Probabilistic Analysis Method, System, Device and Storage Medium for Available Transmission Capacity

Abstract

The present invention discloses a method, system, device and storage medium for probabilistic analysis of available transfer capability, which includes obtaining probability calculation samples of random input variables; substituting the probability calculation samples of random input variables into an available transfer capability calculation model based on a radial basis neural network to solve the available transfer capability corresponding to each probability calculation sample of random input variables; and obtaining the probability distribution of the available transfer capability based on the available transfer capability and outputting it. The present invention can obtain an available transfer capability calculation model based on a radial basis neural network in an offline state. This model has the characteristics of accurate results and high calculation efficiency, and can be used for online probabilistic calculation and analysis of available transfer capability.

Description

Technical Field

[0001] The present invention belongs to the field of power system operation, and particularly relates to a method, system, device and storage medium for probabilistic analysis of available transfer capability based on radial basis neural network. Background Art

[0002] Available transfer capability is an important indicator for the economic and secure operation of power systems. Due to the large-scale integration of wind power and the diversification of user electricity consumption behaviors, the calculation of available transfer capability must consider the randomness brought by factors such as wind speed and load. In the operation analysis of power systems, since the calculation of available transfer capability is complex and time-consuming, especially in random scenarios, it involves repeated solutions for available transfer capability, and its solution efficiency is difficult to meet the requirements of real-time calculation. Summary of the Invention

[0003] The purpose of the present invention is to provide a method, system, device and storage medium for probabilistic analysis of available transfer capability based on radial basis neural network, so as to overcome the difficulties in the prior art that the calculation of available transfer capability in random scenarios is complex and difficult to meet the requirements of real-time calculation.

[0004] To achieve the above object, the present invention adopts the following technical solutions:

[0005] A method for probabilistic analysis of available transfer capability, the method comprising:

[0006] Obtaining probability calculation samples of random input variables;

[0007] Substituting the probability calculation samples of the random input variables into an available transfer capability calculation model based on radial basis neural network to solve the available transfer capability corresponding to each probability calculation sample of the random input variables; wherein the available transfer capability calculation model based on radial basis neural network is established through historical data of probability uncertainty sources and an available transfer capability calculation model;

[0008] Obtaining the probability distribution of available transfer capability according to the available transfer capability and outputting it.

[0009] Further, the method for obtaining the probability calculation samples of the random input variables is: converting the probability calculation samples of the probability uncertainty sources through a model of power injection from the probability uncertainty sources to grid nodes;

[0010] The method for obtaining the random input variables is: through a model of power injection from the probability uncertainty sources to grid nodes, obtaining the node power injection objects affected by the probability uncertainty sources, and taking the node power injection objects as the random input variables;

[0011] The probability calculation samples of the probability uncertainty sources are generated through an uncertainty source probability model;

[0012] The uncertain source probability model is established based on the historical data of the probability uncertain source;

[0013] The probability uncertain source is determined by statistically analyzing the historical data of new energy and load in the power grid;

[0014] The model of the power injection from the probability uncertain source to the power grid nodes is established according to the type of the probability uncertain source.

[0015] Furthermore, the available transfer capability calculation model is constructed by means of the power flow constraint, generator output constraint, voltage magnitude constraint, branch power constraint, sending-end load constraint and receiving-end generator constraint under the current operating state of the power grid. The specific construction method of the available transfer capability calculation model is as follows:

[0016] Based on the power grid topology and line parameters, establish the power flow equation of the power grid as the power flow constraint;

[0017] Based on the parameters of each generator in the power grid, determine the upper and lower limits of the active and reactive power outputs of the generators as the generator output constraint;

[0018] Based on the regulations on the node voltage magnitudes in the power grid, establish the voltage magnitude constraint;

[0019] Based on the parameters and operating environment of each transmission line in the power grid, determine the maximum power that the transmission line is allowed to carry as the branch power constraint;

[0020] Based on the principles of constant sending-end power grid load and constant receiving-end generator output, establish the sending-end load constraint and the receiving-end generator constraint;

[0021] Based on the power flow constraint, generator output constraint, voltage magnitude constraint, branch power constraint, sending-end load constraint and receiving-end generator constraint, establish an optimization model with the maximization of the receiving-end power grid load increment as the optimization objective;

[0022] According to the optimization model, when the receiving-end power grid load is maximized, obtain the power P1 transmitted by the sending-end power grid through the tie line;

[0023] According to the actual operating state of the power grid, obtain the power P2 currently transmitted by the sending-end power grid through the tie line;

[0024] According to the difference between P1 and P2, obtain the random output variable under the current operating state of the power grid;

[0025] Construct the available transfer capability calculation model according to the random input variable and the random output variable.

[0026] Further, the tie line is specifically: according to the power grid topology and the actual operation status, determine the power grid partition, where the power grid partition includes a sending-end power grid and a receiving-end power grid, determine the transmission line between the sending-end power grid and the receiving-end power grid, and the transmission line serves as the tie line.

