A new energy power grid power transmission reliability margin probability evaluation method and system

Abstract

The application discloses a new energy power grid power transmission reliability margin probability evaluation method and system. Under the uncertainty caused by large-scale renewable energy grid connection and load change, an innovative solution is provided for the probability power transmission reliability margin evaluation in the power grid by combining the uncertainty quantification method of orthogonal polynomial approximation, the power transmission reliability margin can be accurately and efficiently evaluated, and remarkable effects are exhibited in improving the calculation efficiency, improving the evaluation accuracy, supporting decision making and promoting renewable energy integration, and the application has important practical application value for the planning and operation of a modern power system.

Description

TECHNICAL FIELD

[0001] The application relates to a new energy power grid power transmission reliability margin probability evaluation method and system, and belongs to the field of power system planning and operation optimization. BACKGROUND

[0002] With the wide access of renewable energy (such as wind energy and solar photovoltaic) and new type of load (such as electric vehicles), more and more uncertainties are brought to the power grid, which poses new challenges to the power transmission reliability margin evaluation of renewable energy power grid. These challenges mainly come from renewable energy, such as the high uncertainty of wind energy and solar energy and the volatility of their output, which increases the complexity of the power network in maintaining stability and reliable power transmission capability. Although the existing probability evaluation methods, such as the Monte Carlo simulation method, can provide certain evaluation results, the method is very time-consuming and inefficient in calculation because a large number of samples are needed to obtain statistical accuracy. Therefore, developing a new method that can effectively deal with uncertainty and significantly improve the calculation efficiency and evaluation accuracy is crucial for the planning and optimization of renewable energy in the power network. SUMMARY

[0003] The application aims at the deficiencies of the prior art and provides a new energy power grid power transmission reliability margin probability evaluation method and system, which introduces a calculation algorithm combined with orthogonal polynomials to accurately evaluate the power transmission reliability margin while significantly reducing the calculation workload.

[0004] Technical scheme: In a first aspect, a new energy power grid power transmission reliability margin probability evaluation method comprises the following steps:

[0005] Step S1: input the data of the power network, including the probability distribution and parameters of renewable energy and load;

[0006] Step S2: for each random input, select an independent standard variable and its corresponding orthogonal polynomial;

[0007] Step S3: generate an experimental design, generate samples in the standard normal space using the Latin hypercube sampling method, then convert the random variables of the standard normal space to the original input variable space, evaluate the converted samples using the probability continuous power flow equation PCPF, and obtain the deterministic response value;

[0008] Step S4: determine the best orthogonal polynomial approximation PCE of each response through an adaptive sparse process;

[0009] Step S5: if the accuracy of the orthogonal polynomial approximation of all responses meets the specified requirements, go to step S6, otherwise go to step S3 to generate additional samples, and then go to step S4;

[0010] Step S6: Evaluate the specified large number of samples using the obtained best orthogonal polynomial approximation to determine the response values;

[0011] Step S7: Calculate the statistical data of each response and generate a result report.

[0012] Further, in the step S1, the renewable energy includes wind power generation and solar power generation, and each uncertain input model is represented as a random variable related to a probability density function X ~ f X (x), for wind power generation, the statistical characteristics of wind speed v in the short term are fitted by a normal distribution, and the corresponding probability density function is: where v is the wind speed, the expectation μ v and the variance are obtained from historical data;

[0013] For solar power generation, the statistical characteristics of solar radiation r in the short term are also fitted by a normal distribution, and the corresponding probability density function is: where r is the actual solar radiation, the expectation μ r and the variance are obtained from historical data;

[0014] The uncertainty of load forecasting is described using a normal distribution, and the load active power P L forecasting result is the mean value provided by the load forecaster, and the prediction variance is represented by σ

[0015] Further, for each realization of wind speed v, the corresponding wind power injected into the terminal bus is calculated from the wind speed-power output relationship:

[0016]

[0017] where v in , v out and v rated are the cut-in, cut-out and rated wind speeds, respectively, P r is the rated wind power, and P w is the wind power output active power;

[0018] The solar radiation-power output relationship is expressed as:

[0019]

[0020] where P pv represents the power output of solar power generation, r c is the specified radiation point, r std is the solar radiation amount under standard environment, and P rs is the rated capacity of solar photovoltaic.

