Probabilistic power flow calculation method and device considering offshore wind power and frequency

Abstract

The application discloses a kind of offshore wind power and frequency's probabilistic power flow calculation method and device, wherein the method comprises: respectively constructing the active-frequency static characteristic of unit and the active-frequency static characteristic of load;Static frequency characteristic is considered to build power flow model based on the active-frequency static characteristic of unit and the active-frequency static characteristic of load;According to the power flow model considering static frequency characteristic, decoupling linear power flow model considering static frequency characteristic is derived;The decoupling linear power flow model is analyzed based on UT transformation and NATAF transformation to obtain the probabilistic power flow analysis result.The application is based on the active-frequency static characteristic of unit and the active-frequency static characteristic of load to build the power flow model considering static frequency characteristic, and decoupling linear power flow model considering static frequency characteristic is derived, can effectively improve the calculation efficiency of probabilistic power flow, and then can improve the efficiency of probabilistic power flow analysis in random scene.

Description

Technical Field

[0001] The present invention relates to the technical field of power system steady-state analysis, and in particular to a method and device for calculating a probabilistic power flow taking into account offshore wind power and frequency. Background Art

[0002] Compared to onshore wind power, offshore wind energy resources offer 20% to 40% higher energy efficiency than onshore wind farms. They also offer advantages such as minimal land occupation, high wind speeds, minimal dust, high power generation, stable operation, and zero dust emissions. They also reduce wear and tear on turbines, extending their service life and making them suitable for large-scale development.

[0003] From 2017 to 2021, global offshore wind power capacity installations showed an upward trend. As more and more countries around the world develop offshore wind power, the growth rate of global offshore wind power installations has significantly accelerated. According to data from GWEC's "Global Offshore Wind Report 2022," global offshore wind power installations reached 21.1 GW in 2021, a 207.89% increase from 2020, the largest increase on record. The cumulative installed capacity of offshore wind power in 2021 reached 56 GW, a 55.56% increase from 2020. However, offshore wind power output exhibits significant randomness and volatility. This randomness can lead to significant power imbalances in the system. Active power imbalances in the power system can cause frequency fluctuations across the entire grid. Therefore, it is necessary to consider static frequency characteristics in steady-state analysis of power systems involving stochastic scenarios.

[0004] At present, the existing probabilistic power flow calculation methods usually use nonlinear AC power flow models to consider static frequency characteristics, which limits their application in large-scale calculation scenarios. As the penetration rate of wind power continues to increase, it is difficult to ensure the calculation accuracy of the probabilistic power flow when the range of random injection amount varies greatly. Summary of the Invention

[0005] The present invention provides a probabilistic tidal current calculation method and device taking into account offshore wind power and frequency, so as to solve the technical problem that the existing probabilistic tidal current calculation method is difficult to ensure the calculation accuracy of the probabilistic tidal current when the random injection amount varies within a large range.

[0006] An embodiment of the present invention provides a probabilistic power flow calculation method considering offshore wind power and frequency, including:

[0007] Construct the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load respectively;

[0008] Building a power flow model that considers static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load;

[0009] According to the power flow model considering static frequency characteristics, a decoupled linear power flow model considering static frequency characteristics is derived;

[0010] A probabilistic power flow analysis is performed on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain a probabilistic power flow analysis result.

[0011] Furthermore, the respectively constructing of the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load includes:

[0012] According to the active power output, active power-frequency static characteristic coefficient and no-load frequency of the unit, the active power-frequency static characteristic of the unit is constructed:

[0013] P Gi =K Gi (f 0i -f)

[0014] Among them, P Gi is the active power output of the unit at node i, K Gi is the active power-frequency static characteristic coefficient of the unit at node i. The corresponding empirical value range of the steam turbine generator set is 20~33.3 (pu), f 0i is the no-load frequency of the unit at node i, and f is the system frequency;

[0015] According to the load active power, the load rated active power and the load active power-frequency static characteristic coefficient, the load active power-frequency static characteristic is constructed:

[0016] P Li =P LNi +K Li (ff N )

[0017] Among them, P Li is the active power of the load on node i, P LNi is the rated active power of the load on node i, K Li is the active power-frequency static characteristic coefficient of the load, and the empirical value corresponding to the comprehensive load is 1.5 (pu).

[0018] Furthermore, the constructing of a power flow model considering static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load includes:

[0019] According to the unified iterative method, a power flow model considering the static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load:

[0020] ΔP i (V,θ,f)=P Gi (f)-PLi (f)-P i (V,θ)=0

[0021] ΔQ i (V,θ)=Q Gi -Q Li -Q i (V,θ)=0

[0022] Where ΔP i , ΔQ i is the power imbalance of node i, P i , Q i is the power equation of node i, V, θ are the voltage amplitude and phase angle, Q Gi is the reactive power output of the unit at node i, Q Li is the reactive power of the load on node i.

[0023] Furthermore, the decoupled linear power flow model considering static frequency characteristics is derived based on the power flow model considering static frequency characteristics, including:

[0024] By solving the power flow model considering the static frequency characteristics, a Newton-Raphson power flow model considering the static frequency characteristics is constructed. The Newton-Raphson power flow model is:

[0025]

[0026] Among them, J 11 、J 12 、J 21 and J 22 The element expressions of are the same as those of the conventional Newton-Raphson power flow method, where K and F are related to the static frequency characteristics;

[0027] A decoupled linear power flow model considering static frequency characteristics is derived based on the Newton-Raphson power flow model considering static frequency characteristics.

[0028] Furthermore, the Newton-Raphson power flow model considering the static frequency characteristics is deduced to obtain a decoupled linear power flow model considering the static frequency characteristics, including:

[0029] Based on the Newton-Raphson power flow model that takes into account static frequency characteristics, a linear expression of the active power equation and a decoupled linear expression of the reactive power equation are calculated according to the node power equation and the elements of the node admittance matrix of the AC power flow, and a linear power flow model with voltage amplitude and phase angle decoupling is determined based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation;

[0030] The linear power flow model is transcribed and the static frequency characteristics are introduced to obtain a decoupled linear power flow model considering the static frequency characteristics.

