTMs can more comprehensively capture complex patterns and trends in power load data, enhancing the model's adaptability and data processing capabilities, and providing strong support for the efficient operation of smart grids.

[0119] For example, step S40, which involves establishing a dynamic frequency regulation control model considering the primary frequency regulation success rate and frequency regulation energy consumption based on the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by changes in the grid frequency, according to the gas turbine load set and grid frequency set, includes steps S401 to S405. The following is in conjunction with... Figure 7 Steps S401 to S405 will be explained in detail.

[0120] Step S401: Obtain the gas turbine load set and the power grid frequency set, standardize the gas turbine load set and the power grid frequency set, process the standardized gas turbine load set and the power grid frequency set based on principal component analysis, and extract the feature factors of the gas turbine load set and the power grid frequency set.

[0121] Step S402: Load feature factors, construct a dynamic frequency modulation control model based on the partial least squares method of neural network combined with feature factors, and use neural network to solve the internal relationship of dynamic frequency modulation control model.

[0122] Step S403: Load the dynamic frequency modulation control model, use a three-layer BP network to solve the input-output mapping relationship of the feature factors, and determine the number of neural network nodes of the dynamic frequency modulation control model based on the cross-validation method, and output the converged dynamic frequency modulation control model.

[0123] Step S404: Introduce the primary frequency modulation success rate and frequency modulation energy consumption as constraints into the dynamic frequency modulation control model, and calculate the primary frequency modulation success rate and frequency modulation energy consumption residuals.

[0124] Step S405: Determine the final dynamic frequency modulation control model based on the calculation results of the first frequency modulation success rate and frequency modulation energy consumption residual.

[0125] For example, the objective function of the dynamic frequency modulation control model is defined as:

[0126]

[0127] Where f(·) represents the objective function of the dynamic frequency modulation control model, P sf s Let be the input representations of the gas turbine load set and the grid frequency set, respectively, min(y s ) represents the success rate of a single frequency modulation (FM) y. s Minimization is a constraint condition for Q. s Indicates frequency modulation energy consumption, C s T s u s These are the specific heat capacity influence coefficient, temperature, and fuel flow rate, respectively. R and L represent the gas constant and equipment load torque, respectively, V s Let υ represent fuel velocity, η represent reduced velocity, and η represent gas turbine efficiency.

[0128] Another embodiment of the present invention relates to a dynamic frequency regulation control system for a gas turbine power generation device, such as... Figure 8 As shown, it includes a parameter acquisition module 100, a model construction module 200, and a dynamic frequency tuning module 300.

[0129] The parameter acquisition module 100 is used to acquire time-series operating parameters of the gas turbine power generation equipment and grid frequency parameters, and to normalize the operating parameters of the gas turbine power generation equipment and grid frequency parameters to obtain a normalized frequency regulation dataset.

[0130] The model building module 200 is used to construct a gas turbine dynamic model and a power grid dynamic load model as islands based on the normalized frequency regulation dataset; execute the gas turbine dynamic model and the power grid dynamic load model, and obtain the gas turbine load set and the power grid frequency set based on the gas turbine dynamic model and the power grid dynamic load model, respectively.

[0131] The dynamic frequency regulation module 300 is used to establish a dynamic frequency regulation control model that considers the success rate of primary frequency regulation and frequency regulation energy consumption based on the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by the change of grid frequency, according to the gas turbine load set and grid frequency set. The gas turbine load set and grid frequency set are used as inputs to the dynamic frequency regulation control model, and the dynamic frequency regulation control model is executed to perform dynamic frequency regulation control on the gas turbine power generation equipment.

[0132] It should be noted that the parameter acquisition module 100, the model building module 200, and the dynamic frequency tuning module 300 can be connected via Bluetooth or a local area network, and the parameter acquisition module 100, the model building module 200, and the dynamic frequency tuning module 300 can be devices that can communicate, such as laptops, personal digital assistants (PDAs), and mobile phones.

[0133] For example, such as Figure 8As shown, the parameter acquisition module 100 specifically includes an outlier processing unit 110, a normalization processing unit 120, and a normalization unit 130.

[0134] The outlier processing unit 110 is used to load the operating parameters of the gas turbine power generation equipment and the grid frequency parameters in real time, perform outlier processing on the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, and delete outliers in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters.

[0135] The normalization processing unit 120 is used to standardize the operating parameters of the gas turbine power generation equipment and the grid frequency parameters after removing outliers, and to normalize the standardized operating parameters of the gas turbine power generation equipment and the grid frequency parameters, outputting a normalized set; wherein, the normalization processing method is Box-Cox transformation.