[0027] Further, the method for establishing the available transfer capability calculation model based on the radial basis neural network is specifically as follows:

[0028] Determine the distribution range of the random input variable according to the historical data and actual requirements of the random input variable;

[0029] Uniformly generate random samples of the random input variable according to the distribution range of the random input variable, substitute the random samples of the random input variable into the available transfer capability calculation model to obtain the corresponding available transfer capability output value, use all the random samples of the random input variable as the sample centers of the radial basis neural network, then use all the random samples of the random input variable as the input of the radial basis neural network, and establish a linear equation system with the same dimension as the number of random samples of the random input variable according to the output;

[0030] Obtain the available transfer capability model based on the radial basis neural network by solving the linear equation system.

[0031] Further, the method for establishing the probability model of the uncertainty source is specifically as follows:

[0032] Draw the probability distribution image of the probability uncertainty source through the historical data of the probability uncertainty source, and obtain the probability distribution type that the probability uncertainty source follows according to the characteristics of the probability distribution image of the probability uncertainty source;

[0033] According to the historical data of the probability uncertainty source, calculate the Nth-order origin moment of the probability uncertainty source, where the order of the Nth-order origin moment is equal to the number of distribution parameters of the corresponding probability distribution type;

[0034] According to the probability distribution type of the probability uncertainty source, calculate the Nth-order origin moment with distribution parameters of the probability uncertainty source through the integral method with parameters;

[0035] Establish an equation by making the Nth-order origin moment equal to the Nth-order origin moment with distribution parameters, and use the nonlinear equation solving method to calculate the distribution parameters of the probability uncertainty source, so as to obtain the marginal distribution model for describing the probability characteristics of a single probability uncertainty source;

[0036] For any two probability uncertainty sources, calculate the correlation coefficient between any two probability uncertainty sources, and all the correlation coefficients together form a correlation coefficient matrix for describing the probability characteristics between different probability uncertainty sources;

[0037] The marginal distribution model and the correlation coefficient matrix together constitute the probability model of the uncertainty source.

[0038] Further, the probability distribution of the available transmission capacity obtained according to the available transmission capacity is specifically obtained by a probability calculation method.

[0039] Further, the probability calculation method is specifically as follows:

[0040] Generate a number of probability calculation samples of probability uncertain sources according to the uncertain source probability model;

[0041] Convert the probability calculation samples of the probability uncertain sources into probability calculation samples of random input variables through the model of injecting power from the probability uncertain sources to the power grid nodes;

[0042] Substitute the probability calculation samples of the random input variables into the available transmission capacity calculation model based on the radial basis neural network to obtain a number of available transmission capacities;

[0043] Obtain the probability distribution of the available transmission capacity through statistical means.

[0044] The available transmission capacity probability analysis system includes an acquisition module, a calculation module, and a probability distribution module, where:

[0045] The acquisition module: is used to acquire the probability calculation samples of the random input variables;

[0046] The calculation module: is used to substitute the probability calculation samples of the random input variables into the available transmission capacity calculation model based on the radial basis neural network to solve the available transmission capacity corresponding to each probability calculation sample of the random input variable; among them, the available transmission capacity calculation model based on the radial basis neural network is established through the historical data of the probability uncertain sources and the available transmission capacity calculation model;

[0047] The probability distribution module: is used to obtain the probability distribution of the available transmission capacity according to the available transmission capacity and output it.

[0048] Further, the method for obtaining the probability calculation samples of the random input variables is: converting the probability calculation samples of the probability uncertain sources through the model of injecting power from the probability uncertain sources to the power grid nodes;

[0049] The method for obtaining the random input variables is: through the model of injecting power from the probability uncertain sources to the power grid nodes, obtaining the node injection power objects affected by the probability uncertain sources, and taking the node injection power objects as the random input variables;

[0050] The probability calculation samples of the probability uncertain sources are generated by the uncertain source probability model;

[0051] The uncertain source probability model is established according to the historical data of the probability uncertain sources;

[0052] The probability uncertainty source is determined by statistically analyzing the historical data of new energy and load in the power grid;

[0053] The model for the injection power of the probability uncertainty source into the power grid nodes is established according to the type of the probability uncertainty source.

[0054] Furthermore, the available transfer capability calculation model is constructed by means of power flow constraints, generator output constraints, voltage magnitude constraints, branch power constraints, sending - end load constraints, and receiving - end generator constraints under the current operating state of the power grid. The specific construction method of the available transfer capability calculation model is as follows:

[0055] Based on the power grid topology and line parameters, establish the power flow equation of the power grid as the power flow constraint;

[0056] Based on the parameters of each generator in the power grid, determine the upper and lower limits of the active and reactive power outputs of the generators as the generator output constraints;

[0057] Based on the regulations on the node voltage magnitudes of the power grid, establish the voltage magnitude constraints;

[0058] Based on the parameters and operating environment of each transmission line in the power grid, determine the maximum power that the transmission line is allowed to carry as the branch power constraint;

[0059] According to the principles of constant sending - end power grid load and constant receiving - end generator output, establish the sending - end load constraints and receiving - end generator constraints;

[0060] Based on the power flow constraints, generator output constraints, voltage magnitude constraints, branch power constraints, sending - end load constraints, and receiving - end generator constraints, with the maximization of the receiving - end power grid load increment as the optimization objective, establish an optimization model;

[0061] According to the optimization model, when the receiving - end power grid load is maximized, obtain the power P1 transmitted by the sending - end power grid through the tie - line;

[0062] According to the actual operating state of the power grid, obtain the power P2 currently transmitted by the sending - end power grid through the tie - line;

[0063] According to the difference between P1 and P2, obtain the random output variable under the current operating state of the power grid;

[0064] Based on the random input variable and the random output variable, construct the available transfer capability calculation model.

