Method and device for obtaining linear frequency lowest point constraint based on support vector machine

Abstract

The present invention proposes a method and device for obtaining a linear frequency minimum point constraint based on a support vector machine, which belongs to the technical field of power system operation control. The method includes: establishing a power system frequency response model based on the inertia center; using Monte Carlo simulation to obtain data samples of the start and stop status of the power system units; performing numerical calculations on the frequency response model according to the samples to obtain the maximum frequency deviation of each sample under a preset disturbance; determining the maximum frequency deviation to obtain a label marking result of whether the sample corresponding to the maximum frequency deviation meets the frequency minimum point constraint; and using support vector machine technology to classify and learn the samples according to the label marking result to obtain the frequency minimum point constraint. The present invention takes into account a variety of primary frequency modulation equipment, and proposes a method for obtaining a linear frequency minimum point constraint based on support vector machine technology for dispatching and controlling a high-proportion new energy power system. The result takes into account both accuracy and reliability.

Description

Technical Field

[0001] The present invention belongs to the technical field of operation control of power systems, and in particular relates to a method and device for obtaining linear frequency minimum point constraints based on a support vector machine. Background Art

[0002] With the integration of large-scale renewable energy units into the grid, the inertia of the power system is gradually decreasing, and the issue of frequency security is becoming increasingly prominent. Frequency security has become a key bottleneck restricting the safe integration of large-scale renewable energy. It is urgent to fully utilize the primary frequency regulation resources of the power grid to improve the safe and economic operation of the power grid in response to the high proportion of renewable energy feed-in.

[0003] To ensure power system frequency security, frequency stability constraints need to be incorporated into the dispatch model to prevent rapid drops in the power system frequency from triggering underfrequency load shedding devices. Compared to other constraints in the dispatch model, frequency stability constraints involve high-order differential algebraic equations. The lowest point in the frequency drop process is a complex nonlinear function of the current system operating conditions. Nonlinear constraints can slow down the solution of the dispatch model and even render it unsolvable. This can adversely impact dispatch decisions and make it impossible to guarantee power system frequency security. Summary of the Invention

[0004] The purpose of this invention is to fill a gap in the existing technology by proposing a method and apparatus for obtaining linear frequency minimum point constraints based on support vector machines. This invention considers various primary frequency modulation devices and proposes a method for obtaining linear frequency minimum point constraints based on support vector machine technology for dispatch control of power systems with a high proportion of renewable energy. The results are both accurate and reliable and can be directly embedded in existing dispatch models to guide dispatch decisions for power systems.

[0005] The first embodiment of the present invention proposes a method for obtaining a linear frequency minimum point constraint based on a support vector machine, comprising:

[0006] According to the speed regulator models of each generator set and station in the power system, a power system frequency response model based on the inertia center is established;

[0007] Monte Carlo simulation is used to obtain data samples of the start and stop states of the power system units;

[0008] Performing numerical calculation on the frequency response model according to the sample to obtain the maximum frequency deviation of the sample under a preset disturbance;

[0009] By determining the maximum frequency deviation, a label marking result is obtained to determine whether the sample corresponding to the maximum frequency deviation meets the frequency minimum point constraint;

[0010] According to the label marking results, support vector machine technology is used to perform classification learning on the samples to obtain the frequency minimum point constraint.

[0011] In a specific embodiment of the present invention, the speed regulator model of each generator set and station includes:

[0012] The speed regulator model of thermal power unit is expressed as follows:

[0013]

[0014] Where i∈N G , N G is the collection of thermal power units; u i is the start and stop variable of unit i, which is a 0-1 variable; K mi is the mechanical power gain coefficient of unit i; R i is the droop control coefficient of unit i; T Ri is the inertia time constant of unit i; F Hi is the steam turbine power transfer coefficient of unit i; ΔP mi is the mechanical power increment of unit i; Δf is the inertia center frequency deviation;

[0015] The wind farm station speed regulator model is expressed as follows:

[0016]

[0017] Where j∈N W , N W R is the wind farm station collection; j is the droop control coefficient of station j; T j is the inertia time constant of station j; ΔP mj is the mechanical power increment of station j.

