Power grid planning method and system based on load prediction and energy storage cooperation

Abstract

The invention discloses a power grid planning method and system based on load prediction and energy storage cooperation, and relates to the technical field of load prediction, and the method comprises the following steps: obtaining load data which comprises load change and load time sequence data; dynamically mapping the load change into a pseudo seismic wave amplitude, and converting the pseudo seismic wave amplitude into a pseudo seismic wave signal matched with the time-space characteristic of the load time sequence data based on a Morlet wavelet basis function; predicting a future seismic wave signal according to the pseudo seismic wave signal, and converting the future seismic wave signal into a predicted load value based on a dynamic integral reduction algorithm; and verifying the predicted load value and performing power grid planning. According to the method, load change is converted into a seismic wave form through mathematical mapping, and high-sensitivity load prediction is realized by multiplexing a seismic wave analysis technology.

Description

Technical Field

[0001] The present invention relates to the technical field of load forecasting, and more specifically, to a power grid planning method and system based on the coordination of load forecasting and energy storage. Background Art

[0002] Load forecasting, as a core technology for power system planning and operation, has evolved from traditional statistical methods to artificial intelligence-driven approaches: In the early days, it relied on time series models (such as ARIMA, exponential smoothing) and regression analysis to establish linear correlations between historical data and external factors such as temperature and holidays; with the expansion of data scale, machine learning (such as SVM, random forest) and neural networks (ANN) have been gradually applied, enhancing the ability to capture non-linear relationships; in recent years, deep learning models (such as LSTM, Transformer), with their advantages in modeling complex time series features and global dependencies, have become the mainstream technology, and hybrid models combining decomposition techniques, physical laws, and multi-source data (meteorological, economic indicators) have emerged to address challenges such as renewable energy fluctuations and emergencies; future trends focus on frontier technologies such as federated learning and graph neural networks to strengthen multi-region collaborative forecasting and power grid topology modeling, promoting accurate and intelligent load forecasting for high-proportion new energy power systems.

[0003] For example, the day-ahead load forecasting method, device, and load forecasting system disclosed in the invention patent announcement with the publication number CN117895480A include obtaining first predicted meteorological data corresponding to a target day and first historical load data within a historical time period before the target day; inputting the first predicted meteorological data and the first historical load data into a target prediction model to obtain the target day-ahead load output by the target prediction model; wherein, the target prediction model is obtained by fusing a time series model and an artificial intelligence model using the Stacking mechanism. The day-ahead load forecasting method of this application can integrate the advantages of two types of models, achieve deep cross of multiple models, and improve the accuracy and precision of day-ahead load forecasting by using the Stacking strategy to fuse a time series model and an artificial intelligence model.

[0004] For example, the short-term electricity load forecasting method, device and load forecasting system announced in the invention patent announcement with the announcement number of: CN117895477A, including obtaining the first predicted meteorological data corresponding to the target area at the moment to be measured and the first historical load data within the historical time period before the moment to be measured; processing the first historical load data based on at least two time scales to obtain at least two categories of second historical load data; each category of the second historical load data includes load data corresponding to multiple unit times under the time scale corresponding to the second historical load data; based on the at least two categories of second historical load data and the first predicted meteorological data, predicting the target load data corresponding to the target area at the moment to be measured. The short-term electricity load forecasting method of the present application can improve the accuracy and authenticity of short-term load forecasting.

[0005] In the above disclosed technical solution, there are at least the following technical problems: existing load forecasting methods are difficult to capture the non-linear characteristics and sudden fluctuations of load data. Seismic wave signal analysis technology has mature applications in waveform pattern recognition and mutation detection, but its association with load has not been explored.

[0006] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0007] In order to overcome the above defects of the prior art, the embodiments of the present invention provide a power grid planning method and system based on the coordination of load forecasting and energy storage. Through mathematical mapping, the load change is transformed into the form of seismic waves, and the load is predicted through seismic wave signal analysis technology to solve the problem that existing load forecasting methods are difficult to capture the non-linear characteristics and sudden fluctuations of load data.

