Method and device for determining reserve capacity of power system, electronic equipment and storage medium

Abstract

The invention discloses a method and device for determining the reserve capacity of an electric power system, electronic equipment and a storage medium, and belongs to the technical field of reserve capacity management.The method comprises the steps that empirical mode decomposition is conducted on a predicted net load curve and an actual net load curve, and a high-frequency component, an intermediate-frequency component and a low-frequency component of the predicted net load curve and the actual net load curve are generated; calculating and predicting a high-frequency error sequence, an intermediate-frequency error sequence and a low-frequency error sequence; carrying out probability distribution modeling to generate a high-frequency error probability density function, an intermediate-frequency error probability density function and a low-frequency error probability density function; and determining high-frequency reserve capacity, intermediate-frequency reserve capacity and low-frequency reserve capacity of the power system according to the high-frequency error probability density function, the intermediate-frequency error probability density function and the low-frequency error probability density function based on a preset confidence coefficient. By implementing the invention, the problem that the reserve capacity of the power system is not accurately determined in the prior art can be solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of spare capacity management, and particularly relates to a method and device for determining spare capacity of a power system, an electronic device and a storage medium. BACKGROUND

[0002] Spare capacity of a power system is a key resource for ensuring safe, stable and reliable operation of a power grid. In the process of dispatching management, accurate determination and optimized configuration of spare capacity requirements not only meet the rigid requirements of power supply reliability, but also are important links for improving the economic efficiency of the entire power system operation and realizing efficient use of resources, and are of great significance for ensuring the smooth and orderly production and life of society.

[0003] At present, the method for determining spare capacity in the prior art mainly relies on simple statistical processing of historical load data, such as using the mean value or interval statistical data in a fixed period as a spare capacity benchmark. This method is widely used in actual power grid dispatching due to its simplicity and ease of implementation, but it fails to fully capture the characteristics of multi-scale fluctuations of system net load, especially in the identification of high-frequency, medium-frequency and low-frequency load fluctuations, which reduces the accuracy of spare capacity determination and easily leads to insufficient spare capacity configuration or resource waste. SUMMARY

[0004] The embodiments of the present application provide a method and device for determining spare capacity of a power system, an electronic device and a storage medium, which can solve the problem of inaccurate determination of spare capacity of a power system in the prior art.

[0005] An embodiment of the present application provides a method for determining spare capacity of a power system, comprising:

[0006] obtaining a predicted net load curve and an actual net load curve of a typical day of the power system;

[0007] performing empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component;

[0008] performing empirical mode decomposition on the actual net load curve to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component;

[0009] calculating the difference between the predicted high-frequency component and the actual high-frequency component to generate a high-frequency error sequence, calculating the difference between the predicted medium-frequency component and the actual medium-frequency component to generate a medium-frequency error sequence, and calculating the difference between the predicted low-frequency component and the actual low-frequency component to generate a low-frequency error sequence;

[0010] modeling probability distribution of the high frequency error sequence, the medium frequency error sequence and the low frequency error sequence respectively to generate high frequency error probability density function, medium frequency error probability density function and low frequency error probability density function;

[0011] Based on the preset confidence, the high frequency reserve capacity, the medium frequency reserve capacity and the low frequency reserve capacity of the power system are determined according to the high frequency error probability density function, the medium frequency error probability density function and the low frequency error probability density function respectively.

[0012] Further, the predicted net load curve is empirically mode decomposed to generate predicted high frequency component, predicted medium frequency component and predicted low frequency component, including:

[0013] The predicted intrinsic mode function extraction operation is repeatedly performed until the current prediction residual curve meets the preset decomposition termination condition to generate a set of predicted intrinsic mode functions;

[0014] Each predicted intrinsic mode function in the set of predicted intrinsic mode functions is subjected to frequency analysis to generate predicted high frequency component, predicted medium frequency component and predicted low frequency component;

[0015] The predicted intrinsic mode function extraction operation includes:

[0016] Local maximum values and local minimum values in the current prediction curve are extracted to generate each current prediction local maximum value and each current prediction local minimum value; wherein the initial prediction curve is the predicted net load curve;

[0017] Based on the cubic spline interpolation method, the upper envelope function of the current prediction curve and the lower envelope function of the current prediction curve are constructed according to each current prediction local maximum value and each current prediction local minimum value;

[0018] The envelope mean function of the current prediction curve is calculated and generated according to the upper envelope function of the current prediction curve and the lower envelope function of the current prediction curve;

[0019] The predicted intrinsic mode function of the current prediction curve is determined according to the envelope mean function of the current prediction curve and the current prediction curve;

[0020] The current prediction residual curve is calculated and generated according to the predicted intrinsic mode function of the current prediction curve and the current prediction curve;

[0021] The predicted intrinsic mode function of the current prediction curve is added to the set of predicted intrinsic mode functions;

[0022] It is judged whether the current prediction residual curve meets the preset decomposition termination condition, if yes, the set of predicted intrinsic mode functions is output; if not, the current prediction residual curve is updated as the current prediction curve.

[0023] Further, the actual net load curve is subjected to empirical mode decomposition to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component, including:

[0024] The actual intrinsic mode function extraction operation is repeatedly performed until the current actual residual curve meets a preset decomposition termination condition, to generate an actual intrinsic mode function set;

[0025] Each actual intrinsic mode function in the actual intrinsic mode function set is subjected to frequency analysis to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component;

[0026] The intrinsic mode function extraction operation includes:

[0027] Local maximum values and local minimum values in the current actual curve are extracted to generate respective current actual local maximum values and respective current actual local minimum values; wherein the initial actual curve is the actual net load curve;

[0028] Based on the respective current actual local maximum values and the respective current actual local minimum values, an upper envelope function of the current actual curve and a lower envelope function of the current actual curve are constructed based on a cubic spline interpolation method;

[0029] The envelope mean function of the current actual curve is calculated based on the upper envelope function of the current actual curve and the lower envelope function of the current actual curve;

[0030] The actual intrinsic mode function of the current actual curve is determined based on the envelope mean function of the current actual curve and the current actual curve;

[0031] The current actual residual curve is calculated based on the actual intrinsic mode function of the current actual curve and the current actual curve;

[0032] The actual intrinsic mode function of the current actual curve is added to the actual intrinsic mode function set;

[0033] It is determined whether the current actual residual curve meets the preset decomposition termination condition, and if so, the actual intrinsic mode function set is output; if not, the current actual residual curve is updated as the current actual curve.

[0034] Further, the high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence are subjected to probability distribution modeling respectively to generate a high-frequency error probability density function, a medium-frequency error probability density function and a low-frequency error probability density function, including:

[0035] The high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence are sampled respectively to generate respective high-frequency error sample points, respective medium-frequency error sample points and respective low-frequency error sample points.