[0136] Normalization unit 130 is used to load the normalized set and process the normalized set using max-min normalization; after obtaining the normalized normalized set, white noise is introduced to perform iterative filtering on the normalized normalized set to obtain the normalized frequency modulation dataset.

[0137] The specific implementation method of the dynamic frequency regulation control system for gas turbine power generation equipment provided in the embodiments of the present invention can be found in the description of the dynamic frequency regulation control method for gas turbine power generation equipment provided in the embodiments of the present invention, and will not be repeated here.

[0138] The dynamic frequency regulation control system for gas turbine power generation equipment provided in this invention overcomes the problem that existing methods cannot accurately respond to changes in grid frequency by combining primary frequency regulation success rate and frequency regulation energy consumption, thus failing to accurately respond to primary frequency regulation under complex conditions and reducing the primary frequency regulation success rate. By establishing a dynamic frequency regulation control model that considers primary frequency regulation success rate and frequency regulation energy consumption, the primary frequency regulation success rate of the dynamic frequency regulation control model can be improved by combining gas turbine load and grid frequency, and frequency regulation energy consumption can be reduced. This allows for more precise control of the gas turbine output power, more accurate response to changes in grid frequency, improved primary frequency regulation success rate, and faster response to grid frequency changes, thus helping to maintain grid stability.

[0139] Another embodiment of the present invention relates to an electronic device comprising at least one processor and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed, enables the at least one processor to perform the dynamic frequency regulation control method for a gas turbine power generation device described in the above embodiments.

[0140] Memory, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the dynamic frequency regulation control method for gas turbine power generation equipment provided in the above embodiments of the present invention. Memory may include a program storage area and a data storage area. The program storage area may store the operating system and application programs required for at least one function. The data storage area may store data created by using the dynamic frequency regulation control method for gas turbine power generation equipment. Furthermore, memory may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the local module via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0141] Another embodiment of the present invention relates to a computer-readable storage medium storing a computer program that, when executed by a processor, implements the dynamic frequency regulation control method for a gas turbine power generation device as described in the above embodiments.

[0142] Computer-readable storage media (e.g., memory) can be volatile or non-volatile, or may include both. By way of example, and not limitation, non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which can act as external cache memory. By way of example, and not limitation, RAM can be obtained in various forms, such as synchronous RAM (DRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The storage devices disclosed herein are intended to include, but are not limited to, these and other suitable types of memory.

[0143] The various exemplary logic blocks, modules, and circuits described herein can be implemented or performed using the following components designed to perform the functions herein: general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination of these components. A general-purpose processor may be a microprocessor, but alternatively, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors combined with a DSP, and / or any other such configuration.

[0144] Another embodiment of the present invention relates to a computer program product, including a computer program that, when executed by a processor, implements the dynamic frequency regulation control method for gas turbine power generation equipment described in the above embodiments.

[0145] Those skilled in the art will understand that the above embodiments are specific implementations of the present invention, and in practical applications, various changes can be made in form and detail without departing from the spirit and scope of the present invention.

Claims

1. A dynamic frequency regulation control method for a gas turbine power generation device, characterized in that, The dynamic frequency modulation control method includes: The operating parameters of the gas turbine power generation equipment and the grid frequency parameters are obtained in a time sequence. The operating parameters of the gas turbine power generation equipment and the grid frequency parameters are normalized to obtain a normalized frequency regulation dataset. Based on the normalized frequency regulation dataset, a dynamic model of the gas turbine and a dynamic load model of the power grid, which are respectively isolated, are constructed. Execute the gas turbine dynamic model and the power grid dynamic load model, and obtain the gas turbine load set and the power grid frequency set based on the gas turbine dynamic model and the power grid dynamic load model, respectively; Based on the gas turbine load set and the power grid frequency set, and considering the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by power grid frequency changes, a dynamic frequency regulation control model is established that takes into account the success rate of primary frequency regulation and frequency regulation energy consumption. The gas turbine load set and the power grid frequency set are used as inputs to the dynamic frequency regulation control model. The dynamic frequency regulation control model is executed, and the gas turbine power generation equipment is dynamically regulated using the dynamic frequency regulation control model.