[0008] To achieve the above object, the present invention provides the following technical solutions: A power grid planning method based on the coordination of load forecasting and energy storage, including the following steps: obtaining load data, where the load data includes load changes and load time series data; dynamically mapping the load changes into pseudo-seismic wave amplitudes, and converting the pseudo-seismic wave amplitudes into pseudo-seismic wave signals that match the spatio-temporal characteristics of the load time series data based on the Morlet wavelet basis function; predicting future seismic wave signals according to the pseudo-seismic wave signals, and converting the future seismic wave signals into predicted load values based on the dynamic integral reduction algorithm; verifying the predicted load values and performing power grid planning.

[0009] In a preferred embodiment, the obtaining of the load data is specifically: obtaining load time series data based on a preset data source and a preset sampling frequency and performing data processing; approximately calculating the load change rate at each time point of the processed load time series data based on the difference method as the load change, and the load change rate is obtained by approximately differentiating the time based on discrete sampling.

[0010] In a preferred embodiment, the dynamic mapping of the load change to the pseudo-seismic wave amplitude is specifically as follows: establish a mapping relationship equation between the pseudo-seismic wave amplitude and the load change rate; based on the mapping relationship equation, dynamically map the load change to the pseudo-seismic wave amplitude.

[0011] In a preferred embodiment, the conversion of the pseudo-seismic wave amplitude to a pseudo-seismic wave signal matching the spatio-temporal characteristics of the load time series data based on the Morlet wavelet basis function is specifically as follows: map a preset load period to the pseudo-seismic wave energy attenuation range, and simulate a Gaussian attenuation term based on the pseudo-seismic wave energy attenuation range; establish a mapping relationship equation between the pseudo-seismic wave vibration frequency and the load change, and simulate an oscillation term based on the load change; construct a Morlet wavelet basis function based on the simulated oscillation term and the Gaussian attenuation term, and superimpose the Morlet wavelet basis functions to generate a continuous pseudo-seismic wave signal; compare the time-frequency distribution of the pseudo-seismic wave signal with the load change to obtain a pseudo-seismic wave signal matching the spatio-temporal characteristics of the load time series data.

[0012] In a preferred embodiment, the acquisition of the pseudo-seismic wave signal matching the spatio-temporal characteristics of the load time series data is specifically obtained by adjusting the mapping relationship equation and the Gaussian attenuation term to ensure that the load peak corresponds to the high-amplitude pulse of the pseudo-seismic wave and to ensure that the main frequency falls within a preset frequency band.

[0013] In a preferred embodiment, the prediction of the future seismic wave signal based on the pseudo-seismic wave signal is specifically as follows: obtain an LSTM-CNN hybrid network model and train the time-frequency feature extraction ability of the hybrid network model to obtain a pseudo-seismic wave prediction model; use the pseudo-seismic wave signal as the input signal of the CNN layer of the pseudo-seismic wave prediction model to extract the waveform spatial features and output a waveform spatial feature sequence; use the waveform spatial feature sequence as the input signal of the LSTM layer of the pseudo-seismic wave prediction model to output the temporal dependence feature sequence of the pseudo-seismic wave; input the temporal dependence feature sequence into the fully connected layer to output the future seismic wave signal.

[0014] In a preferred embodiment, the conversion of the future seismic wave signal to a predicted load value based on the dynamic integral reduction algorithm is specifically as follows: based on the dynamic integral reduction algorithm, combine the future seismic wave signal with the initial load value and gradually calculate the future load value, where the load value is the magnitude of the load time series data at any moment; adjust the algorithm parameters in real time according to the future load value.

[0015] In a preferred embodiment, the verification of the predicted load value and the power grid planning are specifically as follows: calculate the absolute error between the predicted load value and the actual value based on the error function MAE of the predicted load value and the actual value; construct a combined loss function of SSIM+MAE based on SSIM and MAE, and optimize the pseudo-seismic wave prediction model according to the combined loss function.