[0065] Furthermore, the specific establishment method of the available transfer capability calculation model based on the radial basis neural network is as follows:

[0066] Determine the distribution range of the random input variable according to the historical data and actual requirements of the random input variable;

[0067] Uniformly generate random samples of random input variables according to the distribution range of random input variables, substitute the random samples of random input variables into the available transfer capability calculation model to obtain the corresponding available transfer capability output values, take all the random samples of random input variables as the sample centers of the radial basis neural network, and then use all the random samples of random input variables as the inputs of the radial basis neural network. Establish a system of linear equations with the same dimension as the number of random samples of random input variables according to the output;

[0068] By solving the system of linear equations, an available transfer capability model based on the radial basis neural network is obtained.

[0069] Furthermore, the method for establishing the probability model of the uncertainty source is specifically as follows:

[0070] Draw the probability distribution image of the probability uncertainty source through the historical data of the probability uncertainty source, and obtain the probability distribution type followed by the probability uncertainty source according to the characteristics of the probability distribution image of the probability uncertainty source;

[0071] According to the historical data of the probability uncertainty source, calculate the Nth-order origin moment of the probability uncertainty source, and the order of the Nth-order origin moment is equal to the number of distribution parameters of the corresponding probability distribution type;

[0072] According to the probability distribution type of the probability uncertainty source, calculate the Nth-order origin moment with distribution parameters of the probability uncertainty source through the integral method with parameters;

[0073] Establish an equation by making the Nth-order origin moment equal to the Nth-order origin moment with distribution parameters, and use the nonlinear equation solving method to calculate the distribution parameters of the probability uncertainty source, so as to obtain the marginal distribution model for describing the probability characteristics of a single probability uncertainty source;

[0074] For any two probability uncertainty sources, calculate the correlation coefficient between any two probability uncertainty sources, and all the correlation coefficients together form a correlation coefficient matrix for describing the probability characteristics between different probability uncertainty sources;

[0075] The marginal distribution model and the correlation coefficient matrix together constitute the probability model of the uncertainty source.

[0076] Furthermore, the method for obtaining the probability distribution of the available transfer capability according to the available transfer capability specifically uses a probability calculation method, and the probability calculation method is specifically as follows:

[0077] Generate probability calculation samples of several probability uncertainty sources according to the probability model of the uncertainty source;

[0078] Convert the probability calculation samples of the probability uncertainty source into probability calculation samples of random input variables through the model of the probability uncertainty source injecting power into the grid node;

[0079] Bring the probability calculation samples of random input variables into the available transfer capability calculation model based on the radial basis neural network to obtain several available transfer capabilities.

[0080] By means of statistics, obtain the probability distribution of the available transfer capability.

[0081] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the available transfer capability probability analysis method.

[0082] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, it implements the steps of the available transfer capability probability analysis method.

[0083] Compared with the prior art, the present invention has the following beneficial technical effects:

[0084] The present invention can obtain an approximate calculation model of the available transfer capability of the power grid in an offline state, that is, the available transfer capability calculation model based on the radial basis neural network. The mathematical expression form of this approximate model is more concise than that of the available transfer capability calculation model, and the solution efficiency is higher. At the same time, because the radial basis neural network has a good approximation effect on any continuous model, the available transfer capability calculation model based on the radial basis neural network has high accuracy. Therefore, the present invention has the characteristics of accurate calculation results and high calculation efficiency, and is used for the online probability calculation of the available transfer capability. Description of the Drawings

[0085] The drawings in the specification are used to provide a further understanding of the present invention and constitute a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0086] Figure 1 It is a flowchart of the available transfer capability probability analysis method based on the radial basis neural network in an embodiment of the present invention;

[0087] Figure 2 It is a schematic diagram of the power system structure in an embodiment of the present invention;

[0088] Figure 3 It is a structure diagram of the radial basis neural network in an embodiment of the present invention. Detailed Embodiments

[0089] The following further details the present invention in conjunction with the drawings and specific embodiments.

[0090] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0091] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned accompanying drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0092] The present invention will be further described in detail below in conjunction with the accompanying drawings:

[0093] See Figure 1 , in an embodiment of the present invention, a method for probabilistic analysis of available transfer capability based on a radial basis neural network is proposed. This method is mainly applied to: realizing the probabilistic analysis of the available transfer capability of the power grid quickly under random scenarios, so as to guide the current power market transactions in a timely manner and guide the optimal allocation of market resources. Specifically, the method for probabilistic analysis of available transfer capability based on a radial basis neural network of the present invention includes the following steps.