[0036] determine the number of high-frequency error sample points, the number of medium-frequency error sample points and the number of low-frequency error sample points according to each high-frequency error sample point, each medium-frequency error sample point and each low-frequency error sample point respectively;

[0037] determine the high-frequency bandwidth according to each high-frequency error sample point and the number of high-frequency error sample points, determine the medium-frequency bandwidth according to each medium-frequency error sample point and the number of medium-frequency error sample points, and determine the low-frequency bandwidth according to each low-frequency error sample point and the number of low-frequency error sample points;

[0038] generate a high-frequency error probability density function based on the high-frequency bandwidth by using a Gaussian kernel function, generate a medium-frequency error probability density function based on the medium-frequency bandwidth by using a Gaussian kernel function, and generate a low-frequency error probability density function based on the low-frequency bandwidth by using a Gaussian kernel function.

[0039] On the basis of the above method embodiment, the application provides a device embodiment.

[0040] An embodiment of the application provides a device for determining the reserve capacity of a power system, comprising a data acquisition module, an empirical mode decomposition module, an error sequence determination module, a probability modeling module and a reserve capacity determination module.

[0041] The data acquisition module is configured to acquire a predicted net load curve and an actual net load curve of a typical day of the power system.

[0042] The empirical mode decomposition module is configured to perform empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component, and perform empirical mode decomposition on the actual net load curve to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component.

[0043] The error sequence determination module is configured to calculate the difference between the predicted high-frequency component and the actual high-frequency component to generate a high-frequency error sequence, calculate the difference between the predicted medium-frequency component and the actual medium-frequency component to generate a medium-frequency error sequence, and calculate the difference between the predicted low-frequency component and the actual low-frequency component to generate a low-frequency error sequence.

[0044] The reserve capacity determination module is configured to determine the high-frequency reserve capacity, the medium-frequency reserve capacity and the low-frequency reserve capacity of the power system according to the high-frequency error probability density function, the medium-frequency error probability density function and the low-frequency error probability density function respectively based on a preset confidence level.

[0045] Further, the empirical mode decomposition module is configured to perform empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component, including:

[0046] The predicted intrinsic mode function extraction operation is repeatedly performed until the current predicted residual curve meets a preset decomposition termination condition, and a predicted intrinsic mode function set is generated.

[0047] The frequency analysis is performed on each predicted intrinsic mode function in the predicted intrinsic mode function set to generate the predicted high-frequency component, the predicted medium-frequency component and the predicted low-frequency component.

[0048] The predicted intrinsic mode function extraction operation includes:

[0049] The local maximum values and the local minimum values in the current predicted curve are extracted to generate respective current predicted local maximum values and respective current predicted local minimum values; wherein the initial predicted curve is the predicted net load curve.

[0050] The upper envelope function of the current predicted curve and the lower envelope function of the current predicted curve are constructed based on the cubic spline interpolation method according to the respective current predicted local maximum values and the respective current predicted local minimum values.

[0051] The envelope mean function of the current predicted curve is calculated and generated according to the upper envelope function of the current predicted curve and the lower envelope function of the current predicted curve.

[0052] The predicted intrinsic mode function of the current predicted curve is determined according to the envelope mean function of the current predicted curve and the current predicted curve.

[0053] The current predicted residual curve is calculated and generated according to the predicted intrinsic mode function of the current predicted curve and the current predicted curve.

[0054] The predicted intrinsic mode function of the current predicted curve is added to the predicted intrinsic mode function set.

[0055] It is determined whether the current predicted residual curve meets the preset decomposition termination condition, if yes, the predicted intrinsic mode function set is output, and if not, the current predicted residual curve is updated as the current predicted curve.

[0056] Further, the empirical mode decomposition module is configured to perform empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component, including:

[0057] The predicted intrinsic mode function extraction operation is repeatedly performed until the current predicted residual curve meets a preset decomposition termination condition, and a predicted intrinsic mode function set is generated.

[0058] For each actual eigenmode function in the actual eigenmode function set, frequency analysis is performed to generate actual high-frequency components, actual medium-frequency components and actual low-frequency components;

[0059] The eigenmode function extraction operation includes:

[0060] Local maximum values and local minimum values in the current actual curve are extracted to generate current actual local maximum values and current actual local minimum values; wherein, the initial actual curve is an actual net load curve;

[0061] Based on the current actual local maximum values and the current actual local minimum values, an upper envelope function of the current actual curve and a lower envelope function of the current actual curve are constructed based on a cubic spline interpolation method;

[0062] Based on the upper envelope function of the current actual curve and the lower envelope function of the current actual curve, an envelope mean function of the current actual curve is calculated and generated;

[0063] Based on the envelope mean function of the current actual curve and the current actual curve, actual eigenmode functions of the current actual curve are determined;

[0064] Based on the actual eigenmode functions of the current actual curve and the current actual curve, a current actual residual curve is calculated and generated;

[0065] The actual eigenmode functions of the current actual curve are added to the actual eigenmode function set;

[0066] It is judged whether the current actual residual curve meets a preset decomposition termination condition, if yes, the actual eigenmode function set is output; if not, the current actual residual curve is updated as the current actual curve.

[0067] Further, the probability modeling module is configured to model the probability distribution of the high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence respectively to generate a high-frequency error probability density function, a medium-frequency error probability density function and a low-frequency error probability density function, including:

[0068] The high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence are sampled respectively to generate high-frequency error sample points, medium-frequency error sample points and low-frequency error sample points;

[0069] The number of high-frequency error sample points, the number of medium-frequency error sample points and the number of low-frequency error sample points are determined respectively according to the high-frequency error sample points, the medium-frequency error sample points and the low-frequency error sample points;

[0070] The high-frequency bandwidth is determined according to each high-frequency error sample point and the number of high-frequency error sample points; the medium-frequency bandwidth is determined according to each medium-frequency error sample point and the number of medium-frequency error sample points; and the low-frequency bandwidth is determined according to each low-frequency error sample point and the number of low-frequency error sample points.

[0071] The high-frequency error probability density function is generated based on the high-frequency bandwidth by using a Gaussian kernel function; the medium-frequency error probability density function is generated based on the medium-frequency bandwidth by using a Gaussian kernel function; and the low-frequency error probability density function is generated based on the low-frequency bandwidth by using a Gaussian kernel function.

[0072] On the basis of the above-mentioned method embodiment, the application correspondingly provides an electronic device embodiment.

[0073] An embodiment of the application provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the computer program is executed by the processor, the determination method of the reserve capacity of the power system in any one of the above-mentioned method embodiments is realized.