[0021] Furthermore, step S2 includes: selecting a suitable orthogonal polynomial as the basis function based on the probability distribution type of the random variable, and constructing a surrogate model. In the formula, the random vector ξ=(ξ1,ξ2,...,ξ) n It has n independent components; Ψ k (ξ) is related to f ξ Orthogonal multivariate polynomials; c k These are the coefficients of an orthogonal polynomial; Where the subscript i j f represents the j-th degree of the i-th univariate polynomial basis. ξ It is the joint probability density function of random vectors. It is about Orthogonal basis.

[0022] Furthermore, step S3 includes: generating M in the standard normal space using the Latin hypercube sampling method. C Sample

[0023] Using the variable transformation principle, the random variable ξ in the standard normal space is transformed... C Transform to the original input variable space u C : Here u C This represents wind speed v, solar radiation r, and active power of the load P. L , is u C The inverse cumulative distribution function, φ is ξ C The cumulative distribution function;

[0024] Evaluate the transformed sample u using PCPF C To obtain deterministic response values

[0025] Furthermore, using v, r, and P L Let be random vectors representing wind speed, solar radiation, and load variation, respectively. For a PV-type node system, the three-phase probabilistic continuous power flow (PCPF) equations for an N-node system are expressed as:

[0026]

[0027] In the formula, Is node i in phase The voltage amplitude, M is the number of phases; and It is the phase k between nodes i and j. The conductance and susceptance of the admittance matrix; and are the cosine and sine of the voltage phase angle difference between nodes i and j at phase k and λ is the Lagrange multiplier for the handling of the constraints in the optimization problem; and are the actual power changes of wind power, solar photovoltaic, load and other types of distributed generators at bus i phase ; and are the reactive power changes due to wind speed changes and load changes; is the reactive power generation of node i at phase ; min,i and Q max,i are the minimum and maximum values of the reactive power; the above equations are expressed in a compact form as:

[0028] f(x, μ, λ, U) = f(x, μ) - λb(U) = 0

[0029] where x is the state vector, μ is the control parameter vector, U = [v, r, P L ] is the random vector describing wind speed, solar radiation, load active power; the load change vector b of the system is:

[0030]

[0031] The optimization problem of the power transmission reliability margin probability is expressed as:

[0032] max λ

[0033] s.t. f(x, μ) - λb(U) = 0

[0034] V min ≤ V i (x, μ, λ, U) ≤ V max

[0035]

[0036] where V min and V max are the lower and upper limits of the node voltage amplitude; I ij,max is the specified capacity of the line or transformer between node i and node j; λ is the normalized load margin under the given load change vector, and the maximum value of λ under the constraints of the above equation corresponds to the power transmission reliability margin.

[0037] Further, the kth orthogonal polynomial approximation of Y i in the step S4 is constructed by the following steps:

[0038] S41, generate k candidate multi-index sets, and use the formula cut them, where n is the number of random variable inputs, and p is the total order of orthogonal polynomials; evaluate the resulting k-term multivariate polynomial to form the design matrix H k k-1 , k , ΔH k is the matrix corresponding to the newly added kth polynomial;

[0039] S42, apply the least angle regression (LAR) algorithm to select the best reserved design matrix H k The columns of the set are used to estimate the corresponding coefficients by the least squares method, for the random variable The corresponding random response calculated by the deterministic method The coefficients are calculated by solving the least squares minimization problem:

[0040] S43, calculate the leave-one-out error indicator using the formula cloo ε loo ×T(M C ,P) where h=diag[H(H T H) -1 H T ], is the empirical variance of the response vector y C ; if the error reaches the specified accuracy, return the best orthogonal polynomial approximation, otherwise increase the order of the polynomial and return to S41.