[0031] Furthermore, based on the Newton-Raphson power flow model considering the static frequency characteristics, a linear expression of the active power equation and a decoupled linear expression of the reactive power equation are calculated according to the elements of the node power equation and the node admittance matrix of the AC power flow, and a linear power flow model for voltage amplitude and phase angle decoupling is determined according to the linear expression of the active power equation and the decoupled linear expression of the reactive power equation, including:

[0032] The node power equation of the AC power flow is constructed based on the real and imaginary parts of the elements in the node admittance matrix, the voltage amplitude of the node, and the voltage phase angle difference of the node:

[0033]

[0034]

[0035] Among them, G ij 、B ij Represent the real and imaginary parts of the elements in the i-th row and j-th column of the node admittance matrix, V i , V j are the voltage amplitudes of nodes i and j, θ ij =θ i -θ j is the voltage phase angle difference between node i and node j;

[0036] The elements of the node admittance matrix are:

[0037]

[0038] According to the node admittance matrix, an approximate node admittance matrix without considering the parallel admittance is defined. The elements of the approximate node admittance matrix are:

[0039]

[0040] Among them, Y ij is the i-th row and j-th column element of the node admittance matrix Y, y ij is the admittance of the line between node i and node j, y ii is the parallel admittance of node i, Y′ ij is the i-th row and j-th column element of the approximate nodal admittance matrix Y′;

[0041] According to the node power equation of the AC power flow and the elements of the node admittance matrix, the linear expression of the active power equation and the decoupled linear expression of the reactive power equation are calculated, and the linear power flow model with voltage amplitude and phase angle decoupling is determined based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation.

[0042] Furthermore, the decoupled linear power flow model considering the static frequency characteristics is:

[0043]

[0044] Furthermore, the probabilistic power flow analysis is performed on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain the probabilistic power flow analysis results, including:

[0045] Determine the random input variables and the distribution type function of the probabilistic power flow model, and solve the correlation coefficient matrix of random variables in the standard normal distribution domain based on NATAF transformation;

[0046] Performing Cholesky decomposition on the correlation coefficient matrix to obtain a decomposition matrix;

[0047] According to the inverse function of the cumulative distribution function, the cumulative distribution function of the standard normal distribution random variable and the decomposition matrix, the sample matrix of the random variable in the original distribution domain is calculated;

[0048] The sample matrix is ​​brought into the decoupled linear power flow model for calculation to obtain the mean and covariance matrix of the probability power flow variables.

[0049] One embodiment of the present invention provides a probabilistic power flow calculation device that considers offshore wind power and frequency, including:

[0050] Frequency static characteristic construction module, used to construct the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load respectively;

[0051] A first power flow model construction module is used to construct a power flow model that considers static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load;

[0052] A second power flow model construction module is used to derive a decoupled linear power flow model considering static frequency characteristics based on the power flow model considering static frequency characteristics;

[0053] The probabilistic power flow analysis module is used to perform probabilistic power flow analysis on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain probabilistic power flow analysis results.

[0054] One embodiment of the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the probabilistic power flow calculation method considering offshore wind power and frequency as described above.

[0055] The embodiment of the present invention constructs a power flow model that considers static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load, and derives a decoupled linear power flow model that considers static frequency characteristics. It can effectively speed up the calculation speed of a single deterministic power flow while ensuring accuracy, and can also ensure the calculation accuracy of the probabilistic power flow when the range of random injection amount changes is large, thereby effectively improving the calculation efficiency of the probabilistic power flow, and further improving the efficiency of probabilistic power flow analysis in random scenarios.

[0056] Furthermore, the embodiment of the present invention adopts the UT algorithm based on NATAF transformation, combined with the characteristics of the decoupled linear power flow model, which can further improve the calculation efficiency of the probabilistic power flow considering the static frequency characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of a probabilistic power flow calculation method considering offshore wind power and frequency, provided by an embodiment of the present invention;

[0058] Figure 2 Schematic diagram of the mean and standard deviation absolute error of frequencies under three combinations provided by an embodiment of the present invention;

[0059] Figure 3 Schematic diagram of the absolute error of the variance of the power flow results under the S1 combination provided in an embodiment of the present invention;

[0060] Figure 4 Schematic diagram of the absolute error of the variance of the power flow results under the S2 combination provided in an embodiment of the present invention;

[0061] Figure 5 Schematic diagram of the absolute error of the variance of the power flow results under the S3 combination provided in an embodiment of the present invention;

[0062] Figure 6 3 is a schematic structural diagram of a probabilistic power flow calculation device taking into account offshore wind power and frequency, provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0064] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0065] In the description of this application, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.

[0066] See also Figure 1 One embodiment of the present invention provides a probabilistic power flow calculation method considering offshore wind power and frequency, comprising:

[0067] S1. Construct the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load respectively;

[0068] The present invention ignores the weak coupling relationship between reactive power and frequency and considers the static frequency characteristics. The Jingtian frequency characteristics mainly include the active static characteristics, which describe the frequency changes after the system enters a stable state, as well as the relationship between the active power and frequency of the units and loads.

[0069] S2. Based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load, a power flow model considering the static frequency characteristics is constructed;

[0070] S3. Based on the power flow model considering static frequency characteristics, a decoupled linear power flow model considering static frequency characteristics is derived;

[0071] S4. Perform probabilistic power flow analysis on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain the probabilistic power flow analysis results.

[0072] The embodiment of the present invention constructs a power flow model that considers static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load, and derives a decoupled linear power flow model that considers static frequency characteristics. It can effectively speed up the calculation speed of a single deterministic power flow while ensuring accuracy, and can also ensure the calculation accuracy of the probabilistic power flow when the range of random injection amount changes is large, thereby effectively improving the calculation efficiency of the probabilistic power flow, and further improving the efficiency of probabilistic power flow analysis in random scenarios.

[0073] Furthermore, the embodiment of the present invention adopts the UT algorithm based on NATAF transformation, combined with the characteristics of the decoupled linear power flow model, which can further improve the calculation efficiency of the probabilistic power flow considering the static frequency characteristics.