[0074] On the basis of the above-mentioned method embodiment, the application correspondingly provides a storage medium embodiment.

[0075] An embodiment of the application provides a storage medium, which has a computer program stored thereon, and when the computer program is executed, the determination method of the reserve capacity of the power system in any one of the above-mentioned method embodiments is executed by a device in which the storage medium is located.

[0076] Compared with the prior art, the application has the following beneficial effects:

[0077] The embodiment of the application provides a determination method, device, electronic device and storage medium of the reserve capacity of a power system. The method obtains a predicted net load curve and an actual net load curve of a typical day of the power system, and decomposes the predicted net load curve and the actual net load curve into high, medium and low frequency fluctuation components by using an empirical mode decomposition technology, so that different time scale fluctuations are processed respectively. The error sequences of high, medium and low frequencies are obtained by calculating the difference between the prediction and the actual value of each frequency component. The three types of error sequences are independently modeled, and a probability density function is generated for each type of error. According to a preset confidence level, the required high-frequency reserve capacity, medium-frequency reserve capacity and low-frequency reserve capacity are calculated by using the respective probability density functions.

[0078] The present application adopts an empirical mode decomposition technique to decompose the net load curve into high, medium and low frequency components, directly solving the core defect that the traditional method cannot identify and distinguish multi-scale fluctuation characteristics. Furthermore, independent probability distribution modeling is performed on the prediction error of each frequency component, so that the simple statistical processing of the traditional method is replaced by fine probability analysis, so that the standby capacity matched with each fluctuation characteristic can be determined based on the confidence level, overcoming the problem of resource waste or insufficient reliability caused by rough evaluation of the traditional method, and improving the accuracy of the determination of the standby capacity of the power system. BRIEF DESCRIPTION OF DRAWINGS

[0079] Figure 1 is a flowchart of a determination method of a standby capacity of a power system provided by an embodiment of the present application.

[0080] Figure 2 is a structural diagram of a determination device of a standby capacity of a power system provided by an embodiment of the present application. DETAILED DESCRIPTION

[0081] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0082] As Figure 1 shown, to solve the problem of inaccurate determination of the standby capacity of the power system in the prior art, an embodiment of the present application provides a determination method of a standby capacity of a power system, at least including the following steps:

[0083] Step S1, obtaining a predicted net load curve and an actual net load curve of a typical day of a power system.

[0084] Specifically, for a typical day, the predicted and actual power consumption load data and the predicted and actual new energy output data recorded at a preset time interval for 24 hours are obtained.

[0085] The predicted net load curve is a time sequence obtained by subtracting the predicted new energy output at each time point from the corresponding predicted power consumption load. Similarly, the actual net load curve is a time sequence obtained by subtracting the actual new energy output at each time point from the corresponding actual power consumption load. By obtaining the two complete net load curves, complete and necessary data inputs are provided for subsequent accurate analysis of prediction errors and quantification of fluctuation uncertainty at different time scales.

[0086] Step S2, performing empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component.

[0087] In a preferred embodiment, performing empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component comprises:

[0088] Repeating the predicted intrinsic mode function extraction operation until the current predicted residual curve meets a preset decomposition termination condition to generate a predicted intrinsic mode function set;

[0089] Performing frequency analysis on each predicted intrinsic mode function in the predicted intrinsic mode function set to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component;

[0090] The predicted intrinsic mode function extraction operation comprises:

[0091] Extracting local maximum values and local minimum values in the current predicted curve to generate respective current predicted local maximum values and respective current predicted local minimum values; wherein the initial predicted curve is the predicted net load curve;

[0092] Constructing an upper envelope function of the current predicted curve and a lower envelope function of the current predicted curve based on a cubic spline interpolation method according to the respective current predicted local maximum values and the respective current predicted local minimum values;

[0093] Calculating a envelope mean function of the current predicted curve according to the upper envelope function of the current predicted curve and the lower envelope function of the current predicted curve;

[0094] Determining a predicted intrinsic mode function of the current predicted curve according to the envelope mean function of the current predicted curve and the current predicted curve;

[0095] Calculating a current predicted residual curve according to the predicted intrinsic mode function of the current predicted curve and the current predicted curve;

[0096] Adding the predicted intrinsic mode function of the current predicted curve to the predicted intrinsic mode function set;

[0097] Determining whether the current predicted residual curve meets a preset decomposition termination condition, and if so, outputting the predicted intrinsic mode function set, and if not, updating the current predicted residual curve to the current predicted curve.

[0098] For example, the predicted net load curve is processed using an empirical mode decomposition method, and the main steps include:

[0099] 1. Data point determination and extreme value detection: Determine the coordinate (x, y) of each data point in the original signal u(x). At the same time, detect all local maximum and minimum values in the signal, and divide the signal into N intervals according to the abscissa of these extreme points.

[0100] 2. Envelope construction by cubic spline interpolation: Construct the upper envelope function U upper (x) and the lower envelope function U lower (x) of the original signal by cubic spline interpolation method, as follows:

[0101] a. In the ith sub-interval [xi, xi+1], where i = 1, 2…N-1, construct a cubic polynomial:

[0102] U i (x) = a i +b i (x-x i )+c i (x-x i ) 2 +d i (x-x i ) 3

[0103] where a i , b i , c i , d i are undetermined coefficients, and x i and x i+1 represent the starting point and the ending point of the ith sub-interval, respectively.

[0104] b. Solve the corresponding coefficients by interpolation conditions, continuity conditions and boundary conditions:

[0105] U i (x i ) = y i and U i (x i+1 ) = y i+1 (interpolation conditions)

[0106] U′ i (x i ) = U′ i+1 (x i+1 ) and U″ i (x i+1 ) = U″ i+1 (x i+1 )(first and second derivative continuity conditions)

[0107] U″1(x1) = 0 and U″ N (x N+1 ) = 0 (natural boundary conditions).

[0108] c. Substitute the obtained coefficient into the cubic polynomial U i (x) to obtain the envelope function of each sub-interval in the original signal.

[0109] 3. Calculation of candidate intrinsic function: Calculate the average value of the upper envelope and the lower envelope as the envelope average value, and subtract the average value from the original signal to obtain the candidate intrinsic mode function (IMF), as follows

[0110] a. The envelope average value is calculated as follows:

[0111]

[0112] b. The candidate intrinsic function (IMF) is calculated as follows:

[0113] z(x) = u(x) - m(x)

[0114] 4. IMF determination and iterative extraction: Determine whether the IMF meets the corresponding conditions, if not, local iterative processing needs to be performed on z(x) until the IMF that meets the conditions is extracted. The specific steps are as follows:

[0115] a. IMF condition determination: The difference between the extreme point and the zero-crossing point in the entire data range is not more than 1; the local mean value in the entire data range is approximately 0 (or changes very small).