[0094] Step S1: Determine the probabilistic uncertainty sources and the corresponding random input variables. According to the historical data of each element in the power grid, determine the probabilistic uncertainty sources of the power grid. In this embodiment, the active load and the wind speed of the wind farm are used as the probabilistic uncertainty sources, but the applicable scope of the present invention is not limited to these two types of probabilistic uncertainty sources. Further, convert the probabilistic uncertainty sources into random input variables of the available transfer capability model, such as converting the wind speed into the output of the wind farm while keeping the load unchanged.

[0095] Step S2: Establish a calculation model for available transfer capability (ATC) based on optimization methods: Based on the power grid topology and line parameters, establish the power flow equation of this power grid as the power flow constraint of the ATC calculation model; based on the parameters of each generator in the power grid, determine the upper and lower limits of the active and reactive power outputs of the generators as the generator output constraints, that is, the minimum value of generator active / reactive power output ≤ generator active / reactive power output ≤ the maximum value of generator active / reactive power output; based on the regulations on the node voltage amplitude of the power grid, establish the voltage amplitude constraint, that is, the minimum value of voltage amplitude ≤ voltage amplitude ≤ the maximum value of voltage amplitude; based on the parameters and operating environment of each transmission line in the power grid, determine the maximum power that the transmission line is allowed to carry, and use this maximum power as the branch power constraint, that is, 0 ≤ the apparent power actually flowing through the branch ≤ the branch power upper limit; according to the unchanged load of the sending-end power grid and the unchanged generator output of the receiving-end power grid, establish the sending-end load constraint and the receiving-end generator constraint, that is, the sending-end power grid load = the original load of the sending-end power grid, and the active power output of the receiving-end power grid generator = the original active power output of the receiving-end power grid generator; based on the above constraints, with the maximization of the load increment of the receiving-end power grid as the optimization goal, establish an optimization model, and according to the optimization model, obtain the power P1 transmitted by the sending-end power grid through the tie line when the load of the receiving-end power grid is the largest; according to the actual operating state of the power grid, obtain the power P2 currently transmitted by the sending-end power grid through the tie line; according to the difference between P1 and P2, obtain the random output variable under the current operating state of the power grid; construct an ATC calculation model based on the random input variable and the random output variable.

[0096] Specifically, first, according to the actual situation of the target power grid, determine the sending-end power grid, receiving-end power grid, and tie line as shown in Figure 2 . Solve the power flow of this power grid through the power flow equation shown below.

[0097]

[0098]

[0099] where N is the number of nodes in this power grid; P i and Q i are the injected active power and reactive power of node i respectively; V i , V j and θ ij represent the voltage amplitudes and their phase differences of nodes i and j; G ij and B ij are the real and imaginary parts of the element in the i-th row and j-th column of the node admittance matrix of the power grid respectively.

[0100] According to the power flow result of this power grid, further calculate the actual active power flowing through a single tie line of this power grid according to the following formula.

[0101]

[0102] Among them, P S→R is the actual active power flowing from the sending-end power grid to the receiving-end power grid on a single tie line. The operation "real()" represents taking the real part of a complex number, and represent the voltage phasors of the sending-end power grid node and the receiving-end power grid node connected by the same tie line. Y TL represents the admittance of this tie line, and the superscript "*" represents taking the conjugate complex number.

[0103] Furthermore, add up the active powers flowing through all tie lines to obtain the total power transmitted from the sending-end power grid to the receiving-end power grid, that is, P1.

[0104] According to the following optimization model, find the maximum value that the load of the receiving-end power grid can increase to through the power transmission of the sending-end power grid under the constraint conditions.

[0105]

[0106] s.t.

[0107]

[0108]

[0109] P G,i = P G,i,0 , i ∈ R

[0110] P L,i = P L,i,0 , i ∈ S

[0111]

[0112]

[0113]

[0114]

[0115] Among them, R and S represent the sets of node numbers of the receiving-end power grid and the sending-end power grid, and B represents the set of node numbers of the entire power grid. P L,i and P L,i,0 respectively represent the active load of node i and the corresponding actual value; P G,i and P G,i,0 respectively represent the active power output of the generator at node i and the corresponding actual value; P wind,i and Q wind,i respectively represent the active and reactive power outputs of wind power at node i; V i , V j and θ ijDenote the voltage magnitudes of nodes \(i\) and \(j\) and their phase difference; \(Q\) G,i and \(Q\) L,i respectively denote the reactive power output and reactive power load of the generator at node \(i\); \(S\) ij denotes the apparent power flowing from node \(i\) to node \(j\); and the superscripts “max” and “min” respectively represent the upper and lower limits of the variable.

[0116] Based on the power flow results obtained from the above optimization model, use the following formula again to calculate the power flowing from the sending - end power grid to the receiving - end power grid on each tie - line under the condition of the maximum load in the receiving - end power grid.

