Bus load prediction method based on high-order multivariate Markov chain and related device

Abstract

The invention discloses a bus load prediction method based on a high-order multivariate Markov chain and a related device, and the method comprises the steps: obtaining the daily load data of a plurality of adjacent buses from a power grid regulation and control operation system, and building a bus load sequence sample in combination with the number of load sampling points of each bus; converting the bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between the bus load sequence samples; determining the order of the high-order multivariate Markov chain model matrix; estimating a transition probability matrix of the high-order multivariate Markov chain model matrix through historical data; estimating the inter-column high-order weight of the high-order multivariate Markov chain model matrix; and substituting the determined order, the transition probability matrix and the inter-column high-order weight into the corrected high-order multivariate Markov chain model matrix, and predicting the bus load by using the corrected high-order multivariate Markov chain model matrix. According to the invention, influence factors of time and space are brought into a consideration range of bus load prediction at the same time, and the prediction result is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a bus load prediction method based on a high-order multivariate Markov chain and a related device. Background Art

[0002] Bus load refers to the sum of the terminal loads in the area powered by the main transformer of the substation. Bus load forecasting allows control personnel to understand the development and change trends of the load in a timely manner and make arrangements for the power generation of various power sources in advance. At the same time, it also lays a computational foundation for the calculation of the load-side regulation capacity. It provides strong support for the stable and reliable operation of the new power system. Affected by factors such as geographical distribution, system operation mode, and load transfer, there is a coupling relationship between the operating characteristics of bus loads with similar physical distances. In addition, as a time series, the change trend of bus load is also affected by historical loads.

[0003] In the relevant prior art, for example, the patent application with publication number CN117933451A, “An interpretable bus load forecasting method based on attention mechanism”, solves the problem that there are a large number of buses in the prior art, and the traditional bus load forecasting model adopts traditional machine learning or deep learning models, and the model is complex and difficult to interpret, and it is difficult to perform root cause analysis and summary. By selecting historical load data and multiple feature variables related to it as input, and using the attention mechanism to adaptively adjust the importance of different feature variables, high-precision and highly interpretable reasoning performance is achieved, and finally a time series model containing multiple feature variables is constructed to predict future bus loads. In addition, the prediction results are analyzed, the performance and robustness of the model are evaluated, and the model structure or parameters are further adjusted to improve the prediction accuracy and stability. For another example, the patent application with publication number CN117498311A, "A bus load prediction method and system considering the impact of distributed photovoltaics", includes the following steps: generating bus power load data and distributed photovoltaic output data at different times in real time according to the current operation mode, analyzing the impact of different weather factors on bus power load and distributed photovoltaics, establishing power load prediction models and distributed photovoltaic prediction models according to the weather factor analysis results and data, inputting historical data and predicted weather factor data, obtaining bus power load prediction results and distributed photovoltaic output prediction results, and calculating the prediction results of bus load off-grid according to the connection relationship between the bus and distributed photovoltaics. By adopting real-time topological relationship analysis, the active data of bus power load is accurately restored, providing accurate sample data for bus load prediction, and effectively improving the accuracy of bus load prediction. However, the above existing methods are mainly based on analyzing the impact of weather, time, temperature and humidity on bus load, and using neural networks and other methods to predict bus load according to historical loads, ignoring the impact of load transfer and adjacent bus loads under different operating modes, and the prediction accuracy is difficult to improve to a higher level. Summary of the invention

[0004] The purpose of the present invention is to provide a bus load prediction method and related device based on high-order multivariate Markov chain to solve the problems in the prior art. The high-order multivariate Markov chain is used to predict the bus load, and the time and space influencing factors are taken into consideration to make the prediction more accurate. According to the load of the adjacent bus and the historical time series, the hidden coupling relationship between the bus loads is effectively excavated, and the bus load is predicted, which is more in line with reality.