[0018] In a specific embodiment of the present invention, the method for calculating the inertial center frequency deviation is as follows:

[0019] Aggregating the frequency deviations of each generator set and station in the power system to a virtual central node, the inertia center frequency deviation expression is as follows:

[0020]

[0021] Among them, S i Indicates the capacity of unit or station i; H i Indicates the inertia of unit or station i; Δf i Indicates the frequency deviation of unit or station i.

[0022] In a specific embodiment of the present invention, the power system frequency response model expression is as follows:

[0023]

[0024] Where ΔP e is the preset load disturbance size; D is the load damping coefficient; s is the Laplace operator.

[0025] In a specific embodiment of the present invention, the frequency response model is numerically calculated based on the sample to obtain the maximum frequency deviation of the sample under a preset disturbance, which is expressed as follows:

[0026]

[0027] Where Δf max is the maximum frequency deviation vector, and its dimension is the same as the number of data samples of the start and stop status of the power system units; Indicates the maximum frequency deviation of the lth sample;

[0028] The data sample expression of the start and stop status of the power system unit is as follows:

[0029]

[0030] Where X represents the data sample matrix, the number of rows of matrix X is equal to the total number of thermal power units in the power system, and the number of columns is equal to the number of samples; Indicates the start / stop status of the i-th thermal power unit in the l-th sample, which is a 0-1 variable; Indicates that the unit is put into operation. Indicates that the unit is in shutdown state.

[0031] In a specific embodiment of the present invention, determining the maximum frequency deviation to obtain a label marking result of whether the sample corresponding to the maximum frequency deviation satisfies the frequency minimum point constraint includes:

[0032] Determine the maximum frequency deviation of each sample: If Then let the data label y of the lth sample sl =1, indicating that the sample meets the frequency minimum point constraint; otherwise, let y sl =-1, indicating that the sample does not meet the frequency minimum point constraint; is the maximum frequency deviation allowed;

[0033] After all samples are judged, the label vector is obtained:

[0034] Y=[y s1 y s2 y s3 … y sl …]

[0035] Here, the dimension of Y is the same as the number of samples.

[0036] In a specific embodiment of the present invention, the sample classification learning is performed using support vector machine technology based on the label marking result to obtain the frequency minimum point constraint, including:

[0037] 1) Construct a one-way soft margin support vector machine model, the expression is as follows:

[0038]

[0039]

[0040] Where ω represents the normal vector of the separating hyperplane, and its dimension is equal to the total number of thermal power units in the power system; b is the offset of the separating hyperplane to the origin; C is the regularization parameter, which represents the penalty coefficient for misclassified samples; m and n are the indices of different types of samples, respectively, where m corresponds to samples that meet the frequency minimum point constraint, and n corresponds to samples that do not meet the frequency minimum point constraint; ξ m is the slack variable corresponding to sample m, indicating the distance of the misclassified point from the classification boundary; S m Indicates the total number of samples that meet the frequency minimum point constraint; x sm and x sn are the characteristic vectors of different types of samples, representing the column elements of the data sample matrix X, and the dimension is equal to the total number of thermal power units in the power system; sm and y sn are the data labels of different types of samples, with values of ±1, corresponding to the elements of the label vector Y;

[0041] 2) Solve the one-way soft margin support vector machine model to obtain the separating hyperplane of the data sample matrix X and the label Y, and finally obtain the equivalent linear expression of the frequency minimum point constraint as follows:

[0042] ωu t +b≥0

[0043] ω=(ω1 ω2 ω3 … ω i …)

[0044]

[0045] Among them, u t is the unit start-stop vector corresponding to time period t, and its dimension is equal to the total number of thermal power units in the power system; is the start and stop variable of unit i in period t, which is a 0-1 variable; ω i is the element of the hyperplane normal vector corresponding to unit i.