[0041] Secondly, a new energy power grid transmission reliability margin probability evaluation system, comprising:

[0042] Data input module, for inputting the data of the power network, including the probability distribution and parameters of renewable energy and load;

[0043] Orthogonal polynomial selection module, for selecting an independent standard variable and its corresponding orthogonal polynomial for each random input;

[0044] Experimental design module, for generating an experimental design, using the Latin hypercube sampling method to generate samples in the standard normal space, then converting the random variables of the standard normal space to the original input variable space, using the probabilistic continuous power flow equation (PCPF) to evaluate the converted samples, and obtaining the deterministic response value;

[0045] PCE determination module, for determining the best orthogonal polynomial approximation PCE of each response through an adaptive sparse process;

[0046] ​A response value determination module is configured to determine whether the accuracy of the orthogonal polynomial approximation of all responses meets the specified requirements. If yes, the obtained best orthogonal polynomial approximation is used to evaluate a large number of samples to determine the response value. If not, the experimental design module is called to generate additional samples, and then the PCE determination module is called for processing.

[0047] A result output module is configured to calculate statistical data of each response according to the determined response value and generate a result report.

[0048] In a third aspect, a computer device includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of the new energy power grid power transmission reliability margin probability evaluation method according to the first aspect of the present application.

[0049] In a fourth aspect, a computer storage medium stores a computer program, and the computer program, when executed by a processor, implements the steps of the new energy power grid power transmission reliability margin probability evaluation method according to the first aspect of the present application.

[0050] Beneficial effects: The present application provides an innovative solution for evaluating the probability of power transmission reliability margin in the power grid by combining the orthogonal polynomial approximation uncertainty quantification method. This method provides an accurate and efficient evaluation method for the uncertainty caused by renewable energy grid connection and load changes. Compared with traditional Monte Carlo simulation methods, the present application has the following advantages:

[0051] (1) Significantly improve the calculation efficiency: By using the adaptive sparse method to construct the orthogonal polynomial approximation, the present application significantly reduces the number of samples required for evaluating the probability of power transmission reliability margin, thereby significantly improving the calculation efficiency.

[0052] (2) Improve the evaluation accuracy: By combining the adaptive sparse method to construct the orthogonal polynomial approximation, the present application not only optimizes the calculation process, but also improves the evaluation accuracy. This method can more accurately capture and process uncertainty, ensuring that the evaluation results are more close to the actual network operation.

[0053] (3) Support decision making: By providing more accurate and efficient evaluation of the probability of power transmission reliability margin, the present application supports power grid planners and operators to make more scientific and reasonable decisions in the face of uncertainty. Especially in the context of increasing renewable energy, the present application provides strong technical support for the reliability and stability management of the power grid.

[0054] (4) Promote the integration of renewable energy: the present application helps to optimize the access and use of renewable energy by accurately assessing the probabilistic power transmission reliability margin of the power network. This not only improves the operation efficiency of the power grid, but also promotes the wider application of renewable energy and supports the green transformation of energy structure. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 is a flow chart of the method for evaluating the power transmission reliability margin of the new energy power grid.

[0056] Figure 2 is the P-V curve and power transmission reliability margin of the deterministic evaluation of the power transmission reliability margin of the renewable energy power grid in the embodiment.

[0057] Figure 3 is a schematic diagram of the cumulative distribution function (CDF) of the voltage out-of-limit reliability margin calculated by the MCS and orthogonal polynomial approximation method in the embodiment. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0059] The present application proposes a method for evaluating the power transmission reliability margin of the power grid under the uncertainty caused by the large-scale integration of renewable energy and load changes. Referring to Figure 1 , the method comprises the following steps:

[0060] Step S1: input the data of the power network, including the probability distribution and parameters of renewable energy and load;

[0061] The renewable energy considers wind power and solar power, and each uncertain input model can be represented as a random variable related to the probability density function (PDF) X~f X (x). For wind power, the statistical characteristics of wind speed v in the short term can be fitted by a normal distribution, and the corresponding probability density function is:

[0062]

[0063] In the formula, v is the wind speed, the expectation μ v and the variance can be obtained from historical data. For each realization of wind speed v, the corresponding wind power injected at the terminal bus can be calculated by the wind speed-power output relationship:

[0064]

[0065] In the formula, v in , v out and v rated are the cut-in, cut-out and rated wind speeds (m / s), respectively, and Pr For rated wind power, P w This refers to the active power output of wind power.