[0016] A power grid planning system based on the coordination of load forecasting and energy storage, comprising: a pseudo-seismic wave generation module, configured to dynamically map load changes to pseudo-seismic wave amplitudes, and convert the pseudo-seismic wave amplitudes into pseudo-seismic wave signals that match the spatio-temporal characteristics of load time series data based on the Morlet wavelet basis function; a pseudo-seismic wave prediction module, configured to predict future seismic wave signals according to the pseudo-seismic wave signals; and a load forecasting verification module, configured to verify the predicted load values and perform power grid planning. The technical effects and advantages of a power grid planning method and system based on the coordination of load forecasting and energy storage according to the present invention: 1. The present invention converts the time series fluctuations of load data into multi-scale time-frequency characteristics of seismic waves, and through technologies such as wavelet synthesis and spectrum analysis, explicitly extracts the transient mutations and periodic laws of load changes, solving the problem of insufficient capture of complex time series patterns by traditional methods.

[0017] 2. The present invention significantly reduces the development cost, improves the prediction accuracy and robustness by directly migrating mature signal processing algorithms and pre-trained models in the seismic field, and utilizing their noise resistance, time-frequency analysis ability and physical interpretability. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic flow chart of a power grid planning method based on the coordination of load forecasting and energy storage according to the present invention.

[0019] Figure 2 It is a schematic structural diagram of a power grid planning system based on the coordination of load forecasting and energy storage according to the present invention.

[0020] Figure 3 It is the original load curve.

[0021] Figure 4 It is the generated pseudo-seismic wave signal.

[0022] Figure 5 It is the spectrum comparison between the pseudo-seismic wave signal and the load change. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0024] Embodiment 1 Figure 1 A power grid planning method based on the coordination of load forecasting and energy storage according to the present invention is given, including the following steps: S1. Obtain load data, where the load data includes load variations and load time-series data.

[0025] In this embodiment, obtaining load data, where the load data includes load variations and load time-series data, and generating standardized preprocessed data is specifically as follows: Obtain load time-series data based on a preset data source and a preset sampling frequency; Eliminate outliers in the load time-series data, smooth high-frequency noise in the load time-series data, and perform normalization; Approximately calculate the load change rate at each time point of the load time-series data as the load variation based on the difference method, where the load change rate is obtained by approximately differentiating the time based on discrete sampling.

[0026] The normalization formula is:

[0027] where, is the original data point; is the minimum value in the data; is the maximum value in the data; is the data point after normalization.

[0028] The specific load instantaneous change rate is:

[0029] where, is the load value at time; is the load instantaneous change rate; is the sampling interval.

[0030] It should be noted that due to the characteristics of time-series data, the time order must be retained during processing, otherwise the time dependence of the data will be destroyed.

[0031] S2. Dynamically map the load variation to the amplitude of a pseudo-seismic wave, and convert the amplitude of the pseudo-seismic wave into a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time-series data based on the Morlet wavelet basis function.

[0032] In this embodiment, dynamically mapping the load variation to the amplitude of a pseudo-seismic wave is specifically as follows: Establish a mapping relationship equation between the amplitude of the pseudo-seismic wave and the load variation; Dynamically map the load variation to the amplitude of the pseudo-seismic wave based on the mapping relationship equation.

[0033] The mapping relationship equation of the amplitude of the pseudo-seismic wave is:

[0034] where, is the amplitude of the pseudo-seismic wave; is the proportionality coefficient, which is used to adjust the amplitude dimension to match the physical meaning of the seismic wave.

[0035] It should be noted that when adjusting the proportionality coefficient it is necessary to retain both the balanced noise suppression and the mutation characteristics simultaneously.