[0116] b. Local iteration: If z(x) does not meet the above conditions, repeat the above process with z(x) as the new signal until the IMF meets the conditions.

[0117] c. Decomposition process iteration: Once the first IMF that meets the conditions is obtained, z(x) is considered as the first IMF, i.e. IMF1(x) = z(x)

[0118] d. Update the remaining signal

[0119] r(x) = u(x) - IMF g (x)

[0120] Repeat the above screening process for r(x) to continuously screen candidate IMFs and obtain IMF2(x), IMF3(x)…

[0121] e. Termination condition: When the remaining signal r(x) becomes a monotonic function or the number of extreme points is too small (cannot continue effective interpolation), stop decomposition to obtain the final residual r n (x)

[0122] e. Final decomposition result: Finally, the original signal is represented as the sum of several IMF components and residual, i.e.

[0123]

[0124] wherein, U(x) represents the input signal, IMF g (x) represents the gth intrinsic mode function, r n (x) represents the residual.

[0125] g. Component analysis: through the analysis of the instantaneous frequency and energy of each IMF component, the system net load components can be recombined and classified into three categories: high-frequency components, medium-frequency components and low-frequency components.

[0126] Specifically, after obtaining the predicted net load curve, the present application performs empirical mode decomposition on the predicted net load curve, aiming to adaptively decompose the complex, non-stationary time series signal into a set of limited intrinsic mode function components with different intrinsic time scale characteristics, and finally generate predicted high-frequency components, predicted medium-frequency components and predicted low-frequency components that can reflect different fluctuation rates.

[0127] In a preferred embodiment, the decomposition process is realized by repeatedly performing an intrinsic mode function extraction operation until the final predicted residual curve meets the preset decomposition termination condition, such as becoming a monotonic function. The extraction operation aims to separate a qualified intrinsic mode function from the current predicted curve as its input. Specifically, the operation first extracts the local extreme points of the current predicted curve to construct its upper and lower envelope lines, and calculates the envelope mean, and then subtracts the envelope mean from the current curve to obtain a candidate intrinsic mode function. Subsequently, the candidate intrinsic mode function is judged, if it does not meet the preset convergence condition as an intrinsic mode function, it itself is taken as a new processing object and the steps of constructing envelope, calculating mean and subtracting are repeated, and this iterative process of repeated purification is the screening process. When the screened function meets the convergence condition, it is determined as a qualified predicted intrinsic mode function, and is subtracted from the un-screened current predicted curve to generate a predicted residual curve for the next round of extraction operation.

[0128] After obtaining the set containing multiple predicted intrinsic mode functions through the above repeated operation, frequency analysis is performed on each predicted intrinsic mode function in the set. According to the instantaneous frequency characteristics of each intrinsic mode function, functions with similar frequency characteristics are classified and superimposed, thereby recombining to generate the predicted high-frequency components, predicted medium-frequency components and predicted low-frequency components. Through the above decomposition process, the effective separation of different time scale fluctuation components in the original predicted net load curve is realized, which lays a foundation for subsequent independent analysis and uncertainty modeling of each type of fluctuation.

[0129] Step S3, performing empirical mode decomposition on the actual net load curve to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component.

[0130] In a preferred embodiment, the empirical mode decomposition on the actual net load curve to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component comprises:

[0131] The actual intrinsic mode function extraction operation is repeatedly performed until the current actual residual curve meets the preset decomposition termination condition, and an actual intrinsic mode function set is generated;

[0132] For each actual intrinsic mode function in the actual intrinsic mode function set, frequency analysis is performed to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component;

[0133] The intrinsic mode function extraction operation comprises:

[0134] Extracting local maximum values and local minimum values in the current actual curve to generate each current actual local maximum value and each current actual local minimum value; wherein the initial actual curve is the actual net load curve;

[0135] Based on the cubic spline interpolation method, constructing an upper envelope function of the current actual curve and a lower envelope function of the current actual curve according to each current actual local maximum value and each current actual local minimum value;

[0136] According to the upper envelope function of the current actual curve and the lower envelope function of the current actual curve, calculating and generating an envelope mean function of the current actual curve;

[0137] According to the envelope mean function of the current actual curve and the current actual curve, determining an actual intrinsic mode function of the current actual curve;

[0138] According to the actual intrinsic mode function of the current actual curve and the current actual curve, calculating and generating a current actual residual curve;

[0139] Adding the actual intrinsic mode function of the current actual curve to the actual intrinsic mode function set;

[0140] Determining whether the current actual residual curve meets the preset decomposition termination condition, if yes, outputting the actual intrinsic mode function set; if not, updating the current actual residual curve as the current actual curve.

[0141] In a specific implementation, corresponding to the processing of the predicted net load curve, the present application also performs empirical mode decomposition on the obtained actual net load curve. The purpose of this step is to decompose the complex net load fluctuation that has actually occurred into a set of actual intrinsic mode function components with different intrinsic time scale characteristics, thereby generating actual high frequency components, actual medium frequency components and actual low frequency components that can serve as objective benchmarks.

[0142] The decomposition process is consistent with the method used when processing the predicted curve, and is achieved by repeatedly performing an actual intrinsic mode function extraction operation. This operation aims to separate a qualified actual intrinsic mode function from the current actual curve as its input. Its specific process includes: extracting the local extreme points of the current actual curve to construct its upper and lower envelope lines, and calculating the envelope mean to obtain a candidate intrinsic mode function; and performing an iterative screening process on the candidate function until it meets the preset convergence condition and is determined as a qualified actual intrinsic mode function. Subsequently, this screened and determined intrinsic mode function is subtracted from the current actual curve to generate an actual residual curve for the next round of extraction operation. The entire process is repeated until the final actual residual curve meets the preset decomposition termination condition.

[0143] After obtaining the set containing multiple actual intrinsic mode functions through the above repeated operation, frequency analysis is also performed on each actual intrinsic mode function in the set, and classification and superposition are performed according to its frequency characteristics to recombine the actual high frequency components, actual medium frequency components and actual low frequency components. By performing the same decomposition operation on the actual net load curve, actual fluctuation components corresponding to the predicted components in the time scale are obtained, which provides an objective benchmark for accurately calculating the prediction error of each frequency in the next step.

[0144] Step S4, calculate the difference between the predicted high frequency component and the actual high frequency component to generate a high frequency error sequence. Calculate the difference between the predicted medium frequency component and the actual medium frequency component to generate a medium frequency error sequence. Calculate the difference between the predicted low frequency component and the actual low frequency component to generate a low frequency error sequence.