[0066] For solar power generation, the statistical characteristics of solar radiation r in the short term can also be adopted using a normal distribution, and the corresponding probability density function is:

[0067]

[0068] In the formula, r represents the actual solar radiation, and μ represents the expected solar radiation. r and variance This can be obtained from historical data; the solar radiation-power output relationship is expressed as:

[0069]

[0070] In the formula P pv r represents the power output of solar power generation. c For a given radiation point, it is typically set to 150W / m. 2 r std P represents the amount of solar radiation under standard conditions. rs This refers to the rated capacity of solar photovoltaic power.

[0071] To describe the uncertainty in load forecasting, the most common approach is to use a normal distribution. The active power of the load, P... L Average of the prediction results Typically provided by load forecasters, This represents the prediction variance. Load forecasters typically only provide active power data, while reactive power is determined under the assumption of a constant power factor. Load forecasters can use common load forecasting methods and can achieve high-resolution load forecasting based on the large amount of real-time load data provided by smart meters.

[0072] Step S2: For each random input, select an independent standard variable and its corresponding orthogonal polynomial;

[0073] Based on the probability distribution type of the random variable, select an appropriate orthogonal polynomial as the basis function to construct the surrogate model: In the formula, the random vector ξ=(ξ1,ξ2,...,ξ) n ), has n independent components; Ψ k (ξ) is related to f ξ Orthogonal multivariate polynomials ( Where the subscript i j f represents the j-th degree of the i-th univariate polynomial basis. ξ It is the joint probability density function of random vectors. It is about (orthogonal basis); c kare the coefficients of the orthogonal polynomials. In practical applications, the orthogonal polynomial approximation of the random response is truncated so that the total order is no higher than p (p is the total order of the orthogonal polynomials: ), thus the truncated surrogate model is where P is the number of retained terms.

[0074] The correspondence of common continuous probability distributions is shown in Table 1.

[0075] Table 1 Standard forms of classical continuous distributions and their corresponding orthogonal polynomials

[0076]

[0077] The multivariate orthogonal polynomials of the polynomial chaos expansion are constructed in Step 2, which are composed of the tensor product of univariate orthogonal polynomials. To illustrate the construction of the mixed multivariate orthogonal polynomials, take a two-dimensional input random vector ξ = (ξ1, ξ2) as an example, whose independent variable distribution functions are ξ1 ~ N(0, 1) and ξ2 ~ N(0, 1) respectively, then the multivariate polynomial corresponding to p ≤ 2 is:

[0078]

[0079] where φ1, φ2 are Hermite polynomials, thus the surrogate model Y = c0Ψ0(ξ) + … + c5Ψ5(ξ) based on orthogonal polynomial expansion is obtained. The subscript k and the index a k of the multivariate polynomial Ψ k k1 kn ki define the order j of the i-th univariate polynomial basis, that is, j = a ki .

[0080] Step S3: generating an experimental design, using Latin hypercube sampling to generate samples, then converting random variables in the standard normal space to the original input variable space, and using the probabilistic continuous power flow equation (PCPF) to obtain deterministic response values;

[0081] First, the Latin hypercube sampling method is used to generate M C samples in the standard normal space

[0082] Second, the variable conversion principle is used to convert the random variables ξ C in the standard normal space to the original input variable space u C : Here u C is the wind speed v, solar radiation r and active power load P L ,​​​ is the inverse cumulative distribution function of u C , φ is the cumulative distribution function of ξ C . By this inverse transformation, a sample set of ξ can be converted back to a sample set of u C , and then the converted sample u a is evaluated by PCPF to obtain the deterministic response value