[0046] The second embodiment of the present invention provides a linear frequency minimum point constraint acquisition device based on a support vector machine, comprising:

[0047] The power system frequency response model construction module is used to establish the power system frequency response model based on the inertia center according to the speed regulator models of each generator set and station in the power system;

[0048] A sample acquisition module, configured to obtain data samples of the start and stop states of the power system units using Monte Carlo simulation;

[0049] a sample maximum frequency deviation calculation module, configured to perform numerical calculation on the frequency response model according to the sample to obtain the maximum frequency deviation of the sample under a preset disturbance;

[0050] A sample label marking module is used to determine whether the sample corresponding to the maximum frequency deviation satisfies the frequency minimum point constraint by determining the maximum frequency deviation;

[0051] The frequency minimum constraint construction module is used to classify and learn the samples according to the label marking results using support vector machine technology to obtain the frequency minimum point constraint.

[0052] A third embodiment of the present invention provides an electronic device, including:

[0053] at least one processor; and a memory communicatively coupled to the at least one processor;

[0054] The memory stores instructions that can be executed by the at least one processor, and the instructions are configured to execute the above-mentioned method for obtaining the linear frequency minimum point constraint based on a support vector machine.

[0055] A fourth aspect of the present invention provides a computer-readable storage medium storing computer instructions for enabling the computer to execute the above-mentioned method for obtaining the linear frequency minimum point constraint based on a support vector machine.

[0056] The characteristics and beneficial effects of the present invention are:

[0057] The present invention proposes a method and device for obtaining a linear frequency minimum point constraint based on a support vector machine. First, a frequency response model based on the center of inertia is established; then, Monte Carlo simulation technology is used to obtain data samples of the start-up and shutdown states of the unit; the frequency response model is numerically solved based on the data samples of the start-up and shutdown states of the unit to obtain the maximum frequency deviation of the sample under a preset disturbance; the sample is labeled based on whether the maximum frequency deviation value of the sample exceeds a critical value; finally, support vector machine technology is used to classify and learn the samples according to the labels to obtain a linear classification boundary, i.e., the frequency minimum point constraint.

[0058] (1) Monte Carlo simulation technology is used to obtain data samples of the start and stop status of the unit to drive the calculation and solution of the support vector machine model.

[0059] (2) The frequency response model based on the inertial center is numerically solved based on the data samples of the unit's start-up and shutdown status to obtain the maximum frequency deviation of the sample under the preset disturbance, and the data sample is labeled according to whether the frequency deviation exceeds the critical value.

[0060] (3) Based on support vector machine technology, data samples are classified and learned according to sample labels to obtain linear classification boundaries and then obtain the linear frequency minimum point constraint.

[0061] (4) This invention adopts the concept of data-driven machine learning. Based on the data samples and their safety labels obtained by Monte Carlo simulation, classification learning is performed to obtain classification boundaries, thereby realizing the transformation of nonlinear constraints obtained by differential algebraic equations into equivalent linear constraints. The concept of one-way soft margin is introduced to ensure the conservatism of classification boundaries with respect to samples. Under the condition of acceptable accuracy loss, the safety, reliability, and economy of power grid operation are taken into account. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 The figure is an overall flow chart of a method for obtaining a linear frequency minimum point constraint based on a support vector machine according to an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The present invention proposes a method and device for obtaining a linear frequency minimum point constraint based on a support vector machine, which is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] The first embodiment of the present invention proposes a method for obtaining the linear frequency minimum point constraint based on a support vector machine. The overall process is as follows: Figure 1 As shown, the following steps are included:

[0065] 1) Based on the speed regulator model of each primary frequency regulation device in the power system, a frequency response model based on the inertia center considering multiple primary frequency regulation devices is established. The specific steps are as follows:

[0066] 1-1) Aggregate the frequency deviations of each unit and station in the power system to a virtual central node. The weight of each frequency aggregation is the capacity and inertia of the unit or station. The expression for the inertia center frequency deviation is as follows:

[0067]

[0068] Among them, S i represents the capacity of unit / station i; H i Indicates the inertia of unit / station i; Δf i represents the frequency deviation of unit / station i; Δf is the inertia center frequency deviation, which represents the frequency deviation of the power system.