[0145] Specifically, by subtracting the actual high frequency component from the corresponding predicted high frequency component, a time sequence reflecting the prediction deviation of the high frequency part, i.e. the high frequency error sequence, is obtained. Using the same method, the difference between the predicted medium frequency component and the actual medium frequency component is calculated to generate a medium frequency error sequence, and the difference between the predicted low frequency component and the actual low frequency component is calculated to generate a low frequency error sequence. Through this step, the problem of direct analysis of complex net load curves is transformed into the problem of independent statistical characteristic analysis of high, medium and low frequency prediction error sequences, so that different types of uncertainty can be quantitatively processed more targetedly.

[0146] Step S5, respectively, the high frequency error sequence, the medium frequency error sequence and the low frequency error sequence are modeled probability distribution, generating high frequency error probability density function, medium frequency error probability density function and low frequency error probability density function.

[0147] In a preferred embodiment, respectively, the high frequency error sequence, the medium frequency error sequence and the low frequency error sequence are modeled probability distribution, generating high frequency error probability density function, medium frequency error probability density function and low frequency error probability density function, including:

[0148] Respectively, the high frequency error sequence, the medium frequency error sequence and low frequency error sequence are sampled, generating each high frequency error sample point, each medium frequency error sample point and each low frequency error sample point;

[0149] According to each high frequency error sample point, each medium frequency error sample point and each low frequency error sample point, respectively, the number of high frequency error sample points, the number of medium frequency error sample points and the number of low frequency error sample points are determined;

[0150] According to each high frequency error sample point, and the number of high frequency error sample points, the high frequency bandwidth is determined; according to each medium frequency error sample point, and the number of medium frequency error sample points, the medium frequency bandwidth is determined; according to each low frequency error sample point, and the number of low frequency error sample points, the low frequency bandwidth is determined;

[0151] Using Gaussian kernel function, based on the high frequency bandwidth, generating high frequency error probability density function; using Gaussian kernel function, based on the medium frequency bandwidth, generating medium frequency error probability density function; using Gaussian kernel function, based on the low frequency bandwidth, generating low frequency error probability density function.

[0152] For example, the high frequency part is taken as an example, the obtained high frequency error sample point is fitted with kernel density estimation method. The specific steps are as follows:

[0153] 1. Bandwidth selection: using Gaussian kernel function, and determining the optimal bandwidth h according to Silverman empirical rule, the calculation formula is as follows

[0154]

[0155] In the formula: h is the optimal bandwidth; e is the number of high frequency error sample points; is the sample standard deviation of all time period high frequency net load difference fluctuation data; j is the sample serial number of high frequency error sample point; x j is the value of high frequency error sample point i; x is the sample mean of high frequency error sample point.

[0156] 2. Probability kernel density estimation

[0157] Probability density function of the obtained high frequency net load difference in each 15-minute period may be expressed as:

[0158]

[0159] wherein x 1 ...x 96 represent the high frequency net load difference variables of the 1st…96th 15-minute periods, respectively; represent the jth high frequency net load difference sample data of the 1st…96th 15-minute periods, respectively.

[0160] In a specific implementation, the probability distribution modeling is implemented by using a kernel density estimation method. For each error sequence of high, medium and low frequencies, the method first needs to determine the optimal bandwidth. In a specific implementation, the optimal bandwidth is calculated according to the sample quantity and sample standard deviation of the error sequence according to the Silverman empirical rule. After the optimal bandwidth of each error sequence is determined, the error sequence is processed by using a Gaussian kernel function. The process is centered on each error sample point, and a Gaussian distribution function with a shape determined by the optimal bandwidth is superimposed. The Gaussian distribution functions corresponding to all error sample points are accumulated and normalized, and finally the high frequency error probability density function, the medium frequency error probability density function and the low frequency error probability density function which can smoothly and continuously describe the distribution characteristics of the error sequence are generated.

[0161] By establishing independent probability density functions for each type of error sequence, the application realizes accurate and quantitative description of prediction uncertainty under different time scales, and provides a scientific mathematical model basis for subsequent risk and confidence-based reserve capacity decision-making.

[0162] Step S6, based on the preset confidence, respectively according to the high frequency error probability density function, the medium frequency error probability density function and the low frequency error probability density function, the high frequency reserve capacity, the medium frequency reserve capacity and the low frequency reserve capacity of the power system are determined.

[0163] For example, taking the high frequency part as an example, under a preset confidence level, the upper limit x max of the high frequency net load difference of each period is calculated by using the estimated probability density function, as the required high frequency reserve capacity under the worst case:

[0164]

[0165] Δ high = max(x max , 0)

[0166] wherein: is the high-frequency error probability density function of the net load fluctuation difference; E is the high-frequency net load fluctuation difference; x max is the upper limit of the high-frequency net load fluctuation difference at a certain confidence level (a positive value indicates an increase in the net load); x min is the lower limit of the high-frequency net load fluctuation difference at a certain confidence level (a negative value indicates a decrease in the net load); x mid is the median of the high-frequency net load fluctuation difference probability density function; c is a certain confidence level; Δ high is the high-frequency reserve capacity.

[0167] In a specific implementation, after obtaining the probability density functions of the high, medium and low frequency errors, the high, medium and low frequency reserve capacities required by the power system are determined based on the preset confidence level according to the probability density functions respectively. This step is the decision output link of the method, which aims to convert the probability quantization results of the uncertainty in the previous steps into specific executable reserve capacity megawatt values. The confidence level is a key parameter reflecting the required safety and reliability level of the system, which is preset by the power grid management department according to the operation criteria. The determination process is realized by solving the inverse cumulative distribution functions of the error probability density functions. Specifically, for the high-frequency error probability density function, the system calculates the error upper limit value that can cover the probability interval required by the preset confidence level. The upper limit value represents the high-frequency reserve capacity required by the system to resist high-frequency fluctuation uncertainty at the set reliability level. Similarly, the medium-frequency reserve capacity and the low-frequency reserve capacity required by the system are determined according to the medium-frequency error probability density function and the low-frequency error probability density function respectively.

[0168] Through this step, the application finally generates a set of reserve capacity requirements matched with different time scale fluctuation characteristics, which provides a more scientific and explicit decision basis for subsequent refined and economic dispatching compared with the single reserve value obtained by the traditional method.