[0005] In order to achieve the above object, the present invention has the following technical solutions:

[0006] In a first aspect, an embodiment of the present invention provides a bus load forecasting method based on a high-order multivariate Markov chain, comprising:

[0007] Obtain daily load data of several adjacent buses from the power grid control and operation system, and establish bus load sequence samples based on the number of load sampling points on each bus;

[0008] Based on the correlation between bus load sequence samples, the bus load sequence samples are converted into a high-order multivariate Markov chain model matrix;

[0009] Determine the order of the matrix for a high-order multivariate Markov chain model;

[0010] Estimate the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data;

[0011] Estimate high-order weights between columns of a high-order multivariate Markov chain model matrix;

[0012] The order of the high-order multivariate Markov chain model matrix, the transfer probability matrix and the high-order weights between columns are substituted into the modified high-order multivariate Markov chain model matrix, and the bus load is predicted using the modified high-order multivariate Markov chain model matrix.

[0013] As a preferred solution, in the step of obtaining daily load data of several adjacent buses from the power grid control and operation system and establishing a bus load sequence sample in combination with the number of load sampling points of each bus, the daily load data of s adjacent buses are obtained from the control and operation system, the number of load sampling points of each bus is N, and the bus load sequence sample is expressed as follows:

[0014]

[0015] Wherein, the subscript t represents the tth sampling point, and the superscript k represents the data of the kth bus.

[0016] As a preferred solution, in the step of converting the bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between the bus load sequence samples, the correlation between the bus load sequence samples conforms to the following expression:

[0017]

[0018] In the formula, is the high-order weight between columns, representing the influence coefficient between each bus load sequence; n is the order of the high-order multivariate Markov chain model matrix;

[0019]

[0020] represents the h-step transition probability matrix from the state of the j-th sequence at time r-h+1 to the state of the k-th sequence at time r+1;

[0021] make in,

[0022] Then the mathematical expression of the high-order multivariate Markov chain model matrix is ​​as follows:

[0023]

[0024] In the formula,

[0025]

[0026] As a preferred solution, the step of determining the order of the high-order multivariate Markov chain model matrix uses the Pearson correlation coefficient to calculate the correlation between the two bus load series samples. The calculation expression is as follows:

[0027]

[0028] Where, d = 0, 1, 2, ..., N-1, 1 ≤ i ≤ s, 1 ≤ j ≤ s;

[0029] When γ (ij) (d) When the absolute value is less than the threshold value, it is determined that the correlation between the two bus load sequence samples is weak. By calculating the correlation coefficients between different bus load sequence samples at different times, the correlation step d between the two bus load sequence samples is determined. The largest d between all bus load sequence samples is taken as the order n of the high-order multivariate Markov chain model matrix:

[0030]

[0031] As a preferred solution, the step of estimating the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data includes:

[0032] Divide the bus load into m intervals evenly and calculate the minimum bus load P in history min and the maximum bus load P max , then the load interval corresponding to the kth state is [(k-1)(P max -P min ) / m,k(P max -P min ) / m], where k = 1, 2, ... m; by counting the transition frequencies between historical sample data sequences, the transition probability matrix is ​​obtained:

[0033]

[0034] in,

[0035]

[0036] In the formula, It is represented by the bus load sequence {x (j)} a state i j The bus load sequence {x (k)} a state i k The frequency of transfer.

[0037] As a preferred solution, the step of estimating the high-order weights between columns of the high-order multivariate Markov chain model matrix includes:

[0038] By seeking The minimization method is used to solve the problem, where is the occurrence ratio of each state in the jth sequence;

[0039] The corresponding high-order weights between columns are obtained by solving the following optimization problem:

[0040]

[0041] As a preferred solution, the step of substituting the order of the high-order multivariate Markov chain model matrix, the transition probability matrix and the high-order weights between columns into the modified high-order multivariate Markov chain model matrix, the modified high-order multivariate Markov chain model matrix is ​​expressed as:

[0042]

[0043] because in When using the modified high-order multivariate Markov chain model matrix to predict bus load, the first prediction vector of each bus load is taken Then the bus load at time t+1 is The predicted value is The corresponding value.