2. The dynamic frequency modulation control method according to claim 1, characterized in that, The normalization process of the operating parameters of the gas turbine power generation equipment and the grid frequency parameters to obtain a normalized frequency regulation dataset includes: The operating parameters of the gas turbine power generation equipment and the grid frequency parameters are loaded in real time. Outlier point processing is performed on the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, and abnormal values ​​in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters are deleted. The operating parameters of the gas turbine power generation equipment and the grid frequency parameters after removing outliers are standardized. The standardized operating parameters of the gas turbine power generation equipment and the grid frequency parameters are then normalized, and a normalized set is output. The normalization method is Box-Cox transformation. Load the normalized set and process it using max-min normalization, wherein the max-min normalization formula is: Where, x z X represents the normally distributed data output after max-min normalization, and X represents the normally distributed data in the input normalization set. max X min These are the maximum and minimum values ​​of the normally distributed data in the normalized set, respectively. Obtain the normalized set after normalization, introduce white noise to iteratively filter the normalized set, and obtain the normalized frequency modulation dataset.

3. The dynamic frequency modulation control method according to claim 2, characterized in that, The normalized set after normalization is iteratively filtered by introducing white noise to obtain the normalized frequency-modulated dataset, including: The normalized set is loaded, and white noise with amplitude ε is added to the normalized set. Integrated empirical mode decomposition is then performed on the normalized set with added white noise, and the decomposed components are output. The decomposed components are represented as follows: Where EEMD(IMF,t) represents the decomposed component, t represents the sampling time, a represents the number of iterations of white noise ε, and δ(x) represents the decomposed component. z ) represents the output quantity x of the normalized set. z The Dirac function, where e represents the natural constant; Define a noise tolerance and an iteration counter. Calculate the frequency domain modes of the decomposed components based on the noise tolerance and the iteration counter. Obtain the mean value of the frequency domain modes as the decomposed mean. The decomposition average is expressed as: Wherein, EEMD(IMF,t) τ U represents the decomposition mean. b (x z ) represents the frequency domain mode of the decomposed component at the b-th iteration, u b+1 (x z ) represents the frequency domain mode of the decomposed component at the (b+1)th iteration, b represents the iteration number recorded by the iteration counter, and χ represents the noise tolerance; Perform discrete wavelet transform on the decomposed mean to output the normalized frequency modulated dataset.

4. The dynamic frequency modulation control method according to claim 1, characterized in that, The construction of the gas turbine dynamic model and the power grid dynamic load model as islands based on the normalized frequency regulation dataset includes: Load the normalized frequency regulation dataset and identify the gas turbine type data, start-stop / variable load data, fuel data, and environmental data associated with the gas turbine in the normalized frequency regulation dataset; Using gas turbine type data, start-stop / variable load data, fuel data, and environmental data as constraints, a visual component model of the gas turbine is established. The visual component model of the gas turbine includes a regulating valve module, a regenerative cycle module, a compressor module, a combustion chamber module, a turbine module, a regenerator module, and a moment of inertia model. The loads in the gas turbine visualization component model are divided into rigid loads and flexible loads. The rigid loads and flexible loads in the gas turbine visualization component model are described using a Markov process. The overall disturbance of the gas turbine visualization component model is calculated based on the weighted calculation of rigid load disturbances and flexible load disturbances. The level changes of the rigid load disturbances and flexible load disturbances both follow the probability matrix of the Markov process. Simulation software is used to perform simulation tests on the visualized component model of the gas turbine, and the dynamic model of the gas turbine is output.

5. The dynamic frequency modulation control method according to claim 1, characterized in that, The construction of the gas turbine dynamic model and the power grid dynamic load model as islands based on the normalized frequency regulation dataset includes: Load the normalized frequency regulation dataset and identify the active power, reactive power, power factor, frequency parameters, and response time associated with the power grid in the normalized frequency regulation dataset; The active power, reactive power, power factor, frequency parameters, and response time associated with the power grid in the normalized frequency regulation dataset are fused to obtain a modeling fusion set, which is then divided into a training set and a validation set. An initial model for a dynamic load model of a power grid based on a combination of long short-term memory neural networks and convolutional neural networks is established, and the loss function, iteration rounds, and dynamic hyperparameters of the initial model are defined. The initial model of the power grid dynamic load model is trained based on the training set, the validation set, the loss function, the iteration rounds, and the dynamic hyperparameters to obtain the power grid dynamic load model.

6. The dynamic frequency modulation control method according to claim 5, characterized in that, The process of training the initial model of the power grid dynamic load model based on the training set, the validation set, the loss function, the number of iterations, and the dynamic hyperparameters to obtain the power grid dynamic load model includes: Using the training set as input, the initial model of the power grid dynamic load model is executed. Based on the loss function, the number of iterations, and the dynamic hyperparameters, the initial model of the power grid dynamic load model is iteratively trained to obtain a converged power grid dynamic load model. Using the validation set as input, the converged power grid dynamic load model is executed, and the validation result is output. The validity of the validation result is determined based on the model accuracy threshold. If the validation result meets the model accuracy threshold, the converged power grid dynamic load model is output.