[0169] Another embodiment of the application provides a method for determining the reserve capacity of a power system. In addition to the steps S1-S6 of the above embodiment, the method further comprises an optimized allocation and dispatching step of the reserve capacity after step S6. This step aims to reasonably allocate the high, medium and low frequency reserve capacity requirements calculated in the previous steps according to the adjustment performance, response speed and economic factors of different power generation subjects such as energy storage, gas power plants and coal power units. Specifically, the high-frequency reserve capacity requiring fast response is allocated to fast resources such as energy storage, the medium-frequency reserve capacity is allocated to gas power plants with moderate adjustment performance, and the low-frequency reserve capacity is borne by conventional energy sources such as coal power with slower response but lower cost, thereby forming a hierarchical reserve dispatching scheme accurately matched with multi-scale fluctuation characteristics and taking into account safety and economy.

[0170] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0171] like Figure 2 As shown, an embodiment of the present invention provides a device for determining the reserve capacity of a power system, comprising: a data acquisition module, an empirical mode decomposition module, an error sequence determination module, a probability modeling module, and a reserve capacity determination module;

[0172] The data acquisition module is used to acquire the predicted net load curve and the actual net load curve of a typical day in the power system.

[0173] The empirical mode decomposition module is used to perform empirical mode decomposition on the predicted net load curve to generate predicted high-frequency components, predicted mid-frequency components, and predicted low-frequency components; and to perform empirical mode decomposition on the actual net load curve to generate actual high-frequency components, actual mid-frequency components, and actual low-frequency components.

[0174] The error sequence determination module is used to calculate the difference between the predicted high-frequency component and the actual high-frequency component to generate a high-frequency error sequence; calculate the difference between the predicted intermediate-frequency component and the actual intermediate-frequency component to generate an intermediate-frequency error sequence; and calculate the difference between the predicted low-frequency component and the actual low-frequency component to generate a low-frequency error sequence.

[0175] The probability modeling module is used to model the probability distribution of the high-frequency error sequence, the mid-frequency error sequence, and the low-frequency error sequence respectively, and generate the high-frequency error probability density function, the mid-frequency error probability density function, and the low-frequency error probability density function.

[0176] The reserve capacity determination module is used to determine the high-frequency reserve capacity, medium-frequency reserve capacity, and low-frequency reserve capacity of the power system based on a preset confidence level, according to the high-frequency error probability density function, the medium-frequency error probability density function, and the low-frequency error probability density function, respectively.

[0177] In a preferred embodiment, the empirical mode decomposition module is used to perform empirical mode decomposition on the predicted net load curve to generate predicted high-frequency components, predicted mid-frequency components, and predicted low-frequency components, including:

[0178] Repeat the predicted intrinsic mode function extraction operation until the current predicted residual curve meets the preset decomposition termination condition, and generate a set of predicted intrinsic mode functions;

[0179] For each predicted intrinsic mode function in the predicted intrinsic mode function set, frequency analysis is performed to generate predicted high-frequency components, predicted mid-frequency components, and predicted low-frequency components.

[0180] The predicted intrinsic mode function extraction operation includes:

[0181] extracting local maximum values and local minimum values in the current prediction curve to generate respective current prediction local maximum values and respective current prediction local minimum values; wherein the initial prediction curve is a prediction net load curve;

[0182] constructing, according to the respective current prediction local maximum values and the respective current prediction local minimum values, an upper envelope function of the current prediction curve and a lower envelope function of the current prediction curve based on a cubic spline interpolation method;

[0183] generating, according to the upper envelope function of the current prediction curve and the lower envelope function of the current prediction curve, an envelope mean function of the current prediction curve;

[0184] determining, according to the envelope mean function of the current prediction curve and the current prediction curve, a prediction intrinsic mode function of the current prediction curve;

[0185] generating, according to the prediction intrinsic mode function of the current prediction curve and the current prediction curve, a current prediction residual curve;

[0186] adding the prediction intrinsic mode function of the current prediction curve to the set of prediction intrinsic mode functions;

[0187] judging whether the current prediction residual curve satisfies a preset decomposition termination condition, and if yes, outputting the set of prediction intrinsic mode functions, and if no, updating the current prediction residual curve as the current prediction curve.

[0188] In one preferred embodiment, the empirical mode decomposition module is configured to perform empirical mode decomposition on the actual net load curve to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component, and includes:

[0189] repeating the actual intrinsic mode function extraction operation until the current actual residual curve satisfies the preset decomposition termination condition to generate a set of actual intrinsic mode functions;

[0190] performing frequency analysis on each actual intrinsic mode function in the set of actual intrinsic mode functions to generate the actual high-frequency component, the actual medium-frequency component and the actual low-frequency component;

[0191] The intrinsic mode function extraction operation includes:

[0192] extracting local maximum values and local minimum values in the current actual curve to generate respective current actual local maximum values and respective current actual local minimum values; wherein the initial actual curve is an actual net load curve;

[0193] constructing, according to the respective current actual local maximum values and the respective current actual local minimum values, an upper envelope function of the current actual curve and a lower envelope function of the current actual curve based on a cubic spline interpolation method;

[0194] According to the upper envelope function of the current actual curve and the lower envelope function of the current actual curve, an envelope mean function of the current actual curve is calculated and generated;

[0195] According to the envelope mean function of the current actual curve and the current actual curve, an actual eigenmode function of the current actual curve is determined;

[0196] According to the actual eigenmode function of the current actual curve and the current actual curve, a current actual residual curve is calculated and generated;

[0197] The actual eigenmode function of the current actual curve is added to the set of actual eigenmode functions;

[0198] It is judged whether the current actual residual curve meets a preset decomposition termination condition, if yes, the set of actual eigenmode functions is output; if not, the current actual residual curve is updated as the current actual curve.

[0199] In a preferred embodiment, the probability modeling module is configured to model the probability distribution of the high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence respectively, and generate a high-frequency error probability density function, a medium-frequency error probability density function and a low-frequency error probability density function, including:

[0200] The high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence are sampled respectively to generate high-frequency error sample points, medium-frequency error sample points and low-frequency error sample points;

[0201] The number of high-frequency error sample points, the number of medium-frequency error sample points and the number of low-frequency error sample points are determined respectively according to the high-frequency error sample points, the medium-frequency error sample points and the low-frequency error sample points;

[0202] The high-frequency bandwidth is determined according to the high-frequency error sample points and the number of high-frequency error sample points; the medium-frequency bandwidth is determined according to the medium-frequency error sample points and the number of medium-frequency error sample points; and the low-frequency bandwidth is determined according to the low-frequency error sample points and the number of low-frequency error sample points;

[0203] The high-frequency error probability density function is generated based on the high-frequency bandwidth by using a Gaussian kernel function; the medium-frequency error probability density function is generated based on the medium-frequency bandwidth by using a Gaussian kernel function; and the low-frequency error probability density function is generated based on the low-frequency bandwidth by using a Gaussian kernel function.