[0044] In a second aspect, an embodiment of the present invention provides a bus load prediction system based on a high-order multivariate Markov chain, comprising:

[0045] The bus load sequence sample establishment module is used to obtain the daily load data of several adjacent buses from the power grid control and operation system, and establish the bus load sequence sample in combination with the number of load sampling points of each bus;

[0046] A high-order multivariate Markov chain model matrix conversion module is used to convert bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between bus load sequence samples;

[0047] A matrix order determination module is used to determine the order of the matrix of the high-order multivariate Markov chain model;

[0048] The transition probability matrix estimation module is used to estimate the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data;

[0049] Matrix column-to-column high-order weight estimation module, used to estimate the column-to-column high-order weights of the high-order multivariate Markov chain model matrix;

[0050] The high-order multivariate Markov chain model matrix correction and bus load prediction module is used to substitute the order of the high-order multivariate Markov chain model matrix, the transfer probability matrix and the high-order weights between columns into the corrected high-order multivariate Markov chain model matrix, and use the corrected high-order multivariate Markov chain model matrix to predict the bus load.

[0051] As a preferred solution, the bus load sequence sample establishment module obtains the daily load data of s adjacent buses from the control operation system. The number of load sampling points of each bus is N. The bus load sequence sample is expressed as follows:

[0052]

[0053] Wherein, the subscript t represents the tth sampling point, and the superscript k represents the data of the kth bus.

[0054] As a preferred solution, in the step of converting the bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between the bus load sequence samples, the correlation between the bus load sequence samples conforms to the following expression:

[0055]

[0056] In the formula, is the high-order weight between columns, representing the influence coefficient between each bus load sequence; n is the order of the high-order multivariate Markov chain model matrix;

[0057]

[0058] represents the h-step transition probability matrix from the state of the j-th sequence at time r-h+1 to the state of the k-th sequence at time r+1;

[0059] make in,

[0060] Then the mathematical expression of the high-order multivariate Markov chain model matrix is ​​as follows:

[0061]

[0062] In the formula,

[0063]

[0064] As a preferred solution, the matrix order determination module uses the Pearson correlation coefficient to calculate the correlation between two bus load sequence samples. The calculation expression is as follows:

[0065]

[0066] Where, d = 0, 1, 2, ..., N-1, 1 ≤ i ≤ s, 1 ≤ j ≤ s;

[0067] When γ (ij) (d) When the absolute value is less than the threshold value, it is determined that the correlation between the two bus load sequence samples is weak. By calculating the correlation coefficients between different bus load sequence samples at different times, the correlation step d between the two bus load sequence samples is determined. The largest d between all bus load sequence samples is taken as the order n of the high-order multivariate Markov chain model matrix:

[0068]

[0069] As a preferred solution, the transfer probability matrix estimation module divides the bus load into m intervals on average, and calculates the minimum bus load P in the history. min and the maximum bus load P max , then the load interval corresponding to the kth state is [(k-1)(P max -P min ) / m,k(P max -P min ) / m], where k = 1, 2, ... m; by counting the transition frequencies between historical sample data sequences, the transition probability matrix is ​​obtained:

[0070]

[0071] in,

[0072]

[0073] In the formula, It is represented by the bus load sequence {x (j)} a state i j The bus load sequence {x (k)} a state i k The frequency of transfer.

[0074] As a preferred solution, the matrix column high-order weight estimation module seeks to make The minimization method is used to solve the problem, where is the occurrence ratio of each state of the jth sequence; the corresponding high-order weights between columns are obtained by solving the following optimization problem:

[0075]

[0076] As a preferred solution, the matrix expression of the high-order multivariate Markov chain model after the high-order multivariate Markov chain model matrix correction and bus load forecasting module correction is:

[0077]

[0078] because in When using the modified high-order multivariate Markov chain model matrix to predict bus load, the first prediction vector of each bus load is taken Then the bus load at time t+1 is The predicted value is The corresponding value.

[0079] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the bus load prediction method based on a high-order multivariate Markov chain.