[0117]

[0118] Sum the powers flowing through all tie - lines to obtain the power \(P_2\) transmitted from the sending - end power grid to the receiving - end power grid under the optimization model. Thus, the random output variable under the current operating state of the power grid = the power \(P_1\) transmitted from the sending - end power grid to the receiving - end power grid under the optimization model - the power \(P_2\) transmitted from the sending - end power grid to the receiving - end power grid under the actual operating state of the power grid.

[0119] Step S3: Analyze the characteristics of the probabilistic uncertainty sources and random input variables. Specifically, first establish a probabilistic model of the uncertainty sources, that is, determine the probability distribution type followed by the probabilistic uncertainty sources according to the historical data of the probabilistic uncertainty sources. Specifically, draw the probability distribution image of the probabilistic uncertainty source through the historical data of the probabilistic uncertainty source, and obtain the probability distribution type followed by the probabilistic uncertainty source according to the characteristics of the probability distribution image of the probabilistic uncertainty source. In addition, according to the historical data of the probabilistic uncertainty source, calculate the \(N\) - th order origin moment of the probabilistic uncertainty source, where the order of the \(N\) - th order origin moment is equal to the number of distribution parameters of the corresponding probability distribution type. According to the probability distribution type of the probabilistic uncertainty source, use the integral method with parameters to calculate the \(N\) - th order origin moment with distribution parameters of the probabilistic uncertainty source. Establish an equation by equating the \(N\) - th order origin moment and the \(N\) - th order origin moment with distribution parameters, and use the non - linear equation solving method to calculate the distribution parameters of the probabilistic uncertainty source, so as to obtain the marginal distribution model describing the probability characteristics of a single probabilistic uncertainty source.

[0120] According to the definition of the correlation coefficient, for any two probabilistic uncertainty sources, calculate the correlation coefficient between any two probabilistic uncertainty sources, and all the correlation coefficients together form a correlation coefficient matrix describing the probability characteristics between different probabilistic uncertainty sources.

[0121] Combine the marginal distribution model and the correlation coefficient matrix to form a probabilistic model of the uncertainty sources.

[0122] For the characteristics of the random input variables of the available transmission capacity model, according to the historical data of the probabilistic uncertainty sources (such as load, wind speed, etc.) or the historical data of the random input variables (such as load, wind power, etc.), obtain the numerical range of the random input variables as the distribution range of the random input variables of the available transmission capacity calculation model based on the radial basis neural network. The significance of this input variable range is that the radial basis neural network is established based on the samples of the input variables within this range, which means that when using the available transmission capacity calculation model based on the radial basis neural network, when the numerical values of the input variables are distributed within the specified random input variable distribution range, its calculation results are highly accurate; when the numerical values of the input variables significantly exceed this random input variable distribution range, the calculation results of the available transmission capacity calculation model based on the radial basis neural network are poor.

[0123] Step S4: Establish an available transmission capacity calculation model based on the radial basis neural network. Specifically, according to Figure 3 the radial basis neural network shown, the mathematical expression of the radial basis neural network is obtained as:

[0124]

[0125] where, y RBF represents the output of the radial basis neural network, which is the available transmission capacity here; N C represents the number of neurons in the hidden layer of the radial basis neural network; ω k represents the weight coefficient from the k-th hidden layer neuron to the output; X in represents the random input variable vector; T k represents the sample center corresponding to the k-th hidden layer neuron; ψ represents the radial basis function, specifically a symmetric and bounded function that is high in the middle and low on both sides and is monotonic on both sides of the symmetry axis, such as the Gaussian function; the operation "||||2" represents obtaining the 2-norm of the vector, which represents obtaining the distance from the randomly selected input variable vector to the sample center of the k-th hidden layer neuron here.

[0126] Within the distribution range of the random input variables set in step S3, uniformly generate N s groups of input variable samples, and substitute these input variable samples into the available transmission capacity calculation model established in step S1 to obtain the corresponding available transmission capacity values. Here, "uniformly generate" can be considered that each random input variable follows a uniform distribution on the random input variable distribution range, so as to generate the input variable samples here by generating uniformly distributed random samples. Set N s = N C , take all N s groups of input variable samples as the sample centers of the radial basis neural network, and then substitute all the input variable samples into the mathematical expression of the radial basis function neural network respectively to obtain the following Ns Equations:

[0127]

[0128]

[0129]

[0130]

[0131] The above equation can be expressed as a matrix:

[0132] Hω=y RBF

[0133] Where H is the coefficient matrix, the element in the pth row and kth column is ψ(||T p -T k ||2); ω=[ω1,ω2,..., ω Ns ] T is the weight vector. According to Micchelli's theorem, H is a nonsingular matrix, so the weight vector can be directly solved. This completes the construction of the available power transfer capacity calculation model based on the radial basis function neural network. Given a set of random input variable values, the corresponding available power transfer capacity value can be calculated using the mathematical expression of the radial basis function neural network.

[0134] Step S5: Generate probability calculation samples of random input variables. Specifically, based on the uncertainty source probability model obtained in step S3, obtain the inverse cumulative distribution function of each probability uncertainty source and obtain the correlation coefficient matrix of the random input variables.