[0069] Based on the inertia center frequency deviation, an aggregate frequency can be used to characterize the frequency change of the power system, and then the lowest point value of the power system frequency (corresponding to the maximum frequency deviation) can be obtained.

[0070] 1-2) Establish the governor model of each unit and station in the power system;

[0071] In this embodiment, for any conventional thermal power unit, H i Represents its inertia, and its speed regulator model can be expressed as:

[0072]

[0073] Where i∈N G , N G is the collection of thermal power units; u i is the start and stop variable of unit i, which is a 0-1 variable; K mi is the mechanical power gain coefficient of unit i; R i is the droop control coefficient of unit i; T Ri is the inertia time constant of unit i; F Hi is the steam turbine power transfer coefficient of unit i; ΔP mi is the mechanical power increment of unit i.

[0074] For any wind farm station, use H j Represents its inertia, and its speed regulator model can be expressed as:

[0075]

[0076] Where j∈N W , N W R is the wind farm station collection; j is the droop control coefficient of station j; T j is the inertia time constant of station j; ΔP mjis the mechanical power increment of station j.

[0077] 1-3) Based on the governor model established in step 1-2), a closed-loop transfer function is constructed between the power system frequency deviation and the current operating conditions and load disturbances, and the complex frequency domain expression of the power system frequency deviation is obtained:

[0078]

[0079] Where ΔP e is the preset load disturbance size, which is set in advance by the dispatcher; D represents the load damping coefficient; s is the Laplace operator.

[0080] Formula (4) is the power system frequency response model. By taking the inverse Laplace transform of Formula (4), we can get the time domain expression of the power system frequency deviation Δf(t), and then get the maximum frequency deviation Δf of the power system. max (corresponding to the lowest frequency point).

[0081] 2) Monte Carlo simulation is used to obtain data samples of the start and stop status of power system units:

[0082]

[0083] Where X represents the data sample matrix, and each column in the matrix represents a sample obtained through Monte Carlo simulation; represents the start / stop status of the i-th thermal power unit in the l-th sample, and the symbol s represents the data sample (sample). is a 0-1 variable, Indicates that the unit is put into operation. The number of rows of the matrix X is equal to the total number of thermal power units in the power system, and the number of columns is equal to the number of generated samples. In a specific embodiment of the present invention, the number of samples can be set to 100,000.

[0084] The Monte Carlo simulation calculation in this embodiment can be implemented in common mathematical software, such as MATLAB, etc.; it can also be implemented by calling a mathematical operation library using a basic programming language, such as Python, C, etc.

[0085] 3) Based on the data samples of the start-stop state of the unit obtained in step 2), the frequency response model formula (4) established in step 2) is numerically calculated to obtain the maximum frequency deviation of the sample under the preset disturbance, which is expressed as follows:

[0086]

[0087] Where Δf max is the maximum frequency deviation vector, corresponding to the matrix X, and the dimension of this vector is the same as the number of samples; Indicates the maximum frequency deviation of the lth sample.

[0088] It should be noted that the inverse Laplace transform of (4) can be solved by numerical calculation. The result is the discrete value of the frequency deviation in the time domain, and the calculation time scale can be set to 1ms. This numerical calculation process requires the use of mathematical software such as MATLAB; it can also be implemented by calling mathematical operation libraries using basic programming languages such as Python and C.

[0089] 4) Determine the maximum frequency deviation obtained in step 3) and label the sample based on the determination result;

[0090] In the embodiment of the present invention, according to the calculated maximum frequency deviation vector Δf max , label the data samples of the start and stop status of each unit and obtain the label vector:

[0091] Y=[y s1 y s2 y s3 … y sl …] (7)

[0092] Among them, Y is the label vector, and its dimension is the same as the number of samples; sl Represents the data label of the lth sample. If the maximum frequency deviation corresponding to the sample is less than the set threshold (i.e. ), then y sl =1, indicating that the sample meets the frequency minimum point constraint; otherwise, let y sl =-1, indicating that the sample does not meet the frequency minimum point constraint. The maximum frequency deviation allowed can be the frequency deviation value that triggers the low-frequency load reduction device.