[0074] In one embodiment, step S1, respectively constructing the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load, may further include the following sub-steps:

[0075] S11. Construct the active power-frequency static characteristic of the unit based on the unit's active power output, active power-frequency static characteristic coefficient and no-load frequency:

[0076] P Gi =K Gi (f 0i -f) (1)

[0077] Among them, P Gi is the active power output of the unit at node i, K Gi is the active power-frequency static characteristic coefficient of the unit at node i. The corresponding empirical value range of the steam turbine generator set is 20~33.3 (pu), f 0i is the no-load frequency of the unit at node i, and f is the system frequency;

[0078] S12. Construct the load active power-frequency static characteristic according to the load active power, the load rated active power and the load active power-frequency static characteristic coefficient:

[0079] P Li =P LNi +K Li (ff N ) (2)

[0080] Among them, P Li is the active power of the load on node i, P LNi is the rated active power of the load on node i, K Li is the active power-frequency static characteristic coefficient of the load, and the empirical value corresponding to the comprehensive load is 1.5 (pu).

[0081] In one embodiment, a power flow model considering static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load, including:

[0082] According to the unified iterative method, a power flow model considering static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load:

[0083]

[0084] Where ΔP i , ΔQi is the power imbalance of node i, P i , Q i is the power equation of node i, V, θ are the voltage amplitude and phase angle, Q Gi is the reactive power output of the unit at node i, Q Li is the reactive power of the load on node i.

[0085] In the embodiment of the present invention, the system includes n nodes and m PQ nodes. Formula (3) of the power flow model considering the static frequency characteristics includes n active power equations and m reactive power equations.

[0086] In one embodiment, step S3, deriving a decoupled linear power flow model considering static frequency characteristics based on the power flow model considering static frequency characteristics, further includes the following sub-steps:

[0087] S31. By solving the power flow model considering the static frequency characteristics, a Newton-Raphson power flow model considering the static frequency characteristics is constructed. The Newton-Raphson power flow model is:

[0088]

[0089] Among them, J 11 、J 12 、J 21 and J 22 The element expressions of are the same as those of the conventional Newton-Raphson power flow method, where K and F are related to the static frequency characteristics;

[0090] In the embodiment of the present invention, since the weak coupling relationship between reactive power and frequency is ignored, the elements of F are 0, and the elements of K can be expressed by formula (5).

[0091]

[0092] By iteratively solving equation (4) until convergence, the desired quantities θ, V, and f can be obtained.

[0093] S32. Based on the Newton-Raphson power flow model considering the static frequency characteristics, a decoupled linear power flow model considering the static frequency characteristics is derived.

[0094] In the embodiment of the present invention, starting from the formula (4) of the Newton-Raphson power flow model, a linear power flow model is derived that takes into account the static frequency characteristics and realizes the decoupling of the voltage amplitude and phase angle, so that its accuracy is close to the formula (4) and the speed is close to the DC power flow.

[0095] In one embodiment, step S32, deriving a decoupled linear power flow model considering static frequency characteristics based on a Newton-Raphson power flow model considering static frequency characteristics, further includes the following sub-steps:

[0096] S321. Based on a Newton-Raphson power flow model that takes into account static frequency characteristics, a linear expression for the active power equation and a decoupled linear expression for the reactive power equation are calculated according to the elements of the node power equation and the node admittance matrix of the AC power flow, and a linear power flow model for voltage amplitude and phase angle decoupling is determined based on the linear expression for the active power equation and the decoupled linear expression for the reactive power equation.

[0097] S322. After transcribing the linear power flow model, static frequency characteristics are introduced to obtain a decoupled linear power flow model considering static frequency characteristics.

[0098] In one embodiment, S321, based on a Newton-Raphson power flow model that considers static frequency characteristics, calculates a linear expression of the active power equation and a decoupled linear expression of the reactive power equation according to the node power equation and the elements of the node admittance matrix of the AC power flow, and determines a linear power flow model for voltage amplitude and phase angle decoupling based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation, further comprising the following sub-steps:

[0099] S3211. Construct the node power equation of the AC power flow based on the real and imaginary parts of the elements in the node admittance matrix, the node voltage amplitude, and the node voltage phase angle difference:

[0100]

[0101] Among them, G ij 、B ij Represent the real and imaginary parts of the elements in the i-th row and j-th column of the node admittance matrix, V i , V j are the voltage amplitudes of nodes i and j, θ ij =θ i -θ j is the voltage phase angle difference between node i and node j;

[0102] The elements of the node admittance matrix are:

[0103]

[0104] S3212. Define an approximate nodal admittance matrix without considering parallel admittance based on the nodal admittance matrix. The elements of the approximate nodal admittance matrix are:

[0105]

[0106] Among them, Y ij is the i-th row and j-th column element of the node admittance matrix Y, y ij is the admittance of the line between node i and node j, y ii is the parallel admittance of node i, Y′ ijis the i-th row and j-th column element of the approximate nodal admittance matrix Y′.

[0107] S3213. Calculate the linear expression of the active power equation and the decoupled linear expression of the reactive power equation based on the node power equation and the elements of the node admittance matrix of the AC power flow, and determine the linear power flow model with voltage amplitude and phase angle decoupling based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation.

[0108] In the embodiment of the present invention, formula (7) is substituted into the active power equation of formula (6) to obtain formula (9).

[0109]

[0110] At the same time, the embodiment of the present invention makes some assumptions that are consistent with the physical properties of the power system, as shown in formula (10).

[0111] V i 2 ≈V i ,cosθ ij ≈1,sinθ ij ≈θ i -θ j (10)

[0112] Therefore, the nonlinear term g in equation (9) ij V i (V i -V j cosθ ij ) is linearly approximated to obtain formula (11).

[0113] g ij V i (V i -V j cosθ ij )=g ij (V i -V j ) (11)

[0114] Where, ΔV i Usually higher than V i An order of magnitude smaller, the quadratic term ΔV generated in the middle i 2 and ΔV i ΔV j Can be ignored.

[0115]

[0116] The embodiment of the present invention decouples the voltage amplitude and phase angle by substituting equation (11) into equation (9), and obtains the linear expression (12) of the active power equation.

[0117] Similarly, a similar linear approximation is made to the reactive power equation of Equation (6), and taking into account g ii Generally much smaller than b ii , assuming G′ ij ≈G ij , the decoupled linear expression of the reactive power equation can be obtained (13).

[0118]

[0119] In summary, the linear power flow model with decoupling of voltage amplitude and phase angle can be expressed by Equation (14).

[0120]

[0121] It should be noted that the decoupling in the embodiment of the present invention refers to the decoupling of the voltage amplitude and voltage phase angle of the nonlinear terms in the node power equation, thereby obtaining a linear power flow model that does not require iterative solution.