[0135] According to the number of probabilistic uncertainty sources, generate independent standard normal distribution variable samples U of the same dimension; then convert the independent standard normal distribution samples into correlated standard normal distribution samples Z. The conversion formula is as follows:

[0136] Z=LU

[0137] Where L is the correlation coefficient matrix C of Z Z The decomposition matrix obtained by Cholesky decomposition, namely C Z =LL T . Among them C Z The element in the mth row and nth column of Z,mn . And ρ Z,mn Correlation coefficient ρ with the mth and nth random input variables X,mn Satisfies the following relationship:

[0138]

[0139] in, represents the joint probability density function of the standard normal distribution, Φ represents the cumulative distribution function of the standard normal distribution; μ m and μ n respectively represent the means of the m-th and n-th random input variables, σ m and σ n represent the standard deviations of the m-th and n-th random input variables; and represent the inverse functions of the cumulative distribution functions of the m-th and n-th random input variables.

[0140] For the probability calculation samples of the probability uncertainty sources of the m-th standard normal distribution variables with correlation, they are converted into the calculation samples of the random input variables (loads, wind speeds, etc.) in the power grid through the following conversion formula, denoted as the probability calculation samples of the random input variables:

[0141]

[0142] During the conversion process, the load samples do not need to be converted, while the wind speed samples are converted into the active power output of the wind turbines in the wind farm according to the following formula:

[0143]

[0144] where, v wind represents the wind speed, with the unit of m / s; P T represents the output of a single wind turbine, with the unit of MW. Further, the active power output of the wind farm = the output of a single wind turbine × the number of wind turbines operating in the wind farm.

[0145] Step S6: Calculation of the available transfer capability. Specifically, the probability calculation samples of all the random input variables generated in Step S5 are brought into the available transfer capability calculation model based on the radial basis neural network generated in Step S4 for calculation, and the corresponding available transfer capabilities are obtained, where the number of available transfer capabilities = the number of input variable samples.

[0146] Step S7: Analysis of the probability characteristics of the available transfer capability. Specifically, for the available transfer capabilities calculated in Step S6, their means and variances are obtained, and at the same time, the probability distribution of the available transfer capabilities is obtained, and its probability distribution image is drawn for practical engineering applications.

[0147] The present invention also provides a probability analysis system for the available transfer capability based on a radial basis neural network, including:

[0148] an acquisition module, a calculation module, and a probability distribution module, where:

[0149] Acquisition module: used to acquire probability calculation samples of random input variables; the method for acquiring probability calculation samples of the random input variables is: obtained by converting the probability calculation samples of the probability uncertainty source through a model of the power injection from the probability uncertainty source to the power grid nodes, the method for acquiring the random input variables is: through a model of the power injection from the probability uncertainty source to the power grid nodes, acquiring the power injection object of the nodes affected by the probability uncertainty source and using it as the random input variable, the probability calculation samples of the probability uncertainty source are generated by an uncertainty source probability model, the uncertainty source probability model is established based on the historical data of the probability uncertainty source, the probability uncertainty source is determined by statistically analyzing the historical data of new energy and load in the power grid, and the model of the power injection from the probability uncertainty source to the power grid nodes is established according to the type of the probability uncertainty source;

[0150] Calculation module: used to substitute the probability calculation samples of the random input variables into the available transfer capability calculation model based on the radial basis neural network to solve the available transfer capability corresponding to each probability calculation sample of the random input variables; the available transfer capability calculation model based on the radial basis neural network is established through the historical data of the probability uncertainty source and the available transfer capability calculation model, and the available transfer capability calculation model is constructed by the power flow constraint, generator output constraint, voltage magnitude constraint, branch power constraint, sending-end load constraint and receiving-end generator constraint under the current operating state of the power grid;

[0151] Probability distribution module: used to obtain the probability distribution of the available transfer capability according to the available transfer capability.

[0152] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can be in the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can be in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.

[0153] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for realizing the processes in the Figure 1 each process or multiple processes and / or blocks Figure 1means for the functions specified in one or more boxes.

[0154] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one Figure 1 process or more processes and / or boxes Figure 1 means for the functions specified in one or more boxes.

[0155] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus, such that a series of operation steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one Figure 1 process or more processes and / or boxes Figure 1 means for the functions specified in one or more boxes.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the scope of its protection. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: after reading the present invention, those skilled in the art may still make various changes, modifications or equivalent replacements to the specific implementation manners of the invention, but these changes, modifications or equivalent replacements are all within the scope of protection of the pending claims of the invention.