[0093] 5) Based on the labeling results obtained in step 4), support vector machine technology is used to classify the samples according to the labels to obtain linear classification boundaries, and then obtain the frequency minimum point constraint; the specific steps are as follows:

[0094] 5-1) Construct a one-way soft margin support vector machine model;

[0095] It should be noted that a support vector machine is a data-driven classification learning method that can classify samples based on their corresponding data labels, thereby obtaining the classification boundaries of the data samples. It should be noted that conventional support vector machine models can only be solved if the data samples are linearly separable. In the high-dimensional space generated by the unit's start-up and shutdown state variables, the data sample X under the label vector Y is generally not linearly separable. Therefore, the conventional support vector machine model needs to be modified. In the embodiment of the present invention, one-way soft-margin support vector machine technology is used.

[0096] The one-way soft margin support vector machine model only introduces a slack variable ξ for a certain type of data, that is, it allows this type of data to be misclassified. For a data sample X and its corresponding label Y, when y sl =1 (corresponding to the sample that meets the frequency minimum point constraint), introduce a slack variable; when y sl = -1 (corresponding to samples that do not meet the frequency minimum point constraint), and no slack variables are introduced. That is, only frequency-safe samples are allowed to be misclassified, while frequency-unsafe samples are not allowed to be misclassified. This ensures that all samples that do not meet the frequency minimum point constraint are on one side of the separation boundary, so that the other side of the boundary is all samples that meet the frequency minimum point constraint. In this embodiment, the one-way soft margin support vector machine model can be expressed as:

[0097]

[0098] Where ω represents the normal vector of the separation boundary (hyperplane), and its dimension is equal to the total number of thermal power units in the power system; b is the offset of the separation hyperplane to the origin; C is the regularization parameter, which represents the penalty coefficient for misclassified samples and can be set to 10; m and n are the indices of different types of samples, respectively, m corresponds to samples that meet the frequency minimum point constraint, and n corresponds to samples that do not meet the frequency minimum point constraint; ξ m is the slack variable corresponding to sample m, indicating the distance of the misclassified point from the classification boundary; S m Indicates the total number of samples that meet the frequency minimum point constraint; x sm and x sn are the characteristic vectors of different types of samples, representing the column elements of the data sample matrix X, whose dimension is equal to the total number of thermal power units in the power system; sm and y sn They are data labels for different types of samples, with a value of ±1, corresponding to the elements of the label vector Y.

[0099] It should be pointed out that the solution of model formula (8) requires the use of mathematical software, such as MATLAB, etc.; it can also be implemented by calling mathematical operation libraries using basic programming languages, such as Python, C, etc.

[0100] 5-2) Solving the one-way soft margin support vector machine model constructed in step 5-1) can obtain the separation boundary (hyperplane) between the data sample X and the label Y, and then the equivalent linear expression of the frequency minimum point constraint can be obtained as follows:

[0101] ωu t +b≥0 (9)

[0102] ω=(ω1 ω2 ω3 … ω i …) (10)

[0103]

[0104] Among them, ω is the normal vector of the separation boundary (hyperplane), the dimension of which is the same as the number of thermal power units in the power system, and is solved by support vector machine technology; b is the offset b from the separation hyperplane to the origin, which is also solved by support vector machine technology; u t is the unit start-stop vector corresponding to time period t, and its dimension is equal to the total number of thermal power units in the power system; is the start and stop variable of unit i in period t, which is a 0-1 variable; ω i is the element of the hyperplane normal vector, corresponding to unit i.

[0105] Equations (9)-(11) are the obtained linear frequency minimum point constraints, which can be directly embedded in the unit commitment model of the power system. Since the constraints are linear, they do not reduce the solution speed of the unit commitment model after the constraints are embedded. Therefore, a unit commitment model that considers the frequency minimum point constraint can be obtained and can be solved efficiently and quickly, thereby guiding the power system dispatcher to make dispatch decisions.