[0122] In an embodiment of the present invention, the linear power flow model is transcribed and then the static frequency characteristics are introduced to obtain a decoupled linear power flow model considering the static frequency characteristics, including:

[0123] Equation (14) can be rewritten as Equation (15). The subscripts vθ, pv, and pq correspond to the equilibrium node, PV node, and PQ node, respectively, and the nodes are numbered in this order. This does not affect the versatility of the model but is only for the convenience of analysis.

[0124]

[0125] in,

[0126]

[0127] In the embodiment of the present invention, the static frequency characteristic is introduced into equation (15), and θ, V, and f are taken as the quantities to be determined, thereby obtaining a decoupled linear power flow model considering the static frequency characteristic, as shown in equation (17).

[0128]

[0129] Since the weak coupling relationship between reactive power and frequency is ignored, the elements of F are 0, and the other coefficient matrices can be expressed by equation (18).

[0130]

[0131] The embodiment of the present invention processes the static frequency characteristics to avoid the increase of the nonlinearity of the model, so that the coefficient matrix [HNK; MLF] in equation (17) remains unchanged. By solving equation (17), the desired quantity θ can be directly obtained. pv ,θpq 、V pq and f.

[0132] In one embodiment, step S4, performing a probabilistic power flow analysis on the decoupled linear power flow model based on the UT transformation and the NATAF transformation to obtain a probabilistic power flow analysis result, further includes the following sub-steps:

[0133] S41. Determine the random input variables of the probabilistic power flow model and the distribution type function they obey, and solve the correlation coefficient matrix of the random variables in the standard normal distribution domain according to the NATAF transformation;

[0134] In this embodiment of the invention, only wind power and load rate models are considered. The key factor affecting wind farm capacity is wind speed, which is highly random. This embodiment of the invention uses Weibull and Lognormal distributions to describe the marginal distribution characteristics of wind speed.

[0135] In the embodiment of the present invention, the wind speed is set to obey Weibull distribution and Lognormal distribution, and their probability density functions are respectively expressed as formula (19) and formula (20).

[0136]

[0137]

[0138] Where v is the wind speed, k and c represent the shape parameter and scale parameter, respectively, and a and b represent the logarithmic mean and standard deviation, respectively.

[0139] The wind speed-power conversion model of the embodiment of the present invention is shown in formula (21).

[0140]

[0141] Among them, P WT is the output power of a single fan, P r is the rated output power of a single fan, v, v cut-in 、v rated and v cut-out They are wind speed, cut-in, rated and cut-out wind speeds.

[0142] The load in the embodiment of the present invention obeys the Normal distribution, and the probability density function is expressed by formula (22):

[0143]

[0144] Where, P L is the load active power, μ L represents the mean active power of the load, σ LThe power factor of the embodiment of the present invention remains unchanged, and all reactive loads are determined according to the corresponding active load and power factor.

[0145] In the embodiment of the present invention, the idea of ​​NATAF transformation is as follows: Assume that X represents a random variable that obeys an arbitrary distribution, and Z represents a random variable that obeys a standard Gaussian distribution. Formula (23) is the expression of X in terms of Z:

[0146]

[0147] Among them, Φ(Z i ) is Z i The cumulative distribution function of . For X i The inverse of the cumulative distribution function. The correlation matrix of variable X is shown in formula (24).

[0148]

[0149] in, Represents the asymmetric distribution X i X j Pearson correlation coefficient.

[0150] From formula (25), we can get the X Correlation of an arbitrarily distributed random variable X.

[0151]

[0152] Among them, U is an independent standard Gaussian distributed random variable, L is the lower triangular decomposition matrix, and the matrix can be solved according to Cholesky decomposition.

[0153] The idea of ​​UT conversion in the embodiment of the present invention is as follows:

[0154] Assume that X represents an n-dimensional random input variable, represents its mean, P XX is the covariance. Y represents the mean of the output variable, P YY is the covariance matrix of the output variable. The following is the probability power flow calculation process based on the UT algorithm:

[0155] Step 1: Obtain the sample points on the probability distribution according to formula (26):

[0156]

[0157] Where n is the dimension of the variable. yes The k-th column element of . According to Satisfy P XX =AA ΤCan be obtained

[0158] Step 2: According to formula (27), the weight value corresponding to the sample can be determined:

[0159]

[0160] Step 3: Substitute the sample into the deterministic power flow model formula (28):

[0161] Y k =f(X k ) (28)

[0162] Step 4: According to formula (29), the covariance and mean of the probability power flow output Y can be obtained:

[0163]

[0164] The features of the UT algorithm in the embodiment of the present invention include:

[0165] (1) The number of samples is 2n+1. Since only 2n+1 sample points are selected from the original input probability distribution, only 2n+1 deterministic power flow calculations are required to obtain the probability result.

[0166] (2) The UT method usually determines the sample points based on the mean and covariance of the original input variables, so the obtained sample points carry the Pearson correlation information between random variables.

[0167] In the embodiment of the present invention, the random input variable X of the probability power flow model and the distribution type function it obeys are determined, the dimension of X is n, and the correlation coefficient matrix R of the random variable under the standard normal distribution domain is solved according to the NATAF transformation. Z .

[0168] S42, performing Cholesky decomposition on the correlation coefficient matrix to obtain a decomposition matrix;

[0169] In the embodiment of the present invention, R Z Use Cholesky decomposition to obtain the decomposition matrix L, L satisfies R Z =LL T .

[0170] S43. Calculate a sample matrix of the original distribution domain random variable based on the inverse function of the cumulative distribution function, the cumulative distribution function of the standard normal distribution random variable, and the decomposition matrix;

[0171] In the embodiment of the present invention, the weight value W is calculated according to formulas (30) to (33): 0 、W k and W k+n, in order to obtain the Gaussian distribution Z. Where W 0 Represents the initial weight value, and n represents the matrix dimension, that is, the number of random input variables.

[0172]

[0173]

[0174]

[0175]

[0176] In the embodiment of the present invention, according to formula (26), sampling is performed on an independent n-dimensional standard normal distribution to obtain a sample matrix G = [G0, G1, ..., G 2n+1 ]. Since G is an independent sample of standard normal distribution variables, the sampling formula can be simplified to (34).

[0177]

[0178] Step 5: According to formula (35), the sample is converted into the original distribution domain to obtain the sample matrix X of the random variable X in the original distribution domain sam .