Claims

1. Probability analysis method for available transmission capacity, characterized in that The method includes: Obtaining probability calculation samples of random input variables; Substituting the probability calculation samples of the random input variables into the available transfer capability calculation model based on a radial basis neural network to solve the available transfer capability corresponding to each probability calculation sample of the random input variable; wherein the available transfer capability calculation model based on a radial basis neural network is established through historical data of probability uncertain sources and an available transfer capability calculation model; Obtaining the probability distribution of the available transfer capability according to the available transfer capability and outputting it; Among them, the method for obtaining the probability calculation samples of the random input variables is: converting the probability calculation samples of the probability uncertain sources through a model of the power injection from the probability uncertain sources to the power grid nodes; The method for obtaining the random input variables is: through a model of the power injection from the probability uncertain sources to the power grid nodes, obtaining the node injection power objects affected by the probability uncertain sources, and taking the node injection power objects as random input variables; The probability calculation samples of the probability uncertain sources are generated by an uncertain source probability model; The uncertain source probability model is established according to the historical data of the probability uncertain sources; The probability uncertain sources are determined by statistically analyzing the historical data of new energy and load in the power grid; The model of the power injection from the probability uncertain sources to the power grid nodes is established according to the type of the probability uncertain sources; The available transfer capability calculation model is constructed by power flow constraints, generator output constraints, voltage magnitude constraints, branch power constraints, sending - end load constraints and receiving - end generator constraints under the current operating state of the power grid. The specific construction method of the available transfer capability calculation model is: Establishing the power flow equation of the power grid through the power grid topology structure and line parameters as the power flow constraint; Determining the upper and lower limits of the active and reactive power outputs of the generators through the parameters of each generator in the power grid as the generator output constraints; Establishing the voltage magnitude constraint through the regulations on the node voltage magnitudes in the power grid; Determining the maximum power allowed to flow through the transmission lines as the branch power constraints through the parameters and operating environment of each transmission line in the power grid; Establishing the sending - end load constraint and the receiving - end generator constraint according to the principles of constant sending - end power grid load and constant receiving - end generator output; Based on the power flow constraint, generator output constraint, voltage magnitude constraint, branch power constraint, sending - end load constraint and receiving - end generator constraint, establishing an optimization model with the maximization of the receiving - end power grid load increment as the optimization objective; According to the optimization model, when the receiving - end power grid load is the largest, obtaining the power P1 transmitted by the sending - end power grid through the tie - line; According to the actual operating state of the power grid, obtaining the power P2 currently transmitted by the sending - end power grid through the tie - line; According to the difference between P1 and P2, obtaining the random output variable under the current operating state of the power grid; Constructing the available transfer capability calculation model according to the random input variables and the random output variables; The specific method for establishing the available transfer capability calculation model based on a radial basis neural network is: Determining the distribution range of the random input variables according to the historical data and actual requirements of the random input variables; Generate random samples of random input variables uniformly according to the distribution range of random input variables. Substitute the random samples of random input variables into the available transfer capability calculation model to obtain the corresponding available transfer capability output values. Use all the random samples of random input variables as the sample centers of the radial basis neural network, and then use all the random samples of random input variables as the inputs of the radial basis neural network. Establish a system of linear equations with the same dimension as the number of random samples of random input variables according to the output. Obtain the available transfer capability model based on the radial basis neural network by solving the system of linear equations. The method for establishing the probability model of the uncertainty source is specifically as follows: Draw the probability distribution image of the probability uncertainty source through the historical data of the probability uncertainty source, and obtain the probability distribution type that the probability uncertainty source follows according to the characteristics of the probability distribution image of the probability uncertainty source. According to the historical data of the probability uncertainty source, calculate the Nth-order origin moment of the probability uncertainty source, and the order of the Nth-order origin moment is equal to the number of distribution parameters of the corresponding probability distribution type. According to the probability distribution type of the probability uncertainty source, use the integral method with parameters to calculate the Nth-order origin moment with distribution parameters of the probability uncertainty source. Establish an equation by making the Nth-order origin moment equal to the Nth-order origin moment with distribution parameters, and use the nonlinear equation solving method to calculate the distribution parameters of the probability uncertainty source, so as to obtain the marginal distribution model for describing the probability characteristics of a single probability uncertainty source. For any two probability uncertainty sources, calculate the correlation coefficient between any two probability uncertainty sources, and all the correlation coefficients together form a correlation coefficient matrix for describing the probability characteristics between different probability uncertainty sources. The marginal distribution model and the correlation coefficient matrix together constitute the probability model of the uncertainty source. The method for obtaining the probability distribution of the available transfer capability according to the available transfer capability specifically uses a probability calculation method, and the probability calculation method is specifically as follows: Generate probability calculation samples of several probability uncertainty sources according to the probability model of the uncertainty source. Convert the probability calculation samples of the probability uncertainty sources into probability calculation samples of random input variables through the model of the power injection from the probability uncertainty source to the grid nodes. Substitute the probability calculation samples of the random input variables into the available transfer capability calculation model based on the radial basis neural network to obtain several available transfer capabilities. Obtain the probability distribution of the available transfer capability through statistical means.

2. The probabilistic analysis method for available transmission capacity according to claim 1, wherein The tie line is specifically: According to the grid topology structure and the actual operation state, determine the grid partition, where the grid partition includes the sending-end grid and the receiving-end grid, and determine the transmission line between the sending-end grid and the receiving-end grid, and the transmission line is used as the tie line.