[0204] It should be noted that the above-described embodiments of the apparatus correspond to the above-described embodiments of the application, and can implement the determination method of the backup capacity of the power system according to any one of the above-described embodiments of the application. In addition, the above-described embodiments of the apparatus are only illustrative, and the modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical units, i.e., can be located in one place or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment. In addition, the connection relationship between the modules in the apparatus embodiment provided by the application indicates that there is a communication connection between them, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0205] On the basis of the above-described method embodiments of the application, an electronic device embodiment is provided.

[0206] An embodiment of the application provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, when the computer program is executed by the processor, a determination method of the backup capacity of the power system according to any one of the above-described embodiments of the application is implemented, or when the computer program is executed by the processor, the functions of the modules in the above-described apparatus embodiments are implemented.

[0207] For example, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the application. The one or more modules can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the terminal device.

[0208] The terminal device can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing devices. The terminal device can include, but is not limited to, a processor and a memory.

[0209] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The processor is a control center of the terminal device, and connects all parts of the terminal device through various interfaces and lines.

[0210] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the terminal device by running or executing the computer program and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store an operating system, at least one application program required by a function, etc. The data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0211] On the basis of the above-mentioned method embodiment, the application provides a storage medium embodiment;

[0212] Another embodiment of the application provides a storage medium, which comprises a stored computer program. When the computer program runs, the device where the storage medium is located performs the method for determining the reserve capacity of the power system.

[0213] The storage medium is a computer readable storage medium, and the computer program includes computer program code in the form of source code, object code, an executable file, or some intermediate form, etc. The computer readable medium can include any entity or device capable of carrying the computer program code, a recording medium, a U disk, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, a software distribution medium, etc.

[0214] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0215] The above is the preferred embodiment of the present application. It should be pointed out that for those skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which are also considered to be within the scope of protection of the present application.

Claims

1. A method of determining reserve capacity of an electric power system, characterized by, The method comprises the following steps: obtaining a predicted net load curve and an actual net load curve of a typical day of a power system; performing empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component; performing empirical mode decomposition on the actual net load curve to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component; calculating the difference between the predicted high-frequency component and the actual high-frequency component to generate a high-frequency error sequence; calculating the difference between the predicted medium-frequency component and the actual medium-frequency component to generate a medium-frequency error sequence; calculating the difference between the predicted low-frequency component and the actual low-frequency component to generate a low-frequency error sequence; respectively modeling the probability distribution of the high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence to generate a high-frequency error probability density function, a medium-frequency error probability density function and a low-frequency error probability density function; based on a preset confidence level, determining the high-frequency reserve capacity, the medium-frequency reserve capacity and the low-frequency reserve capacity of the power system according to the high-frequency error probability density function, the medium-frequency error probability density function and the low-frequency error probability density function respectively.

2. The method of claim 1, wherein the reserve capacity of the power system is determined based on the power system model and the power system operating condition. The method comprises the following steps: performing empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component, comprising: repeatedly performing a predicted intrinsic mode function extraction operation until a current predicted residual curve meets a preset decomposition termination condition to generate a set of predicted intrinsic mode functions; performing frequency analysis on each predicted intrinsic mode function in the set of predicted intrinsic mode functions to generate the predicted high-frequency component, the predicted medium-frequency component and the predicted low-frequency component; wherein the predicted intrinsic mode function extraction operation comprises: extracting local maximum values and local minimum values in a current predicted curve to generate respective current predicted local maximum values and respective current predicted local minimum values; wherein the initial predicted curve is the predicted net load curve; constructing an upper envelope function of the current predicted curve and a lower envelope function of the current predicted curve based on a cubic spline interpolation method according to the respective current predicted local maximum values and the respective current predicted local minimum values; calculating and generating an envelope mean function of the current predicted curve according to the upper envelope function of the current predicted curve and the lower envelope function of the current predicted curve; determining a predicted intrinsic mode function of the current predicted curve according to the envelope mean function of the current predicted curve and the current predicted curve; calculating and generating a current predicted residual curve according to the predicted intrinsic mode function of the current predicted curve and the current predicted curve; adding the predicted intrinsic mode function of the current predicted curve to the set of predicted intrinsic mode functions; 3. The method of claim 2, wherein the step of determining the reserve capacity of the power system comprises the steps of: determining the reserve capacity of the power system based on the power system's reserve capacity formula. determining whether the current predicted residual curve meets the preset decomposition termination condition, and if so, outputting the set of predicted intrinsic mode functions; if not, updating the current predicted residual curve as the current predicted curve. The method comprises the following steps: performing empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component, comprising: repeatedly performing an actual intrinsic mode function extraction operation until a current actual residual curve meets a preset decomposition termination condition to generate a set of actual intrinsic mode functions; For each actual eigenmode function in the actual eigenmode function set, frequency analysis is performed to generate actual high-frequency components, actual medium-frequency components and actual low-frequency components; The eigenmode function extraction operation includes: Extracting local maximum values and local minimum values in the current actual curve to generate current actual local maximum values and current actual local minimum values; the initial actual curve is an actual net load curve; Based on the cubic spline interpolation method, an upper envelope function of the current actual curve and a lower envelope function of the current actual curve are constructed according to the current actual local maximum values and the current actual local minimum values; The envelope mean function of the current actual curve is calculated according to the upper envelope function of the current actual curve and the lower envelope function of the current actual curve; The actual eigenmode function of the current actual curve is determined according to the envelope mean function of the current actual curve and the current actual curve; The current actual residual curve is calculated according to the actual eigenmode function of the current actual curve and the current actual curve; The actual eigenmode function of the current actual curve is added to the actual eigenmode function set; If the current actual residual curve meets the preset decomposition termination condition, the actual eigenmode function set is output; if not, the current actual residual curve is updated to the current actual curve.

4. The method for determining the reserve capacity of a power system as described in claim 3, characterized in that, The high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence are respectively subjected to probability distribution modeling to generate a high-frequency error probability density function, a medium-frequency error probability density function and a low-frequency error probability density function, including: The high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence are respectively sampled to generate high-frequency error sample points, medium-frequency error sample points and low-frequency error sample points; The number of high-frequency error sample points, the number of medium-frequency error sample points and the number of low-frequency error sample points are determined according to the high-frequency error sample points, the medium-frequency error sample points and the low-frequency error sample points respectively; The high-frequency bandwidth is determined according to the high-frequency error sample points and the number of high-frequency error sample points; the medium-frequency bandwidth is determined according to the medium-frequency error sample points and the number of medium-frequency error sample points; and the low-frequency bandwidth is determined according to the low-frequency error sample points and the number of low-frequency error sample points; The high-frequency error probability density function is generated based on the high-frequency bandwidth by using a Gaussian kernel function; the medium-frequency error probability density function is generated based on the medium-frequency bandwidth by using a Gaussian kernel function; and the low-frequency error probability density function is generated based on the low-frequency bandwidth by using a Gaussian kernel function.