7. The dynamic frequency modulation control method according to claim 1, characterized in that, The step of establishing a dynamic frequency regulation control model based on the gas turbine load set and the power grid frequency set, and considering the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by power grid frequency changes, and taking into account the primary frequency regulation success rate and frequency regulation energy consumption, includes: The gas turbine load set and the power grid frequency set are obtained, and the gas turbine load set and the power grid frequency set are standardized. The standardized gas turbine load set and the power grid frequency set are processed based on principal component analysis to extract the feature factors of the gas turbine load set and the power grid frequency set. The feature factors are loaded, and the dynamic frequency modulation control model is constructed based on the partial least squares method of the neural network combined with the feature factors. The internal relationship of the dynamic frequency modulation control model is solved by the neural network. The dynamic frequency modulation control model is loaded, and the input-output mapping relationship of the feature factors is solved using a three-layer BP network. The number of neural network nodes of the dynamic frequency modulation control model is determined based on the cross-validation method, and the converged dynamic frequency modulation control model is output. The first frequency modulation success rate and the frequency modulation energy consumption are introduced into the dynamic frequency modulation control model as constraints, and the first frequency modulation success rate and the frequency modulation energy consumption residual are calculated. The final dynamic frequency modulation control model is determined based on the calculation results of the first frequency modulation success rate and the frequency modulation energy consumption residual.

8. The dynamic frequency modulation control method according to claim 7, characterized in that, The objective function of the dynamic frequency modulation control model is defined as: Where f(·) represents the objective function of the dynamic frequency modulation control model, P s f s Let be the input representations of the gas turbine load set and the power grid frequency set, respectively, min(y s ) represents the success rate of a single frequency modulation (FM) adjustment. s Minimization is a constraint condition for Q. s Indicates frequency modulation energy consumption, C s T s u s These are the specific heat capacity influence coefficient, temperature, and fuel flow rate, respectively. R and L represent the gas constant and equipment load torque, respectively, V s Let υ represent fuel velocity, η represent reduced velocity, and η represent gas turbine efficiency.

9. A dynamic frequency regulation control system for a gas turbine power generation equipment, characterized in that, The dynamic frequency modulation control system includes: The parameter acquisition module is used to acquire time-series operating parameters of the gas turbine power generation equipment and grid frequency parameters, and to normalize the operating parameters of the gas turbine power generation equipment and the grid frequency parameters to obtain a normalized frequency regulation dataset. The model building module is used to construct a gas turbine dynamic model and a power grid dynamic load model as islands based on the normalized frequency regulation dataset; execute the gas turbine dynamic model and the power grid dynamic load model, and obtain the gas turbine load set and the power grid frequency set based on the gas turbine dynamic model and the power grid dynamic load model, respectively; The dynamic frequency regulation module is used to establish a dynamic frequency regulation control model that considers the success rate of primary frequency regulation and frequency regulation energy consumption based on the frequency regulation dynamic characteristics of the gas turbine dynamic model triggered by changes in the grid frequency, according to the gas turbine load set and the grid frequency set; the dynamic frequency regulation control model is executed by using the gas turbine load set and the grid frequency set as inputs to the dynamic frequency regulation control model, and the dynamic frequency regulation control model is used to perform dynamic frequency regulation control on the gas turbine power generation equipment.

10. The dynamic frequency modulation control system according to claim 9, characterized in that, The parameter acquisition module specifically includes: An outlier processing unit is used to load the operating parameters of the gas turbine power generation equipment and the grid frequency parameters in real time, perform outlier processing on the operating parameters of the gas turbine power generation equipment and the grid frequency parameters, and delete outliers in the operating parameters of the gas turbine power generation equipment and the grid frequency parameters. The normalization processing unit is used to standardize the operating parameters of the gas turbine power generation equipment and the grid frequency parameters after removing outliers, and to normalize the standardized operating parameters of the gas turbine power generation equipment and the grid frequency parameters, outputting a normalized set; wherein, the normalization processing method is Box-Cox transform; A normalization unit is used to load the normalized set, process the normalized set using max-min normalization, obtain the normalized set, introduce white noise to iteratively filter the normalized set, and obtain the normalized frequency modulation dataset.

Images