[0036] In this embodiment, based on the Morlet wavelet basis function, the amplitude of the pseudo-seismic wave is converted into a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time-series data. Specifically: Map the preset load period to the pseudo-seismic wave energy attenuation range, and simulate the Gaussian attenuation term based on the pseudo-seismic wave energy attenuation range; Establish a mapping relationship equation between the pseudo-seismic wave vibration frequency and the load change, and simulate the oscillation term based on the load change; Construct a Morlet wavelet basis function based on the simulated oscillation term and the Gaussian attenuation term, and superimpose the Morlet wavelet basis function to generate a continuous pseudo-seismic wave signal; Compare the time-frequency distribution of the pseudo-seismic wave and the load change, and adjust the mapping relationship equation and the Gaussian attenuation term to ensure that the load spike corresponds to the high-amplitude pulse of the pseudo-seismic wave and ensure that the main frequency falls within the preset frequency band, so as to obtain a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time-series data.

[0037] The specific mathematical expression of the Morlet wavelet basis is as follows:

[0038] The mapping relationship equation between the pseudo-seismic wave vibration frequency and the load change is:

[0039] The pseudo-seismic wave synthesis algorithm is as follows:

[0040] where, is the Gaussian attenuation term; is the Morlet wavelet function; is the parameter that controls the time diffusion range of the waveform, which is aligned with the load change period; is the oscillation term; is the pseudo-seismic wave frequency; is the parameter that controls the dynamic range of the frequency; is the pseudo-seismic wave signal; is the time component; is the absolute value of the instantaneous load change rate.

[0041] Figure 4Based on the Morlet wavelet basis function, the amplitude of the pseudo-seismic wave is converted into a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time series data. The relationship curve between the amplitude of the pseudo-seismic wave signal and time is shown.

[0042] Figure 5 For the spectral comparison between the pseudo-seismic wave signal and the load change, adjust the mapping relationship equation and the Gaussian attenuation term to ensure that the load peak corresponds to the high-amplitude pulse of the pseudo-seismic wave and ensure that the main frequency falls within the preset frequency band, so as to obtain a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time series data.

[0043] It should be noted that the load change affects both the amplitude and the frequency, which is a reasonable simplification of the relationship between the energy release intensity and the vibration rhythm in real seismic waves. The amplitude is jointly determined by the sign and magnitude of [amplitude factor] (the positive and negative correspond to the polarity of the waveform, and the magnitude corresponds to the amplitude intensity); the frequency is only determined by the absolute value of [frequency factor] and has nothing to do with the sign.

[0044] S3. Predict the future seismic wave signal according to the pseudo-seismic wave signal, and convert the future seismic wave signal into a predicted load value based on the dynamic integral reduction algorithm.

[0045] In this embodiment, predicting the future seismic wave signal according to the pseudo-seismic wave signal is specifically as follows: Based on the public earthquake data set, simulate the seismic waveforms of different magnitudes and focal depths, and train the time-frequency feature extraction ability of the standard LSTM-CNN hybrid network model to obtain a pseudo-seismic wave prediction model; Input the pseudo-seismic wave signal into the CNN layer of the pseudo-seismic wave prediction model, extract the waveform spatial features and output a waveform spatial feature sequence, and the waveform spatial feature sequence includes the pulse peak value, waveform slope, and continuous fluctuation of the pseudo-seismic wave; Input the waveform spatial feature sequence output by the CNN layer into the LSTM layer of the pseudo-seismic wave prediction model, and output a time series dependence feature sequence of the pseudo-seismic wave, and the time series dependence feature sequence is the concatenation term of the spatial feature sequence in the time dimension; Input the time series dependence feature sequence output by the LSTM layer into the fully connected layer, and output the future seismic wave signal.

[0046] It should be noted that in the training stage, a public seismic wave data set (such as USGS earthquake records) is used to simulate waveforms of different magnitudes and focal depths, and the model is trained to predict the subsequent propagation waveform of the seismic wave (input the historical seismic wave and output the future waveform), so that the CNN-LSTM masters the basic time-frequency feature extraction ability of the seismic wave.