5. An apparatus for determining reserve capacity of an electric power system, characterized by comprising: It includes: A data acquisition module, an empirical mode decomposition module, an error sequence determination module, a probability modeling module and a spare capacity determination module; The data acquisition module is configured to acquire a predicted net load curve and an actual net load curve of a typical day of a power system; The empirical mode decomposition module is configured to perform empirical mode decomposition on the predicted net load curve to generate predicted high-frequency components, predicted medium-frequency components and predicted low-frequency components, and perform empirical mode decomposition on the actual net load curve to generate actual high-frequency components, actual medium-frequency components and actual low-frequency components; The error sequence determination module is configured to calculate a difference between the predicted high-frequency component and the actual high-frequency component to generate a high-frequency error sequence; calculate a difference between the predicted medium-frequency component and the actual medium-frequency component to generate a medium-frequency error sequence; calculate a difference between the predicted low-frequency component and the actual low-frequency component to generate a low-frequency error sequence; The probability modeling module is configured to model probability distributions of the high-frequency error sequence, the medium-frequency error sequence and the low-frequency error sequence respectively to generate a high-frequency error probability density function, a medium-frequency error probability density function and a low-frequency error probability density function; The standby capacity determination module is configured to determine high-frequency standby capacity, medium-frequency standby capacity and low-frequency standby capacity of the power system based on the preset confidence and the high-frequency error probability density function, the medium-frequency error probability density function and the low-frequency error probability density function respectively.

6. The apparatus for determining reserve capacity of a power system of claim 5, wherein, The empirical mode decomposition module is configured to perform empirical mode decomposition on the predicted net load curve to generate a predicted high-frequency component, a predicted medium-frequency component and a predicted low-frequency component, including: repeatedly performing a predicted intrinsic mode function extraction operation until a current predicted residual curve meets a preset decomposition termination condition to generate a predicted intrinsic mode function set; performing frequency analysis on each predicted intrinsic mode function in the predicted intrinsic mode function set to generate the predicted high-frequency component, the predicted medium-frequency component and the predicted low-frequency component; The predicted intrinsic mode function extraction operation includes: extracting local maximum values and local minimum values in a current predicted curve to generate respective current predicted local maximum values and respective current predicted local minimum values; wherein the initial predicted curve is the predicted net load curve; constructing an upper envelope function of the current predicted curve and a lower envelope function of the current predicted curve based on the cubic spline interpolation method according to the respective current predicted local maximum values and the respective current predicted local minimum values; calculating a generated envelope mean function of the current predicted curve according to the upper envelope function of the current predicted curve and the lower envelope function of the current predicted curve; determining a predicted intrinsic mode function of the current predicted curve according to the envelope mean function of the current predicted curve and the current predicted curve; calculating a generated current predicted residual curve according to the predicted intrinsic mode function of the current predicted curve and the current predicted curve; adding the predicted intrinsic mode function of the current predicted curve to the predicted intrinsic mode function set; determining whether the current predicted residual curve meets the preset decomposition termination condition, and if so, outputting the predicted intrinsic mode function set; or if not, updating the current predicted residual curve as the current predicted curve.

7. The apparatus for determining reserve capacity of a power system of claim 6, wherein, The empirical mode decomposition module is configured to perform empirical mode decomposition on the actual net load curve to generate an actual high-frequency component, an actual medium-frequency component and an actual low-frequency component, including: repeatedly performing an actual intrinsic mode function extraction operation until a current actual residual curve meets a preset decomposition termination condition to generate an actual intrinsic mode function set; performing frequency analysis on each actual intrinsic mode function in the actual intrinsic mode function set to generate the actual high-frequency component, the actual medium-frequency component and the actual low-frequency component; The intrinsic mode function extraction operation includes: Extracting local maximum values and local minimum values in the current actual curve to generate each current actual local maximum value and each current actual local minimum value; wherein the initial actual curve is an actual net load curve; According to each current actual local maximum value and each current actual local minimum value, constructing an upper envelope function of the current actual curve and a lower envelope function of the current actual curve based on a cubic spline interpolation method; According to the upper envelope function of the current actual curve and the lower envelope function of the current actual curve, calculating and generating an envelope mean function of the current actual curve; According to the envelope mean function of the current actual curve and the current actual curve, determining an actual intrinsic modal function of the current actual curve; According to the actual intrinsic modal function of the current actual curve and the current actual curve, calculating and generating a current actual residual curve; Adding the actual intrinsic modal function of the current actual curve to the actual intrinsic modal function set; Determining whether the current actual residual curve satisfies a preset decomposition termination condition, and if yes, outputting the actual intrinsic modal function set; if not, updating the current actual residual curve to the current actual curve.

8. The apparatus for determining reserve capacity of a power system of claim 7, wherein, The probability modeling module is configured to model probability distributions of the high-frequency error sequence, the medium-frequency error sequence, and the low-frequency error sequence respectively to generate a high-frequency error probability density function, a medium-frequency error probability density function, and a low-frequency error probability density function, including: The high-frequency error sequence, the medium-frequency error sequence, and the low-frequency error sequence are sampled respectively to generate high-frequency error sample points, medium-frequency error sample points, and low-frequency error sample points; The number of high-frequency error sample points, the number of medium-frequency error sample points, and the number of low-frequency error sample points are determined respectively according to the high-frequency error sample points, the medium-frequency error sample points, and the low-frequency error sample points; A high-frequency bandwidth is determined according to the high-frequency error sample points and the number of high-frequency error sample points, a medium-frequency bandwidth is determined according to the medium-frequency error sample points and the number of medium-frequency error sample points, and a low-frequency bandwidth is determined according to the low-frequency error sample points and the number of low-frequency error sample points; The high-frequency error probability density function is generated based on the high-frequency bandwidth by using a Gaussian kernel function, the medium-frequency error probability density function is generated based on the medium-frequency bandwidth by using the Gaussian kernel function, and the low-frequency error probability density function is generated based on the low-frequency bandwidth by using the Gaussian kernel function.

9. An electronic device, comprising: The storage medium comprises a stored computer program, wherein the storage medium controls a device in which the storage medium is located to execute the determination method of the reserve capacity of the power system according to any one of claims 1 to 4 when the computer program is executed.

10. A storage medium, characterized by The storage medium comprises a stored computer program, wherein the storage medium controls a device in which the storage medium is located to execute the determination method of the reserve capacity of the power system according to any one of claims 1 to 4 when the computer program is executed.

Images