[0083] The present application takes wind speed, solar radiation and load variation as random input variables, and then uses the power flow equation to determine the power transmission reliability margin of the renewable energy power grid. The power flow equation can be expressed as: where x = [θ b , θ c , θ a , V b , V c ] T , represents the voltage phase angle and amplitude of each phase; and represent the initial values of the active power and reactive power of node i at phase (such as phases a, b and c); and represent the calculated values of the active power and reactive power of node i at phase in f(x). Let v, r and P L be random vectors representing wind speed, solar radiation and load variation, then for P-V type nodes, the three-phase probabilistic continuous power flow (PCPF) equation corresponding to an N-node system can be expressed as:

[0084]

[0085] In the formula, is the voltage amplitude of node i at phase , and M is the number of phases; and are the conductance (real part) and susceptance (imaginary part) of the admittance matrix between nodes i and j at phase k and ; and are the cosine and sine of the voltage phase angle difference between nodes i and j at phase k and ; λ is the Lagrange multiplier used for handling the constraint conditions in the optimization problem; and are the actual power variations of wind power, solar photovoltaic, load and other types of distributed generators at bus i phase ; and These represent the changes in reactive power caused by wind speed changes and load changes, respectively. For node i in phase Reactive power generation; Q min,i and Q max,i These are the minimum and maximum values ​​of reactive power, respectively. The above equation can be expressed in compact form:

[0086] f(x,μ,λ,U)=f(x,μ)-λb(U)=0

[0087] In the formula, x is the state vector, μ is the control parameter vector such as the transformer tap ratio, and U = [v, r, P]. L Let be a random vector describing wind speed, solar radiation, and load active power. The load variation vector b of the system is:

[0088]

[0089] Therefore, the following formula for probabilistic power transmission reliability margin can be proposed:

[0090] maxλ

[0091] stf(x,μ)-λb(U)=0

[0092] V min ≤V i (x,μ,λ,U)≤V max

[0093]

[0094] In the formula V min and V max These are the lower and upper limits of the node voltage amplitude, respectively; I ij,max Let λ be the specified capacity of the line or transformer between node i and node j; λ is the normalized load margin under a given load variation vector. The maximum value of λ under the constraints of the above formula corresponds to the transmission reliability margin.

[0095] Step S4: Determine the best orthogonal polynomial approximation (PCE) for each response through an adaptive sparse process;

[0096] Response Y i The k orthogonal polynomial approximations are constructed through the following steps:

[0097] a) Generate k candidate multi-indicator sets, and use formula... Truncate them, where n is the number of random variable inputs, p is the total order of the orthogonal polynomial, this formula is the definition formula for the hyperbolic (q-norm) truncation scheme, a k It is a multi-indicator. It is the order of the univariate orthogonal polynomial truncated by the q-norm. The k-term multivariate polynomials obtained from the evaluation form the design matrix H. k =[H k-1 ΔH k ], H k-1 It contains the H matrix corresponding to the first k-1 polynomials; ΔH k It is the matrix corresponding to the newly added k-th polynomial, by ΔH k Add to H k-1 In the design matrix H k This includes all truncated polynomial basis function values ​​from the 0th term to the kth term.

[0098] b) Apply the Least Angle Regression (LAR) algorithm to select the optimal retention design matrix H k The coefficients of the set of random variables are estimated using the least squares method. The corresponding stochastic response calculated by deterministic methods The coefficients are then calculated by solving the following least-squares minimization problem: Establish a vector that minimizes the objective function. The formula for calculating the fitted value, Then, we use ordinary least squares to obtain the solution to the objective function in the above equation.

[0099] c) Using the formula ε cloo =ε loo ×T(M C ,P) Calculate the leave-one-out error index In the formula

[0100] h = diag[H(H) T H)- 1 H T ], It is the response vector y C empirical variance It uses a proxy model to calculate the value of variable ξ. (i) The response value obtained below, h i It is the i-th component of h; diag indicates extracting the diagonal elements of the matrix; tr represents the trace of the matrix, that is, the sum of the diagonal elements of the matrix. If the specified precision is achieved, i.e. or Then return the best orthogonal polynomial approximation, where ∈ is the specified precision threshold. Otherwise, increase the degree of the polynomial and return a).