[0080] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the bus load prediction method based on a high-order multivariate Markov chain is implemented.

[0081] Compared with the prior art, the first aspect of the present invention has at least the following beneficial effects:

[0082] Based on the correlation between bus load sequence samples, the bus load sequence samples are converted into a high-order multivariate Markov chain model matrix. In order to simultaneously include the influencing factors of time and space into the prediction consideration range and make the prediction accuracy more accurate, the present invention adopts a high-order multivariate Markov chain to predict the bus load. A high-order multivariate Markov chain, that is, a multivariate Markov chain with a multivariate sequence of order n, refers to the state probability distribution of the jth sequence at time r+1, which depends on the probability distribution of all sequences at time r, r-1, ..., r-n+1. By determining the order of the high-order multivariate Markov chain model matrix, and estimating the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data, estimating the high-order weights between the columns of the high-order multivariate Markov chain model matrix, and substituting it into the modified high-order multivariate Markov chain model matrix, the bus load is predicted using the modified high-order multivariate Markov chain model matrix, and the load based on the adjacent bus and the historical time series is realized, and the hidden coupling relationship between the bus loads is effectively excavated, and the result of predicting the bus load is more in line with reality. The present invention establishes a high-order multivariate Markov chain model based on the historical load data of the bus to be predicted and the adjacent buses. When predicting the bus load, the influence of the time series and the geographic space coupling relationship on the bus load is considered at the same time, and the prediction accuracy is higher.

[0083] It can be understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0085] Figure 1 Flow chart of a bus load prediction method based on a high-order multivariate Markov chain according to an embodiment of the present invention;

[0086] Figure 2 A structural block diagram of a bus load prediction system based on a high-order multivariate Markov chain in an embodiment of the present invention. DETAILED DESCRIPTION

[0087] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0088] With the large-scale grid connection of clean energy, the electricity supply and demand situation faces new challenges, and it is necessary to improve the management and utilization of flexible resources on the load side. Accurate load forecasting can allow control personnel to understand the changing trend of the load in advance, and can also provide a basis for the evaluation of the demand side's response capabilities, and provide strong support for the grid to make scheduling decisions. Most traditional load forecasting methods only consider the impact of factors such as weather and date type on the load. However, due to factors such as geographical distribution, system operation mode, and load transfer, bus loads with close physical distances have a coupling relationship. At the same time, as a time series, the changing trend of the bus load is also affected by historical loads. Therefore, an embodiment of the present invention proposes a bus load forecasting method based on a high-order multivariate Markov chain, which takes into account multiple influencing factors of time and space, is more in line with reality, and can make the prediction results more accurate. Please refer to Figure 1 The bus load prediction method of the embodiment of the present invention specifically includes:

[0089] S1. Obtain daily load data of several adjacent buses from the power grid control and operation system, and establish a bus load sequence sample based on the number of load sampling points on each bus;

[0090] S2, based on the correlation between bus load sequence samples, convert the bus load sequence samples into a high-order multivariate Markov chain model matrix;

[0091] S3, determine the order of the high-order multivariate Markov chain model matrix;

[0092] S4, estimating the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data;

[0093] S5. Estimate the high-order weights between columns of the high-order multivariate Markov chain model matrix;

[0094] S6. Substitute the order of the high-order multivariate Markov chain model matrix, the transfer probability matrix and the high-order weights between columns into the modified high-order multivariate Markov chain model matrix, and use the modified high-order multivariate Markov chain model matrix to predict the bus load.

[0095] In a possible implementation, in step S1, the daily load data of several adjacent buses are obtained from the power grid control and operation system. When the bus load sequence sample is established in combination with the number of load sampling points of each bus, the daily load data of s adjacent buses are obtained from the control and operation system. The number of load sampling points of each bus is N. The bus load sequence sample is expressed as follows:

[0096]

[0097] Wherein, the subscript t represents the tth sampling point, and the superscript k represents the data of the kth bus.