3. Probabilistic analysis system for available transmission capacity, characterized in that, It includes an acquisition module, a calculation module, and a probability distribution module, where: The acquisition module: is used to acquire the probability calculation samples of random input variables. The calculation module: is used to substitute the probability calculation samples of the random input variables into the available transfer capability calculation model based on the radial basis neural network, and solve the available transfer capability corresponding to each probability calculation sample of the random input variables; among them, the available transfer capability calculation model based on the radial basis neural network is established through the historical data of the probability uncertainty source and the available transfer capability calculation model. Probability distribution module: used to obtain the probability distribution of available transfer capability according to the available transfer capability and output it; Among them, the method for obtaining the probability calculation samples of the random input variables is: converting the probability calculation samples of the probability uncertainty source through the model of the power injection from the probability uncertainty source to the grid node; The method for obtaining the random input variables is: through the model of the power injection from the probability uncertainty source to the grid node, obtaining the node power injection objects affected by the probability uncertainty source, and taking the node power injection objects as the random input variables; The probability calculation samples of the probability uncertainty source are generated by the uncertainty source probability model; The uncertainty source probability model is established according to the historical data of the probability uncertainty source; The probability uncertainty source is determined by statistically analyzing the historical data of new energy and load in the power grid; The model of the power injection from the probability uncertainty source to the grid node is established according to the type of the probability uncertainty source; The available transfer capability calculation model is constructed by the power flow constraint, generator output constraint, voltage amplitude constraint, branch power constraint, sending-end load constraint and receiving-end generator constraint under the current operating state of the power grid. The specific construction method of the available transfer capability calculation model is as follows: Establish the power flow equation of the power grid through the power grid topology structure and line parameters as the power flow constraint; Determine the upper and lower limits of the active and reactive power outputs of the generators through the parameters of each generator in the power grid as the generator output constraint; Establish the voltage amplitude constraint through the regulations on the node voltage amplitude of the power grid; Determine the maximum power allowed to flow through the transmission line as the branch power constraint through the parameters and operating environment of each transmission line in the power grid; Establish the sending-end load constraint and the receiving-end generator constraint according to the principles of constant sending-end grid load and constant receiving-end grid generator output; According to the power flow constraint, generator output constraint, voltage amplitude constraint, branch power constraint, sending-end load constraint and receiving-end generator constraint, establish an optimization model with the maximization of the receiving-end grid load increment as the optimization goal; According to the optimization model, obtain the power P1 transmitted by the sending-end grid through the tie line when the receiving-end grid load is the largest; Obtain the power P2 currently transmitted by the sending-end grid through the tie line according to the actual operating state of the power grid; According to the difference between P1 and P2, obtain the random output variable under the current operating state of the power grid; Construct the available transfer capability calculation model according to the random input variables and random output variables; The specific method for establishing the available transfer capability calculation model based on the radial basis neural network is as follows: Determine the distribution range of the random input variables according to the historical data and actual requirements of the random input variables; Uniformly generate random samples of the random input variables according to the distribution range of the random input variables. Substitute the random samples of the random input variables into the available transfer capability calculation model to obtain the corresponding available transfer capability output values. Take all the random samples of the random input variables as the sample centers of the radial basis neural network, and then take all the random samples of the random input variables as the inputs of the radial basis neural network. Establish a linear equation system with the same dimension as the number of random samples of the random input variables according to the output; Obtain the available transfer capability model based on the radial basis neural network by solving the linear equation system; The method for establishing the probability model of the uncertainty source is specifically as follows: Draw the probability distribution image of the probability uncertainty source through the historical data of the probability uncertainty source, and obtain the probability distribution type followed by the probability uncertainty source according to the characteristics of the probability distribution image of the probability uncertainty source; According to the historical data of the probability uncertainty source, calculate the N-th order origin moment of the probability uncertainty source, and the order of the N-th order origin moment is equal to the number of distribution parameters of the corresponding probability distribution type; According to the probability distribution type of the probability uncertainty source, calculate the N-th order origin moment with distribution parameters of the probability uncertainty source by the integral method with parameters; Establish an equation by making the N-th order origin moment equal to the N-th order origin moment with distribution parameters, and use the nonlinear equation solving method to calculate the distribution parameters of the probability uncertainty source, so as to obtain the marginal distribution model for describing the probability characteristics of a single probability uncertainty source; For any two probability uncertainty sources, calculate the correlation coefficient between any two probability uncertainty sources, and all the correlation coefficients together form a correlation coefficient matrix for describing the probability characteristics between different probability uncertainty sources; The marginal distribution model and the correlation coefficient matrix together constitute the probability model of the uncertainty source; The method for obtaining the probability distribution of the available transfer capability according to the available transfer capability specifically adopts a probability calculation method, and the probability calculation method is specifically as follows: Generate probability calculation samples of several probability uncertainty sources according to the probability model of the uncertainty source; Convert the probability calculation samples of the probability uncertainty source into probability calculation samples of random input variables through the model of the probability uncertainty source injecting power into the grid node; Substitute the probability calculation samples of the random input variables into the available transfer capability calculation model based on the radial basis neural network to obtain several available transfer capabilities; Through statistical means, obtain the probability distribution of the available transfer capability.

4. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the available transfer capability probability analysis method as described in claim 1 or 2.

5. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the available transfer capability probability analysis method as described in claim 1 or 2.

Images