[0106] To implement the above embodiment, a second embodiment of the present invention proposes a linear frequency minimum point constraint acquisition device based on a support vector machine, comprising:

[0107] The power system frequency response model construction module is used to establish the power system frequency response model based on the inertia center according to the speed regulator models of each generator set and station in the power system;

[0108] A sample acquisition module, configured to obtain data samples of the start and stop states of the power system units using Monte Carlo simulation;

[0109] a sample maximum frequency deviation calculation module, configured to perform numerical calculation on the frequency response model according to the sample to obtain the maximum frequency deviation of the sample under a preset disturbance;

[0110] A sample label marking module is used to determine whether the sample corresponding to the maximum frequency deviation satisfies the frequency minimum point constraint by determining the maximum frequency deviation;

[0111] The frequency minimum constraint construction module is used to classify and learn the samples according to the label marking results using support vector machine technology to obtain the frequency minimum point constraint.

[0112] To implement the above embodiment, a third aspect of the present invention provides an electronic device, including:

[0113] at least one processor; and a memory communicatively coupled to the at least one processor;

[0114] The memory stores instructions that can be executed by the at least one processor, and the instructions are configured to execute the above-mentioned method for obtaining the linear frequency minimum point constraint based on a support vector machine.

[0115] To implement the above embodiment, the fourth aspect of the present invention proposes a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable the computer to execute the above-mentioned method for obtaining the linear frequency minimum point constraint based on a support vector machine.

[0116] It should be noted that the computer-readable medium mentioned above in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0117] The computer-readable medium may be included in the electronic device or may exist independently without being incorporated into the electronic device. The computer-readable medium carries one or more programs. When executed by the electronic device, the one or more programs cause the electronic device to execute the method for obtaining the linear frequency minimum point constraint based on a support vector machine according to the above embodiment.

[0118] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0119] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0120] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0121] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0122] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing it in a suitable manner if necessary, and then storing it in a computer memory.

[0123] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0124] Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above-mentioned embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0125] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0126] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A method for obtaining linear frequency minimum point constraint based on support vector machine, characterized in that: include: According to the speed regulator models of each generator set and station in the power system, a power system frequency response model based on the inertia center is established; Monte Carlo simulation is used to obtain data samples of the start and stop states of the power system units; Performing numerical calculation on the frequency response model according to the sample to obtain the maximum frequency deviation of the sample under a preset disturbance; By determining the maximum frequency deviation, a label marking result is obtained to determine whether the sample corresponding to the maximum frequency deviation meets the frequency minimum point constraint; According to the label marking results, the samples are classified and learned using support vector machine technology to obtain the frequency minimum point constraint; The speed regulator models of the generator sets and stations include: The speed regulator model of thermal power unit is expressed as follows: Where i∈N G , N G is the collection of thermal power units; u i is the start and stop variable of unit i, which is a 0-1 variable; K mi is the mechanical power gain coefficient of unit i; R i is the droop control coefficient of unit i; T Ri is the inertia time constant of unit i; F Hi is the steam turbine power transfer coefficient of unit i; ΔP mi is the mechanical power increment of unit i; Δf is the inertia center frequency deviation; s is the Laplace operator; The wind farm station speed regulator model is expressed as follows: Where j∈N W , N W R is the wind farm station collection; j is the droop control coefficient of station j; T j is the inertia time constant of station j; ΔP mj is the mechanical power increment of station j; The power system frequency response model expression is as follows: Where ΔP e is the preset load disturbance size; D is the load damping coefficient; s is the Laplace operator; H i represents the inertia of unit or station i; The method of performing classification learning on the samples based on the label marking results using support vector machine technology to obtain the frequency minimum point constraint includes: 1) Construct a one-way soft margin support vector machine model, the expression is as follows: Where ω represents the normal vector of the separating hyperplane, and its dimension is equal to the total number of thermal power units in the power system; b is the offset of the separating hyperplane to the origin; C is the regularization parameter, which represents the penalty coefficient for misclassified samples; m and n are the indices of different types of samples, respectively, where m corresponds to samples that meet the frequency minimum point constraint, and n corresponds to samples that do not meet the frequency minimum point constraint; ξ m is the slack variable corresponding to sample m, indicating the distance of the misclassified point from the classification boundary; S m Indicates the total number of samples that meet the frequency minimum point constraint; x sm and x sn are the characteristic vectors of different types of samples, representing the column elements of the data sample matrix X, and the dimension is equal to the total number of thermal power units in the power system; sm and y sn are the data labels of different types of samples, with values of ±1, corresponding to the elements of the label vector Y; 2) Solve the one-way soft margin support vector machine model to obtain the separating hyperplane of the data sample matrix X and the label Y, and finally obtain the equivalent linear expression of the frequency minimum point constraint as follows: oh t +b≥0 ω=(ω1 ω2 ω3…ω i …) Among them, u t is the unit start-stop vector corresponding to time period t, and its dimension is equal to the total number of thermal power units in the power system; is the start and stop variable of unit i in period t, which is a 0-1 variable; ω i is the element of the hyperplane normal vector corresponding to unit i.