[0179] X sam =F -1 [Φ(LG)] (35)

[0180] Among them F -1 is the inverse of the cumulative distribution function of X, and Φ is the cumulative distribution function of a standard normally distributed random variable.

[0181] S44. Bring the sample matrix into the decoupled linear power flow model for calculation to obtain the mean and covariance matrix of the probabilistic power flow variables.

[0182] In the embodiment of the present invention, the samples of X are brought into the deterministic power flow model for calculation, and then the mean and covariance matrix of the output variables are obtained according to formula (29).

[0183] In one embodiment, a probabilistic power flow calculation considering offshore wind power and frequency is analyzed using a specific example.

[0184] The test system of the embodiment of the present invention is obtained by modifying the IEEE-57 test system provided in MATPOWER

[32] . In the test, two probabilistic methods are combined with four deterministic power flow models to compare the probabilistic power flow results of different combinations to verify the effectiveness, accuracy and speed of the decoupled linear model based on untraceable transformation proposed in the embodiment of the present invention.

[0185] The four models are as follows: the Newton-Raphson power flow model without considering the static frequency characteristics is denoted as M2, and the Newton-Raphson, DC, and decoupled linear power flow models considering the static frequency characteristics are denoted as M1, M3, and M4, respectively.

[0186] In the power flow model considering static frequency characteristics, the KG of the unit at the original system balance node is set to 25 p.u., the KG of other units is set to 20 p.u., and the load KL is set to 1.5 pu. The test environment is provided by MATPOWER

[32] on a laptop with an AMD Ryzen 75800H CPU and 16GB of memory.

[0187] The IEEE-57 test system was modified as follows: Nodes 55, 56, and 57 were connected to wind farms, numbered WF1, WF2, and WF3, respectively. Each wind farm had 30 wind turbines, each with a rated power of 1 MW. The wind speed, cut-in, rated, and cut-out wind speeds were set at 2 m / s, 15 m / s, and 25 m / s, respectively.

[0188] All active loads follow a normal distribution, with a mean equal to the original value and a standard deviation equal to 0.05 times the mean. The wind speeds at wind farms 55 and 56 follow a Weibull distribution, with scale and shape parameters of 10.7 and 3.97, respectively. The wind speed at wind farm 57 follows a lognormal distribution, with a logarithmic mean of 2.08 and a logarithmic standard deviation of 0.24. The Pearson correlation coefficients between the various random input variables are shown in Table 1.

[0189] Table 1 Pearson correlation coefficient

[0190] Correlation coefficient load wind speed load 0.6 0.2 wind speed 0.2 0.9

[0191] (1) Deterministic power flow model evaluation:

[0192] The absolute errors of the power flow results for each deterministic model were calculated by varying the load value to 0.85, 1, and 1.15 times the original load level, using the results from M1 as a reference. The results are shown in Table 2. The frequency results represent the absolute error, while the phase angle, amplitude, and branch loss results are the average of the absolute errors for each node or branch. A dash ("----") indicates that the corresponding model could not calculate that result.

[0193] Table 2 Comparison of absolute errors of deterministic power flow results under different load multiples

[0194]

[0195] Refer to Table 2. When the same load multiple is used in Table 2, the frequency results for M3 and M4 are identical. This is because both M3 and M4 are linear models, and the voltage amplitude and phase angle are decoupled during the solution of M4. Furthermore, the two models consider the same frequency mechanism. The above statistical results indicate that when the load multiple is any of the above values, the deterministic model (M4) proposed in this embodiment of the present invention is more accurate than other deterministic models. It can also be seen that the errors are largest when frequency is not considered, indicating that whether or not frequency is considered significantly affects the accuracy of the final power flow results.

[0196] Refer to Tables 3-4. Tables 3 and 4 show the mean and variance deviations of the probabilistic power flow results for M2, M3, and M4, respectively, calculated using the MCS, using the M1 results as a reference. Frequency results represent the average or variance of the absolute deviations of 10,000 samples. Phase angle, amplitude, and branch loss results are the average of the absolute deviations of each node or branch. A dash ("----") indicates that the corresponding model cannot calculate that result.

[0197] Table 3 Comparison of mean deviation of MCS probability power flow results for M2, M3, and M4

[0198] Average Deviation f(Hz) θ(°) V(pu) Ploss(MW) M2 ---- 7.4611 0.0088 0.1704 M3 0.0370 1.1175 ---- ---- M4 0.0370 0.1250 0.0024 0.0060

[0199] Table 4 Comparison of variance deviation of MSC probability power flow results for M2, M3, and M4

[0200] Variance bias f(Hz) θ(°) V(pu) Ploss(MW) M2 ---- 1.7356 <![CDATA[3.22×10 -6 ]]> <![CDATA[2.20×10 -3 ]]> M3 <![CDATA[2.94×10 -6 ]]> <![CDATA[2.95×10 -4 ]]> ---- ---- M4 <![CDATA[2.94×10 -6 ]]> <![CDATA[2.97×10 -5 ]]> 3.88 x 10 -7 ]] <![CDATA[2.83×10 -7 ]]>

[0201] Please continue to refer to Tables 3 and 4. It can be seen from Tables 3 and 4 that the deviations of M4 and M3 are the same in terms of frequency; in terms of voltage phase angle, M4<M3<M2; and the deviation rules for voltage amplitude and branch loss are all: M4<M2. M2 cannot calculate the system frequency, and the average deviation of each item is the largest. This is because the unbalanced power generated by wind power fluctuations is only borne by the balancing node, resulting in a large difference in output between multiple units. This also shows the importance of taking frequency into account in power flow calculations. The above statistical results show that under the MCS method probabilistic power flow, the deterministic model (M4) proposed in the embodiment of the present invention is more accurate than other deterministic models.

[0202] (2) Comparative test of the probabilistic tidal flow results of the MCS method and the UT method based on the Newton model and the decoupled linear model. In the following results, S1 represents the tidal flow result error of UT-M1 with MCS-M1 as a reference; S2 represents the tidal flow result error of MCS-M4 with MCS-M1 as a reference; S3 represents the tidal flow result error of UT-M4 with MCS-M1 as a reference.

[0203] See also Figure 2, are the mean absolute error and standard deviation absolute error of the frequencies under the combinations of S1, S2, and S3 provided in the embodiment of the present invention.