[0098] In a possible implementation, the correlation between bus load sequence samples conforms to the following expression:

[0099]

[0100] In the formula, is the high-order weight between columns, representing the influence coefficient between each bus load sequence; n is the order of the high-order multivariate Markov chain model matrix;

[0101]

[0102] represents the h-step transition probability matrix from the state of the j-th sequence at time r-h+1 to the state of the k-th sequence at time r+1;

[0103] make in,

[0104] Then the mathematical expression of the high-order multivariate Markov chain model matrix is ​​as follows:

[0105]

[0106] In the formula,

[0107]

[0108] In a possible implementation, step S3 uses the Pearson correlation coefficient to calculate the correlation between two bus load series samples, and the calculation expression is as follows:

[0109]

[0110] Where, d = 0, 1, 2, ..., N-1, 1 ≤ i ≤ s, 1 ≤ j ≤ s;

[0111] When γ (ij)(d) When the absolute value is less than the threshold value, it is determined that the correlation between the two bus load sequence samples is weak. By calculating the correlation coefficients between different bus load sequence samples at different times, the correlation step d between the two bus load sequence samples is determined. The largest d between all bus load sequence samples is taken as the order n of the high-order multivariate Markov chain model matrix:

[0112]

[0113] In a possible implementation manner, the estimating the transition probability matrix of the high-order multivariate Markov chain model matrix by historical data in step S4 includes:

[0114] Divide the bus load into m intervals evenly and calculate the minimum bus load P in history min and the maximum bus load P max , then the load interval corresponding to the kth state is [(k-1)(P max -P min ) / m,k(P max -P min ) / m], where k = 1, 2, ... m. When there are enough historical sample data, it can be considered that the data distribution between sequences is the same as the distribution law of historical data, that is, the transition probability matrix can be obtained by counting the transition frequency between historical sample data sequences:

[0115]

[0116] in,

[0117]

[0118] In the formula, It is represented by the bus load sequence {x (j)} a state i j The bus load sequence {x (k)} a state i k The frequency of transfer.

[0119] In a possible implementation manner, the estimating of the high-order weights between columns of the high-order multivariate Markov chain model matrix in step S5 includes: by seeking The minimization method is used to solve the problem, where is the occurrence ratio of each state in the jth sequence;

[0120] The corresponding high-order weights between columns are obtained by solving the following optimization problem:

[0121]

[0122] In a possible implementation, the matrix expression of the modified high-order multivariate Markov chain model in step S6 is:

[0123]

[0124] because in When using the modified high-order multivariate Markov chain model matrix to predict bus load, the first prediction vector of each bus load is taken Then the bus load at time t+1 is The predicted value is The corresponding value.

[0125] See also Figure 2 Another embodiment of the present invention further provides a bus load forecasting system based on a high-order multivariate Markov chain, comprising:

[0126] The bus load sequence sample establishment module is used to obtain the daily load data of several adjacent buses from the power grid control and operation system, and establish the bus load sequence sample in combination with the number of load sampling points of each bus;

[0127] A high-order multivariate Markov chain model matrix conversion module is used to convert bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between bus load sequence samples;

[0128] A matrix order determination module is used to determine the order of the matrix of the high-order multivariate Markov chain model;

[0129] The transition probability matrix estimation module is used to estimate the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data;

[0130] Matrix column-to-column high-order weight estimation module, used to estimate the column-to-column high-order weights of the high-order multivariate Markov chain model matrix;

[0131] The high-order multivariate Markov chain model matrix correction and bus load prediction module is used to substitute the order of the high-order multivariate Markov chain model matrix, the transfer probability matrix and the high-order weights between columns into the corrected high-order multivariate Markov chain model matrix, and use the corrected high-order multivariate Markov chain model matrix to predict the bus load.

[0132] In a possible implementation, the bus load sequence sample establishment module obtains daily load data of s adjacent buses from the control operation system, the number of load sampling points of each bus is N, and the bus load sequence sample is expressed as follows:

[0133]

[0134] Wherein, the subscript t represents the tth sampling point, and the superscript k represents the data of the kth bus.