[0047] In this embodiment, and converting the future seismic wave signal into a predicted load value based on the dynamic integral reduction algorithm is specifically as follows: Based on the dynamic integral reduction algorithm, the future seismic wave signal is combined with the initial load value to gradually calculate the future load value, where the load value is the magnitude of the load time series data at any moment. The algorithm parameters are adjusted in real time in combination with the future load value.

[0048] For the dynamic integral reduction algorithm, the specific calculation formula is as follows:

[0049] The algorithm parameters are adjusted by minimizing the prediction error, and the specific calculation formula is as follows:

[0050] Where, is the actual load value at the starting moment of prediction; is the scaling parameter before adjustment; is the scaling parameter after adjustment; is the pseudo-seismic wave prediction waveform; is the actual load value; is the load value gradually integrated and restored.

[0051] It should be noted that the scaling influence of the scaling parameter is eliminated, and segmented integration and calibration of the initial value are performed to prevent error accumulation. If the integration result exceeds the historical load range, it is replaced by the average value of the previous time period.

[0052] S4. Verify the predicted load value and perform power grid planning.

[0053] In this embodiment, verifying the predicted load value and performing power grid planning specifically include: Based on the error function MAE between the predicted load value and the actual value, calculate the absolute error between the predicted load value and the actual value; Based on SSIM and MAE, construct a combined loss function SSIM + MAE, and optimize the pseudo-seismic wave prediction model according to the combined loss function.

[0054] For the waveform structural similarity function SSIM, the specific calculation formula is as follows:

[0055] For the numerical error function MAE, the specific calculation formula is as follows:

[0056] For the combined loss function SSIM + MAE, the specific calculation formula is as follows:

[0057] For the gradient of the combined loss function, the specific calculation formula is as follows:

[0058] wherein, is the mean value; is the variance; is the mean value of the predicted load value; is the mean value of the actual load value; is the predicted load variance; is the actual load variance; is the covariance; is the stability constant; is the weight coefficient and ; is the predicted waveform of the pseudo-seismic wave; is the actual waveform of the pseudo-seismic wave; is the model parameter; is the model predicted value; is the true value; is the gradient of the joint loss function.

[0059] Example 2, Figure 2 provides a power grid planning system based on load prediction and energy storage coordination according to the present invention, including: A data preprocessing module for obtaining load data, where the load data includes load changes and load time series data; A pseudo-seismic wave generation module for dynamically mapping load changes to pseudo-seismic wave amplitudes and converting the pseudo-seismic wave amplitudes into pseudo-seismic wave signals that match the spatio-temporal characteristics of the load time series data based on the Morlet wavelet basis function; A pseudo-seismic wave prediction module for predicting future seismic wave signals according to the pseudo-seismic wave signals and converting the future seismic wave signals into predicted load values based on the dynamic integral reduction algorithm; A load prediction verification module for verifying the predicted load values and performing power grid planning.

[0060] The above formulas are all dimensionless and take their numerical calculations. The formulas are obtained by collecting a large amount of data for software simulation to get a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0061] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0062] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0063] In addition, the functional modules in each embodiment of this application can be integrated into a processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0064] As described above, it is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0065] Finally: The above description is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A power grid planning method based on the coordination of load forecasting and energy storage, characterized in that It includes the following steps: Obtain load data, where the load data includes load variation and load time-series data; Dynamically map the load variation to pseudo-seismic wave amplitude, and convert the pseudo-seismic wave amplitude into a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time-series data based on the Morlet wavelet basis function; Predict the future seismic wave signal according to the pseudo-seismic wave signal, and convert the future seismic wave signal into a predicted load value based on the dynamic integral reduction algorithm; Verify the predicted load value and conduct power grid planning.

2. The power grid planning method based on load prediction and energy storage collaboration according to claim 1, wherein The obtaining of the load data is specifically as follows: Obtain the load time-series data based on the preset data source and preset sampling frequency and conduct data processing; Approximately calculate the load variation rate at each time point of the processed load time-series data based on the difference method as the load variation, and the load variation rate is obtained by approximately deriving the time based on discrete sampling.