[0204] Depend on Figure 2 It can be seen that the calculated frequency mean absolute error and standard deviation absolute error under the S1 combination are 3.67×10⁻¹⁴ and 7.12×10⁻¹⁴, respectively, with error rates of 7.53×10⁻¹⁴% and 0.41%, respectively. The calculated frequency mean absolute error and standard deviation absolute error under the S2 combination are 3.70×10⁻¹⁴ and 1.68×10⁻¹⁴, respectively, with error rates of 7.51×10⁻¹⁴% and 2.01%, respectively. The calculated frequency mean absolute error and standard deviation absolute error under the S3 combination are 3.74×10⁻¹⁴ and 1.50×10⁻¹⁴, respectively, with error rates of 7.58×10⁻¹⁴% and 1.61%, respectively. The error rates under all three combinations are well within the normal error ranges reported in the literature, fully demonstrating the effectiveness of the untraceable transformation-based decoupled linear probability model proposed in this embodiment of the present invention. A comparison of the S1 and S2 data shows that the S2 combination has a larger error, indicating that the deterministic model has a greater impact on accuracy. The error of the S3 combination is the largest, indicating that the combination of the deterministic model and the probabilistic model proposed in the embodiment of the present invention will further affect the accuracy.

[0205] See also Figure 3-5 ,from Figures 3 to 4 The absolute errors of the node voltage amplitude and phase angle and the variance of each branch loss under the S1, S2, and S3 combinations are shown respectively.

[0206] Depend on Figure 2 、 Figure 3 and Figure 4 It can be seen that the absolute errors in the power flow results for all three combinations are small, falling within the normal error range, further verifying the accuracy of the decoupled linear probability model proposed in this embodiment of the present invention. Analysis of any of the data sets S1, S2, and S3 reveals that the voltage amplitude and phase angle errors at nodes 41 and 56 are both excessively large, leading to an excessively large error in the branch loss between these two nodes. S3 has the largest error value, indicating that combining deterministic and probabilistic models has a greater impact on accuracy.

[0207] (3) Speed ​​test of deterministic and probabilistic power flows:

[0208] In this embodiment of the present invention, 10,000 samples are used for the MCS probability method, and 91 samples are used for the UT probability method. The first two rows of Table 5 record the time consumption of the deterministic power flow calculation. The third and fourth rows record the time consumption of the probabilistic power flow calculation using the four deterministic models combined with the UT method. The fifth and sixth rows record the time consumption of the probabilistic power flow calculation using the four deterministic models combined with the MCS method.

[0209] Table 5 Comparison of time consumption for power flow calculation

[0210] Model M1 M2 M3 M4 Time(s) 0.0434 0.0045 0.0313 0.0320 Model UT-M1 UT-M2 UT-M3 UT-M4 Time(s) 0.6926 0.9633 0.5518 0.5825 Model MCS-M1 MCS-M2 MCS-M3 MCS-M4 Time(s) 25.6606 30.8437 10.2160 14.7347

[0211] Refer to Table 5. The second row shows that, considering frequency effects, the decoupled linear model proposed in this embodiment of the present invention takes a time approaching that of the DC method M3 and less than that of M1. This demonstrates the superior speed of the model (M4) proposed in this embodiment of the present invention. The fourth and fifth rows show that, under the same deterministic model, the probabilistic power flow calculation using the probabilistic UT method takes significantly less time than the MCS method. This is because the MCS model can only complete the probabilistic power flow analysis through 10,000 deterministic power flow calculations, while the UT model only requires 91, resulting in a significant reduction in time. Table 5 shows that UT-M4 takes 0.5825 seconds, while MCS-M4 takes 14.7347 seconds. This indicates that the UT method is almost 25 times faster than the MCS method. Furthermore, MCS-M1 takes 25.6606 seconds. This indicates that the UT-M4 probabilistic power flow calculation is approximately 44 times faster than MCS-M1. The speed-up due to the probabilistic model accounts for 95%, demonstrating that the probabilistic model proposed in this embodiment of the present invention achieves a superior speed-up compared to the deterministic model. In summary, in the probabilistic power flow analysis considering static frequency characteristics, the decoupled linear model based on unscented transformation proposed in the embodiment of the present invention has better speed.

[0212] The implementation of the embodiments of the present invention has the following effects:

[0213] The embodiment of the present invention constructs a power flow model that considers static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load, and derives a decoupled linear power flow model that considers static frequency characteristics. It can effectively speed up the calculation speed of a single deterministic power flow while ensuring accuracy, and can also ensure the calculation accuracy of the probabilistic power flow when the range of random injection amount changes is large, thereby effectively improving the calculation efficiency of the probabilistic power flow, and further improving the efficiency of probabilistic power flow analysis in random scenarios.

[0214] Furthermore, the embodiment of the present invention adopts the UT algorithm based on NATAF transformation, combined with the characteristics of the decoupled linear power flow model, which can further improve the calculation efficiency of the probabilistic power flow considering the static frequency characteristics.

[0215] See also Figure 6 Based on the same inventive concept as the above embodiment, one embodiment of the present invention provides a probabilistic power flow calculation device considering offshore wind power and frequency, comprising:

[0216] The frequency static characteristic construction module 10 is used to respectively construct the active power-frequency static characteristic of the unit and the active power-frequency static characteristic of the load;

[0217] A first power flow model construction module 20 is used to construct a power flow model that considers static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load;

[0218] The second power flow model construction module 30 is used to derive a decoupled linear power flow model considering static frequency characteristics based on the power flow model considering static frequency characteristics;

[0219] The probabilistic power flow analysis module 40 is used to perform probabilistic power flow analysis on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain probabilistic power flow analysis results.

[0220] In one embodiment, the frequency static characteristic building module 10 is further used to:

[0221] According to the active power output, active power-frequency static characteristic coefficient and no-load frequency of the unit, the active power-frequency static characteristic of the unit is constructed:

[0222] P Gi =K Gi (f 0i -f)

[0223] Among them, P Gi is the active power output of the unit at node i, K Gi is the active power-frequency static characteristic coefficient of the unit at node i. The corresponding empirical value range of the steam turbine generator set is 20~33.3 (pu), f 0i is the no-load frequency of the unit at node i, and f is the system frequency;

[0224] According to the load active power, the load rated active power and the load active power-frequency static characteristic coefficient, the load active power-frequency static characteristic is constructed:

[0225] P Li =P LNi +K Li (ff N )

[0226] Among them, P Li is the active power of the load on node i, P LNi is the rated active power of the load on node i, K Li is the active power-frequency static characteristic coefficient of the load, and the empirical value corresponding to the comprehensive load is 1.5 (pu).