[0135] In a possible implementation, in the step of converting the bus load sequence samples into a high-order multivariate Markov chain model matrix by the high-order multivariate Markov chain model matrix conversion module based on the correlation between the bus load sequence samples, the correlation between the bus load sequence samples conforms to the following expression:

[0136]

[0137] In the formula, is the high-order weight between columns, representing the influence coefficient between each bus load sequence; n is the order of the high-order multivariate Markov chain model matrix;

[0138]

[0139] represents the h-step transition probability matrix from the state of the j-th sequence at time r-h+1 to the state of the k-th sequence at time r+1;

[0140] make in,

[0141] Then the mathematical expression of the high-order multivariate Markov chain model matrix is ​​as follows:

[0142]

[0143] In the formula,

[0144]

[0145] In a possible implementation manner, the matrix order determination module uses the Pearson correlation coefficient to calculate the correlation between two bus load sequence samples, and the calculation expression is as follows:

[0146]

[0147] Where, d = 0, 1, 2, ..., N-1, 1 ≤ i ≤ s, 1 ≤ j ≤ s;

[0148] When γ (ij) (d) When the absolute value is less than the threshold value, it is determined that the correlation between the two bus load sequence samples is weak. By calculating the correlation coefficients between different bus load sequence samples at different times, the correlation step d between the two bus load sequence samples is determined. The largest d between all bus load sequence samples is taken as the order n of the high-order multivariate Markov chain model matrix:

[0149]

[0150] In a possible implementation manner, the transfer probability matrix estimation module divides the bus load into m intervals on average, and calculates the minimum bus load P in the history. min and the maximum bus load P max , then the load interval corresponding to the kth state is [(k-1)(P max -P min ) / m,k(P max -P min ) / m], where k = 1, 2, ... m; by counting the transition frequencies between historical sample data sequences, the transition probability matrix is ​​obtained:

[0151]

[0152] in,

[0153]

[0154] In the formula, It is represented by the bus load sequence {x (j)} a state i j The bus load sequence {x (k)} a state i k The frequency of transfer.

[0155] In a possible implementation manner, the matrix column high-order weight estimation module seeks to make The minimization method is used to solve the problem, where is the occurrence ratio of each state of the jth sequence; the corresponding high-order weights between columns are obtained by solving the following optimization problem:

[0156]

[0157] In a possible implementation manner, the high-order multivariate Markov chain model matrix after the high-order multivariate Markov chain model matrix correction and the bus load forecasting module correction is expressed as:

[0158]

[0159] because in When using the modified high-order multivariate Markov chain model matrix to predict bus load, the first prediction vector of each bus load is taken Then the bus load at time t+1 is The predicted value is The corresponding value.

[0160] Another embodiment of the present invention further provides an electronic device, including a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the bus load prediction method based on a high-order multivariate Markov chain.

[0161] Another embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the bus load prediction method based on a high-order multivariate Markov chain is implemented.

[0162] The computer program includes computer program code, which may be in source code form, object code form, executable file or some intermediate form. The computer-readable storage medium may include: any entity or device, medium, USB flash drive, mobile hard disk, disk, optical disk, computer memory, read-only memory, random access memory, electric carrier signal, telecommunication signal and software distribution medium, etc. that can carry the computer program code. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals. For ease of explanation, the above content only shows the part related to the embodiment of the present invention. For specific technical details not disclosed, please refer to the method part of the embodiment of the present invention. The computer-readable storage medium is non-temporary and can be stored in a storage device formed by various electronic devices, which can realize the execution process recorded in the method of the embodiment of the present invention.

[0163] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0164] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0165] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0166] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A bus load forecasting method based on high-order multivariate Markov chain, characterized in that: include: Obtain daily load data of several buses from the power grid control and operation system, and establish bus load sequence samples based on the number of load sampling points on each bus; Based on the correlation between bus load sequence samples, the bus load sequence samples are converted into a high-order multivariate Markov chain model matrix; Determine the order of the matrix for a high-order multivariate Markov chain model; Estimate the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data; Estimate high-order weights between columns of a high-order multivariate Markov chain model matrix; The order of the high-order multivariate Markov chain model matrix, the transfer probability matrix and the high-order weights between columns are substituted into the modified high-order multivariate Markov chain model matrix, and the bus load is predicted using the modified high-order multivariate Markov chain model matrix.