[0101] Step S5: If the PCE of all responses meets the required accuracy, proceed to step S6; otherwise, proceed to step S3 to generate additional ΔM.C A new sample is then returned to step S4.

[0102] Step S6: Once the orthogonal polynomial approximation of all responses is obtained, it can be used to evaluate a large number of samples to determine the response values; the calculation is as follows:

[0103]

[0104] Step S7: Calculate the statistical data of each response and generate a result report.

[0105] The output responses of a large number of samples in step S6 are calculated by the polynomial expansion model, which are the available transmission capacity of the power grid. In step S7, the output data of the samples are statistically analyzed to obtain the statistical indicators and probability distribution curve of the available transmission capacity of the power grid, so as to calculate the system reliability margin at a certain confidence level, thereby providing a reference for power grid performance evaluation and planning.

[0106] In order to verify the performance of the method, the proposed method is applied to study the improved IEEE 13-node system, and the accuracy and performance of the method are verified based on the Latin hypercube sampling Monte Carlo simulation (MCS) as a benchmark. It is assumed that the wind speed, solar radiation and load power all obey the normal distribution. Here it is assumed that the mean value of the load power is equal to the original bus load power, and the standard deviation is equal to 5% of the mean value.

[0107] For simplicity, all wind turbines are equipped with the same v rate , v in and v out , which are 15.0, 4.0 and 25.0 (m / s), respectively. Similarly, it is assumed that all solar generators are allocated the same r C and r std , which are 150.0 and 1000.0 (W / m 2 ), respectively.

[0108] Two solar photovoltaic generators and two wind turbines are added to the IEEE 13-node system, and the total load is 1.733 MW and 1.051 Mvar, respectively. The solar photovoltaic rated power P r of buses 675 and 692 is 180 kW and 240 kW, respectively, and the wind turbine rated power P r of buses 680 and 634 is 450 kW and 300 kW, respectively. Since there are 8 single-phase loads in the feeder, the total random input is 12.

[0109] In order to study the influence of uncertainty on the transmission reliability margin of the system, the P-V curve and the transmission reliability margin of the renewable energy power grid (i.e. without uncertainty) are tracked by using the power flow equation. As shown inFigure 2 As shown, Figure 2 The three-phase P-V curves of the 675-bus system at the sampling points where all random inputs are equal to their average values (i.e., a deterministic system) are shown. Due to the imbalance of network parameters and loads, the P-V curves of different phases differ greatly. As the load increases in the predetermined direction, the C-phase voltage value decreases the fastest. Among the three transmission reliability margins (voltage out-of-limit reliability margin, thermal limit reliability margin, and voltage collapse reliability margin), the voltage out-of-limit reliability margin, which is 0.894 MW, is the smallest reliability margin.

[0110] Table 2 Comparison of statistical quantities of reliability margins under two calculation methods

[0111]

[0112] Table 2 gives the estimated mean and variance of the system transmission reliability margin by the MCS and the orthogonal polynomial approximation method (subscript pce), from which it can be seen that the orthogonal polynomial approximation method can accurately estimate the probabilistic statistical characteristics of the system transmission reliability margin.

[0113] Table 3 Comparison of calculation time under two calculation methods

[0114] Method t ed (s)]]> t sc (s)]]> t es (s)]]> t total (s)]]> PCE 8.713 0.934 0.213 9.860 MCS - - 2810.654 2810.654

[0115] Table 3 lists the experimental design time t ed , the coefficient solving time t sc , the statistical sample evaluation time t es , and the total time t total . Compared with the MCS, the orthogonal polynomial approximation method (PCE) spends negligible time on solving coefficients and evaluating statistical samples. Obviously, the orthogonal polynomial approximation method is more efficient than the MCS.

[0116] Once the statistical data of the probabilistic transmission reliability margin are obtained, the next direction is to study how the uncertainty affects the transmission reliability margin of the power grid.