[0227] In one embodiment, the first power flow model building module 20 is further configured to:

[0228] According to the unified iterative method, a power flow model considering static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load:

[0229] ΔPi (V,θ,f)=P Gi (f)-P Li (f)-P i (V,θ)=0

[0230] ΔQ i (V,θ)=Q Gi -Q Li -Q i (V,θ)=0

[0231] Where ΔP i , ΔQ i is the power imbalance of node i, P i , Q i is the power equation of node i, V, θ are the voltage amplitude and phase angle, Q Gi is the reactive power output of the unit at node i, Q Li is the reactive power of the load on node i.

[0232] In one embodiment, the second power flow model building module 30 is further configured to:

[0233] By solving the power flow model considering the static frequency characteristics, a Newton-Raphson power flow model considering the static frequency characteristics is constructed. The Newton-Raphson power flow model is:

[0234]

[0235] Among them, J 11 、J 12 、J 21 and J 22 The element expressions of are the same as those of the conventional Newton-Raphson power flow method, where K and F are related to the static frequency characteristics;

[0236] Based on the Newton-Raphson power flow model considering the static frequency characteristics, a decoupled linear power flow model considering the static frequency characteristics is obtained.

[0237] In one embodiment, the second power flow model building module 30 is further configured to:

[0238] Based on the Newton-Raphson power flow model that takes into account static frequency characteristics, the linear expression of the active power equation and the decoupled linear expression of the reactive power equation are calculated according to the elements of the nodal power equation and the nodal admittance matrix of the AC power flow. The linear power flow model with voltage amplitude and phase angle decoupling is determined based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation.

[0239] The linear power flow model is transcribed and the static frequency characteristics are introduced to obtain a decoupled linear power flow model considering the static frequency characteristics.

[0240] In one embodiment, based on the Newton-Raphson power flow model considering static frequency characteristics, the linear expression of the active power equation and the decoupled linear expression of the reactive power equation are calculated according to the node power equation of the alternating current power flow and the elements of the node admittance matrix, and the linear power flow model of decoupled voltage amplitude and phase angle is determined according to the linear expression of the active power equation and the decoupled linear expression of the reactive power equation, including:

[0241] The node power equation of the alternating current power flow is constructed according to the real part and the imaginary part of the elements in the node admittance matrix, the voltage amplitude of the node, and the voltage phase angle difference of the node:

[0242]

[0243]

[0244] Wherein, G ij , B ij represent the real part and the imaginary part of the element in the i-th row and the j-th column of the node admittance matrix, V i , V j are the voltage amplitudes of the nodes i and j, θ ij = θ i - θ j is the voltage phase angle difference of the nodes i and j;

[0245] The elements of the node admittance matrix are:

[0246]

[0247] According to the definition of the node admittance matrix, the approximate node admittance matrix not considering the shunt admittance is constructed, and the elements of the approximate node admittance matrix are:

[0248]

[0249] Wherein, Y ij is the element in the i-th row and the j-th column of the node admittance matrix Y, y ij is the admittance of the line between the nodes i and j, y ii is the shunt admittance of the node i, Y′ ij is the element in the i-th row and the j-th column of the approximate node admittance matrix Y′;

[0250] The linear expression of the active power equation and the decoupled linear expression of the reactive power equation are calculated according to the node power equation of the alternating current power flow and the elements of the node admittance matrix, and the linear power flow model of decoupled voltage amplitude and phase angle is determined according to the linear expression of the active power equation and the decoupled linear expression of the reactive power equation.

[0251] In one embodiment, the decoupled linear power flow model considering static frequency characteristics is:

[0252]

[0253] In one embodiment, the probabilistic power flow analysis module 40 is further configured to:

[0254] Determine the random input variables and the distribution type function of the probabilistic power flow model, and solve the correlation coefficient matrix of random variables in the standard normal distribution domain based on NATAF transformation;

[0255] Perform Cholesky decomposition on the correlation coefficient matrix to obtain the decomposition matrix;

[0256] According to the inverse function of the cumulative distribution function, the cumulative distribution function of the standard normal distribution random variable and the decomposition matrix, the sample matrix of the random variable in the original distribution domain is calculated;

[0257] The sample matrix is ​​brought into the decoupled linear power flow model for calculation to obtain the mean and covariance matrix of the probabilistic power flow variables.

[0258] One embodiment of the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the probabilistic power flow calculation method considering offshore wind power and frequency as described above.

[0259] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A probabilistic power flow calculation method considering offshore wind power and frequency, characterized in that: include: Construct the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load respectively; The active power-frequency static characteristic of the unit and the active power-frequency static characteristic of the load are constructed separately, including: constructing the active power-frequency static characteristic of the unit according to the active power output of the unit, the active power-frequency static characteristic coefficient and the no-load frequency: in, P Gi For nodes i The active power output of the upper unit, K Gi For nodes i The active power-frequency static characteristic coefficient of the upper unit, the corresponding empirical value range of the steam turbine generator set is 20~33.3 (pu), f 0i For nodes i The no-load frequency of the upper unit, f is the system frequency; According to the load active power, the load rated active power and the load active power-frequency static characteristic coefficient, the load active power-frequency static characteristic is constructed: in, P Li For nodes i Active power of the load, P LNi For nodes i Rated active power of the upper load, K Li is the active power-frequency static characteristic coefficient of the load, and the empirical value corresponding to the comprehensive load is 1.5 (pu); A power flow model considering static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load; the power flow model considering static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load, including: According to the unified iterative method, a power flow model considering the static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load: Among them, Δ P i , Δ Q i For nodes i The power imbalance, P i 、 Q i For nodes i The power equation, V 、 θ are the voltage amplitude and phase angle, Q Gi For nodes i The reactive power output of the upper unit, Q Li For nodes i Reactive power of upper load; According to the power flow model considering the static frequency characteristics, a decoupled linear power flow model considering the static frequency characteristics is derived; according to the power flow model considering the static frequency characteristics, a decoupled linear power flow model considering the static frequency characteristics is derived, including: By solving the power flow model considering the static frequency characteristics, a Newton-Raphson power flow model considering the static frequency characteristics is constructed. The Newton-Raphson power flow model is: in, J 11 、 J 12 、 J 21 and J 22 The element expressions of are the same as those of the conventional Newton-Raphson power flow method, K 、 F Related to static frequency characteristics; Based on the Newton-Raphson power flow model that takes into account static frequency characteristics, a linear expression of the active power equation and a decoupled linear expression of the reactive power equation are calculated according to the node power equation and the elements of the node admittance matrix of the AC power flow, and a linear power flow model with voltage amplitude and phase angle decoupling is determined based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation; After transcribing the linear power flow model, static frequency characteristics are introduced to obtain a decoupled linear power flow model considering static frequency characteristics; A probabilistic power flow analysis is performed on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain a probabilistic power flow analysis result.