2. The bus load forecasting method based on high-order multivariate Markov chain according to claim 1 is characterized in that: In the step of obtaining daily load data of several buses from the power grid control and operation system and establishing bus load sequence samples in combination with the number of load sampling points of each bus, daily load data of s buses are obtained from the control and operation system, the number of load sampling points of each bus is N, and the bus load sequence samples are expressed as follows: Wherein, the subscript t represents the tth sampling point, and the superscript k represents the data of the kth bus.

3. The bus load forecasting method based on high-order multivariate Markov chain according to claim 2 is characterized in that: In the step of converting the bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between the bus load sequence samples, the correlation between the bus load sequence samples conforms to the following expression: In the formula, is the high-order weight between columns, representing the influence coefficient between each bus load sequence; n is the order of the high-order multivariate Markov chain model matrix; represents the h-step transition probability matrix from the state of the j-th sequence at time r-h+1 to the state of the k-th sequence at time r+1; make in, Then the mathematical expression of the high-order multivariate Markov chain model matrix is ​​as follows: In the formula, 4. The bus load forecasting method based on high-order multivariate Markov chain according to claim 3 is characterized in that: The step of determining the order of the high-order multivariate Markov chain model matrix uses the Pearson correlation coefficient to calculate the correlation between the two bus load sequence samples. The calculation expression is as follows: Where, d = 0, 1, 2, ..., N-1, 1 ≤ i ≤ s, 1 ≤ j ≤ s; When γ (ij) (d) When the absolute value is less than the threshold value, it is determined that the correlation between the two bus load sequence samples is weak. By calculating the correlation coefficients between different bus load sequence samples at different times, the correlation step d between the two bus load sequence samples is determined. The largest d between all bus load sequence samples is taken as the order n of the high-order multivariate Markov chain model matrix:

5. The bus load forecasting method based on high-order multivariate Markov chain according to claim 4 is characterized in that: The step of estimating the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data comprises: Divide the bus load into m intervals evenly and calculate the minimum bus load P in history min and the maximum bus load P max , then the load interval corresponding to the kth state is [(k-1)(P max -P min ) / m,k(P max -P min ) / m], where k = 1, 2, ... m; by counting the transition frequencies between historical sample data sequences, the transition probability matrix is ​​obtained: in, In the formula, It is represented by the bus load sequence {x (j) } a state i j The bus load sequence {x (k) } a state i k The frequency of transfer.

6. The bus load forecasting method based on high-order multivariate Markov chain according to claim 5 is characterized in that: The step of estimating the high-order weights between columns of the high-order multivariate Markov chain model matrix comprises: By seeking The minimization method is used to solve the problem, where is the occurrence ratio of each state in the jth sequence; The corresponding high-order weights between columns are obtained by solving the following optimization problem:

7. The bus load forecasting method based on high-order multivariate Markov chain according to claim 6 is characterized in that: The step of substituting the order of the high-order multivariate Markov chain model matrix, the transition probability matrix and the high-order weights between columns into the modified high-order multivariate Markov chain model matrix, the modified high-order multivariate Markov chain model matrix expression is: because in When using the modified high-order multivariate Markov chain model matrix to predict bus load, the first prediction vector of each bus load is taken Then the bus load at time t+1 is The predicted value is The corresponding value.