[0117] As Figure 2 shown, the voltage out-of-limit reliability margin without considering uncertainty is 0.894 MW, and the corresponding probability at the CDF curve ( Figure 3 ) is 0.518, indicating that there is a 51.8% probability that the transmission reliability margin is less than or equal to 0.894 MW. In practical applications, such a low probability of safety is usually unacceptable. Therefore, in order to ensure that the system will not encounter voltage out-of-limit with 95% confidence, the voltage out-of-limit reliability margin must be reduced from 0.894 MW to 0.868 MW, corresponding to a decrease of 2.91%.

[0118] Compared with the traditional Monte Carlo simulation, the method greatly reduces the calculation workload, provides insights into how uncertainty affects the power transmission reliability margin of the power grid, and provides important information and new perspectives for the planning and operation of a power system containing renewable energy.

[0119] Based on the same technical concept as the method embodiment, the application also provides a new energy power grid power transmission reliability margin probability evaluation system, comprising:

[0120] A data input module is configured to input data of the power grid, including probability distribution and parameters of renewable energy and loads;

[0121] A orthogonal polynomial selection module is configured to select an independent standard variable and its corresponding orthogonal polynomial for each random input;

[0122] An experimental design module is configured to generate an experimental design, generate samples in a standard normal space using a Latin hypercube sampling method, then convert the random variables in the standard normal space to the original input variable space, evaluate the converted samples using a probabilistic continuous power flow equation (PCPF), and obtain deterministic response values;

[0123] A PCE determination module is configured to determine the best orthogonal polynomial approximation PCE for each response through an adaptive sparse process;

[0124] A response value determination module is configured to determine whether the accuracy of the orthogonal polynomial approximation of all responses meets the specified requirements, if yes, call the obtained best orthogonal polynomial approximation to evaluate a large number of specified samples to determine the response values, and if not, call the experimental design module to generate additional samples, and then go to the PCE determination module for processing;

[0125] A result output module is configured to calculate statistical data of each response according to the determined response values, and generate a result report.

[0126] It should be understood that the new energy power grid power transmission reliability margin probability evaluation system in the embodiment of the application can realize all the technical solutions in the above method embodiment, and the functions of each functional module can be realized according to the method in the above method embodiment, and the specific implementation process can be referred to the related description in the above embodiment, which will not be repeated here.

[0127] The application also provides a computer device, comprising: one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs are executed by the processor to realize the steps of the new energy power grid power transmission reliability margin probability evaluation method as described above.

[0128] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement steps of the new energy power grid power transmission reliability margin probability evaluation method.

[0129] Those skilled in the art should understand that embodiments of the present application can be provided as a method, device (system), computer device or computer program product. Therefore, the present application can adopt a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt a computer program product in the form of being implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes.

[0130] The present application is described with reference to flowcharts according to the method of embodiments of the present application. It should be understood that each flow in the flowcharts and the combination of the flows in the flowcharts can be realized by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, a special purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device that implements the functions specified in the flowcharts Figure 1 The device specified in one flow or multiple flows.

[0131] These computer program instructions can also be stored in a computer readable memory capable of guiding a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer readable memory produce a product including instruction devices, which implement the functions specified in the flowcharts Figure 1 The device specified in one flow or multiple flows.

[0132] These computer program instructions can also be loaded into a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the flowcharts Figure 1 The device specified in one flow or multiple flows.

Claims

1. A method for evaluating a reliability margin probability of a new energy power grid transmission, characterized in that, The method comprises the following steps: Step S1: inputting data of the electric network, including probability distribution and parameters of renewable energy and loads; Step S2: selecting an independent standard variable and its corresponding orthogonal polynomial for each random input; Step S3: generating an experimental design, generating samples in the standard normal space by using the Latin hypercube sampling method, then converting the random variables in the standard normal space to the original input variable space, evaluating the converted samples by using the probabilistic continuous power flow equation PCPF to obtain deterministic response values; Step S4: determining the best orthogonal polynomial approximation PCE of each response through an adaptive sparse process; Step S5: if the accuracy of the orthogonal polynomial approximation of all responses reaches the specified requirement, proceeding to Step S6, otherwise, proceeding to Step S3 to generate additional samples, and then proceeding to Step S4; Step S6: evaluating a large number of samples specified by using the obtained best orthogonal polynomial approximation to determine response values; Step S7: calculating statistical data of each response and generating a result report.