2. The probabilistic power flow calculation method considering offshore wind power and frequency according to claim 1, characterized in that: The Newton-Raphson power flow model considering the static frequency characteristics is based on the node power equation of the AC power flow and the elements of the node admittance matrix, and the linear expression of the active power equation and the decoupled linear expression of the reactive power equation are calculated, and the linear power flow model with voltage amplitude and phase angle decoupling is determined based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation, including: The node power equation of the AC power flow is constructed based on the real and imaginary parts of the elements in the node admittance matrix, the voltage amplitude of the node, and the voltage phase angle difference of the node: in, G ij 、 B ij Represents the node admittance matrix i Row, No. j the real and imaginary parts of the column elements, V i , V j Node i 、 j The voltage amplitude, θ ij =θ i -θ j For nodes i and nodes j The voltage phase angle difference; The elements of the node admittance matrix are: According to the node admittance matrix, an approximate node admittance matrix without considering the parallel admittance is defined. The elements of the approximate node admittance matrix are: in, Y ij is the node admittance matrix Y No. i Row, No. j Column elements, y ij is a node i and nodes j The admittance of the line between y ii is a node i The parallel admittance of is the approximate nodal admittance matrix No. i Row, No. j Column elements; According to the node power equation of the AC power flow and the elements of the node admittance matrix, the linear expression of the active power equation and the decoupled linear expression of the reactive power equation are calculated, and the linear power flow model with voltage amplitude and phase angle decoupling is determined based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation.

3. The probabilistic power flow calculation method considering offshore wind power and frequency according to claim 1, characterized in that: The decoupled linear power flow model considering the static frequency characteristics is: 。 4. The probabilistic power flow calculation method considering offshore wind power and frequency according to claim 1, characterized in that: The probabilistic power flow analysis is performed on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain a probabilistic power flow analysis result, including: Determine the random input variables and the distribution type function of the probabilistic power flow model, and solve the correlation coefficient matrix of random variables in the standard normal distribution domain based on NATAF transformation; Performing Cholesky decomposition on the correlation coefficient matrix to obtain a decomposition matrix; According to the inverse function of the cumulative distribution function, the cumulative distribution function of the standard normal distribution random variable and the decomposition matrix, the sample matrix of the random variable in the original distribution domain is calculated; The sample matrix is ​​brought into the decoupled linear power flow model for calculation to obtain the mean and covariance matrix of the probability power flow variables.

5. A probabilistic power flow calculation device considering offshore wind power and frequency, characterized in that: include: Frequency static characteristic construction module, used to construct the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load respectively; The active power-frequency static characteristic of the unit and the active power-frequency static characteristic of the load are constructed separately, including: constructing the active power-frequency static characteristic of the unit according to the active power output of the unit, the active power-frequency static characteristic coefficient and the no-load frequency: in, P Gi For nodes i The active power output of the upper unit, K Gi For nodes i The active power-frequency static characteristic coefficient of the upper unit, the corresponding empirical value range of the steam turbine generator set is 20~33.3 (pu), f 0i For nodes i The no-load frequency of the upper unit, f is the system frequency; According to the load active power, the load rated active power and the load active power-frequency static characteristic coefficient, the load active power-frequency static characteristic is constructed: in, P Li For nodes i Active power of the load, P LNi For nodes i Rated active power of the upper load, K Li is the active power-frequency static characteristic coefficient of the load, and the empirical value corresponding to the comprehensive load is 1.5 (pu); A first power flow model construction module is configured to construct a power flow model that takes into account static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load. The power flow model that takes into account static frequency characteristics based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load comprises: According to the unified iterative method, a power flow model considering the static frequency characteristics is constructed based on the active power-frequency static characteristics of the unit and the active power-frequency static characteristics of the load: Among them, Δ P i , Δ Q i For nodes i The power imbalance, P i 、 Q i For nodes i The power equation, V 、 θ are the voltage amplitude and phase angle, Q Gi For nodes i The reactive power output of the upper unit, Q Li For nodes i Reactive power of the upper load; The second power flow model construction module is configured to derive a decoupled linear power flow model that considers static frequency characteristics based on the power flow model that considers static frequency characteristics; the derivation of the decoupled linear power flow model that considers static frequency characteristics based on the power flow model that considers static frequency characteristics includes: By solving the power flow model considering the static frequency characteristics, a Newton-Raphson power flow model considering the static frequency characteristics is constructed. The Newton-Raphson power flow model is: in, J 11 、 J 12 、 J 21 and J 22 The element expressions of are the same as those of the conventional Newton-Raphson power flow method, K 、 F Related to static frequency characteristics; Based on the Newton-Raphson power flow model that takes into account static frequency characteristics, a linear expression of the active power equation and a decoupled linear expression of the reactive power equation are calculated according to the node power equation and the elements of the node admittance matrix of the AC power flow, and a linear power flow model with voltage amplitude and phase angle decoupling is determined based on the linear expression of the active power equation and the decoupled linear expression of the reactive power equation; After transcribing the linear power flow model, static frequency characteristics are introduced to obtain a decoupled linear power flow model considering static frequency characteristics; The probabilistic power flow analysis module is used to perform probabilistic power flow analysis on the decoupled linear power flow model based on UT transformation and NATAF transformation to obtain probabilistic power flow analysis results.

6. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the probabilistic power flow calculation method considering offshore wind power and frequency according to any one of claims 1 to 4.

Images