2. The method according to claim 1, characterized in that The calculation method of the inertial center frequency deviation is as follows: Aggregating the frequency deviations of each generator set and station in the power system to a virtual central node, the inertia center frequency deviation expression is as follows: Among them, S i Indicates the capacity of unit or station i; H i Indicates the inertia of unit or station i; Δf i Indicates the frequency deviation of unit or station i.

3. The method according to claim 1, characterized in that The frequency response model is numerically calculated based on the sample to obtain the maximum frequency deviation of the sample under a preset disturbance, which is expressed as follows: Where Δf max is the maximum frequency deviation vector, and its dimension is the same as the number of data samples of the start and stop status of the power system units; Indicates the maximum frequency deviation of the lth sample; The data sample expression of the start and stop status of the power system unit is as follows: Where X represents the data sample matrix, the number of rows of matrix X is equal to the total number of thermal power units in the power system, and the number of columns is equal to the number of samples; Indicates the start / stop status of the i-th thermal power unit in the l-th sample, which is a 0-1 variable; Indicates that the unit is put into operation. Indicates that the unit is in shutdown state.

4. The method according to claim 3, characterized in that The determining of the maximum frequency deviation to obtain a label marking result of whether the sample corresponding to the maximum frequency deviation satisfies the frequency minimum point constraint includes: Determine the maximum frequency deviation of each sample: If Then let the data label y of the lth sample sl =1, indicating that the sample meets the frequency minimum point constraint; otherwise, let y sl =-1, indicating that the sample does not meet the frequency minimum point constraint; is the maximum frequency deviation allowed; After all samples are judged, the label vector is obtained: And=[and s1 and s2 and s3 …and sl …] Here, the dimension of Y is the same as the number of samples.

5. A linear frequency minimum point constraint acquisition device based on a support vector machine according to the method of claim 1, characterized in that: include: The power system frequency response model construction module is used to establish the power system frequency response model based on the inertia center according to the speed regulator models of each generator set and station in the power system; A sample acquisition module, configured to obtain data samples of the start and stop states of the power system units using Monte Carlo simulation; a sample maximum frequency deviation calculation module, configured to perform numerical calculation on the frequency response model according to the sample to obtain the maximum frequency deviation of the sample under a preset disturbance; A sample label marking module is used to determine whether the sample corresponding to the maximum frequency deviation satisfies the frequency minimum point constraint by determining the maximum frequency deviation; The frequency minimum constraint construction module is used to classify and learn the samples according to the label marking results using support vector machine technology to obtain the frequency minimum point constraint.

6. An electronic device, characterized in that: include: at least one processor; and, a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are configured to execute the method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method according to any one of claims 1 to 4.

Images