8. A bus load forecasting system based on high-order multivariate Markov chain, characterized in that: include: The bus load sequence sample establishment module is used to obtain the daily load data of several buses from the power grid control and operation system, and establish the bus load sequence sample in combination with the number of sampling points of each bus load; A high-order multivariate Markov chain model matrix conversion module is used to convert bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between bus load sequence samples; A matrix order determination module is used to determine the order of the matrix of the high-order multivariate Markov chain model; The transition probability matrix estimation module is used to estimate the transition probability matrix of the high-order multivariate Markov chain model matrix through historical data; Matrix column-to-column high-order weight estimation module, used to estimate the column-to-column high-order weights of the high-order multivariate Markov chain model matrix; The high-order multivariate Markov chain model matrix correction and bus load prediction module is used to substitute the order of the high-order multivariate Markov chain model matrix, the transfer probability matrix and the high-order weights between columns into the corrected high-order multivariate Markov chain model matrix, and use the corrected high-order multivariate Markov chain model matrix to predict the bus load.

9. The bus load forecasting system based on high-order multivariate Markov chain according to claim 8 is characterized in that: The bus load sequence sample establishment module obtains the daily load data of s buses from the control operation system. The number of load sampling points of each bus is N. The bus load sequence sample is expressed as follows: Wherein, the subscript t represents the tth sampling point, and the superscript k represents the data of the kth bus.

10. The bus load forecasting system based on high-order multivariate Markov chain according to claim 9 is characterized in that: In the step of converting the bus load sequence samples into a high-order multivariate Markov chain model matrix based on the correlation between the bus load sequence samples, the correlation between the bus load sequence samples conforms to the following expression: In the formula, is the high-order weight between columns, representing the influence coefficient between each bus load sequence; n is the order of the high-order multivariate Markov chain model matrix; represents the h-step transition probability matrix from the state of the j-th sequence at time r-h+1 to the state of the k-th sequence at time r+1; make in, Then the mathematical expression of the high-order multivariate Markov chain model matrix is ​​as follows: In the formula, 11. The bus load forecasting system based on high-order multivariate Markov chain according to claim 10, characterized in that: The matrix order determination module uses the Pearson correlation coefficient to calculate the correlation between two bus load sequence samples. The calculation expression is as follows: Where, d = 0, 1, 2, ..., N-1, 1 ≤ i ≤ s, 1 ≤ j ≤ s; When γ (ij) (d) When the absolute value is less than the threshold value, it is determined that the correlation between the two bus load sequence samples is weak. By calculating the correlation coefficients between different bus load sequence samples at different times, the correlation step d between the two bus load sequence samples is determined. The largest d between all bus load sequence samples is taken as the order n of the high-order multivariate Markov chain model matrix:

12. The bus load forecasting system based on high-order multivariate Markov chain according to claim 11, characterized in that: The transfer probability matrix estimation module divides the bus load into m intervals on average and calculates the minimum bus load P in history. min and the maximum bus load P max , then the load interval corresponding to the kth state is [(k-1)(P max -P min ) / m,k(P max -P min ) / m], where k = 1, 2, ... m; by counting the transition frequencies between historical sample data sequences, the transition probability matrix is ​​obtained: in, In the formula, It is represented by the bus load sequence {x (j) } a state i j The bus load sequence {x (k) } a state i k The frequency of transfer.

13. The bus load forecasting system based on high-order multivariate Markov chain according to claim 12, characterized in that: The matrix column high-order weight estimation module seeks to make The minimization method is used to solve the problem, where is the occurrence ratio of each state of the jth sequence; the corresponding high-order weights between columns are obtained by solving the following optimization problem:

14. The bus load forecasting system based on high-order multivariate Markov chain according to claim 13, characterized in that: The matrix expression of the high-order multivariate Markov chain model after the high-order multivariate Markov chain model matrix correction and bus load forecasting module correction is: because in When using the modified high-order multivariate Markov chain model matrix to predict bus load, the first prediction vector of each bus load is taken Then the bus load at time t+1 is The predicted value is The corresponding value.

15. An electronic device, characterized in that: It comprises a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the bus load forecasting method based on a high-order multivariate Markov chain as claimed in any one of claims 1 to 7.

16. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the bus load prediction method based on a high-order multivariate Markov chain as described in any one of claims 1 to 7 is implemented.

Images