3. The power grid planning method based on load prediction and energy storage collaboration according to claim 2, wherein The dynamically mapping the load variation to pseudo-seismic wave amplitude is specifically as follows: Establish a mapping relationship equation between the pseudo-seismic wave amplitude and the load variation rate; Dynamically map the load variation to pseudo-seismic wave amplitude based on the mapping relationship equation.

4. The power grid planning method based on load forecasting and energy storage coordination according to claim 3, wherein The converting the pseudo-seismic wave amplitude into a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time-series data based on the Morlet wavelet basis function is specifically as follows: Map the preset load period to the pseudo-seismic wave energy attenuation range, and simulate the Gaussian attenuation term based on the pseudo-seismic wave energy attenuation range; Establish a mapping relationship equation between the pseudo-seismic wave vibration frequency and the load variation, and simulate the oscillation term based on the load variation; Construct the Morlet wavelet basis function based on the simulated oscillation term and Gaussian attenuation term, and superimpose the Morlet wavelet basis functions to generate a continuous pseudo-seismic wave signal; Compare the time-frequency distribution of the pseudo-seismic wave signal with the load variation to obtain a pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time-series data.

5. The grid planning method based on load forecasting and energy storage collaboration according to claim 4, characterized in that The obtaining of the pseudo-seismic wave signal that matches the spatio-temporal characteristics of the load time-series data is specifically as follows: It is obtained by adjusting the mapping relationship equation and the Gaussian attenuation term to ensure that the load peak corresponds to the high-amplitude pulse of the pseudo-seismic wave and ensure that the main frequency falls within the preset frequency band.

6. The power grid planning method based on load prediction and energy storage coordination according to claim 5, wherein, The predicting the future seismic wave signal according to the pseudo-seismic wave signal is specifically as follows: Obtain the LSTM-CNN hybrid network model and train the time-frequency feature extraction ability of the hybrid network model to obtain a pseudo-seismic wave prediction model; Use the pseudo-seismic wave signal as the input signal of the CNN layer of the pseudo-seismic wave prediction model, extract the waveform spatial features and output the waveform spatial feature sequence; Use the waveform spatial feature sequence as the input signal of the LSTM layer of the pseudo-seismic wave prediction model, and output the time-series dependence feature sequence of the pseudo-seismic wave; Input the time-series dependence feature sequence into the fully connected layer and output the future seismic wave signal.

7. The power grid planning method based on load forecasting and energy storage collaboration according to claim 6, wherein The converting the future seismic wave signal into a predicted load value based on the dynamic integral reduction algorithm is specifically as follows: Based on the dynamic integral reduction algorithm, combine the future seismic wave signal with the initial load value and gradually calculate the future load value, where the load value is the magnitude of the load time-series data at any moment; Adjust the algorithm parameters in real time according to the future load value.

8. The power grid planning method based on load prediction and energy storage collaboration according to claim 7, wherein, The verifying the predicted load value and conducting power grid planning is specifically as follows: Calculate the absolute error between the predicted load value and the actual value based on the mean absolute error (MAE) of the error function between the predicted load value and the actual value; Construct a combined loss function of SSIM+MAE based on SSIM and MAE, and optimize the pseudo-seismic wave prediction model according to the combined loss function.

9. A system using the power grid planning method based on load prediction and energy storage collaboration as described in any one of claims 1-8, characterized in that, Including: A data preprocessing module for obtaining load data, where the load data includes load changes and load time series data; A pseudo-seismic wave generation module for dynamically mapping load changes to pseudo-seismic wave amplitudes, and converting the pseudo-seismic wave amplitudes into pseudo-seismic wave signals that match the spatio-temporal characteristics of the load time series data based on the Morlet wavelet basis function; A pseudo-seismic wave prediction module for predicting future seismic wave signals based on the pseudo-seismic wave signals, and converting the future seismic wave signals into predicted load values based on the dynamic integral reduction algorithm; A load prediction verification module for verifying the predicted load values and performing power grid planning.

Images