2. The method of claim 1, wherein, The step S1, renewable energy includes wind power and solar power, each uncertain input model is expressed as probability density function X ~ f X (x) related random variable, for wind power, the statistical characteristics of wind speed v in short term is fitted by normal distribution, and the corresponding probability density function is: In the formula, v is wind speed, expectation μ v And variance Obtained from historical data; For solar power generation, the statistical characteristics of solar radiation r in the short term also adopt normal distribution, and the corresponding probability density function is: In the formula, r is the actual solar radiation, the expectation μ r And the variance Obtained from historical data; Using a normal distribution to describe the uncertainty of the load forecast, the load active power P L the average value of the prediction result is provided by a load forecaster, denotes the prediction variance.

3. The method of claim 2, wherein, For each realization of the wind speed v, the corresponding wind power injected into the terminal bus is calculated according to the wind speed-power output relationship: where v in , v out and v rated are the cut-in, cut-out and rated wind speed, P r is the rated wind power, and P w is the wind power output active power; The solar radiation-power output relationship is expressed as: where P pv represents the power of solar power generation output, r c is a specified radiation point, r std is the solar radiation amount in a standard environment, P rs is the rated capacity of solar photovoltaics.

4. The method of claim 1, wherein, Step S2 includes: selecting appropriate orthogonal polynomials as basis functions based on the probability distribution type of the random variables, and constructing a surrogate model. In the formula, the random vector ξ=(ξ1,ξ2,…,ξ) n It has n independent components; Ψ k (ξ) is related to f ξ Orthogonal multivariate polynomials; c k These are the coefficients of an orthogonal polynomial; Where the subscript i j f represents the j-th degree of the i-th univariate polynomial basis. ξ It is the joint probability density function of random vectors. It is about The orthogonal basis.

5. The method of claim 4, wherein, The step S3 comprises generating M C samples in the standard normal space using a Latin hypercube sampling method Using the transformation principle of variables, the random variable ξ C in the standard normal space is converted to the original input variable space u C : Here u C represents the wind speed v, solar radiation r and the active power of the load P L , is the inverse cumulative distribution function of u C , and φ is the cumulative distribution function of ξ C . Evaluating the transformed sample u using the PCPF C to obtain a deterministic response value 6. The method of claim 5, wherein, v, r and P L are random vectors representing wind speed, solar radiation and load variation, respectively, and for a P-V type node, the three-phase probabilistic continuation power flow (PCPF) equation for a system of N nodes is expressed as: where, is the voltage amplitude of node i at phase M is the number of phases; and are the conductance and susceptance of the admittance matrix between nodes i and j at phase k and and are the cosine and sine of the voltage phase angle difference between nodes i and j at phase k and λ is the Lagrange multiplier for the handling of the constraints in the optimization problem; and are the actual power variations of the wind, solar photovoltaic, load and other types of distributed generators at bus i phase and are the reactive power variation amounts due to the changes in wind speed and load; is the reactive power generation of node i at phase Q min,i and Q max,i are the minimum and maximum values of the reactive power; the above equations are expressed in a compact form:​​ f(x, μ, λ, U) = f(x, μ) - λb(U) = 0 where x is the state vector, μ is the control parameter vector, U = [v, r, P L ] is the random vector describing the wind speed, solar radiation, and active power of the load; the load generation variation vector b of the system is: Then the optimization problem of the power transmission reliability margin probability is expressed as: max λ s.t. f(x, μ) - λb(U) = 0 V min ≤V i (x,μ,λ,U)≤V max where V min and V max are the lower and upper limits of the node voltage magnitude, respectively. I ij,max Ci,j is the prescribed capacity for the line or transformer between node i and node j; λ is the normalized load margin under the given load variation vector, and the maximum value of λ corresponding to the power transmission reliability margin under the condition of satisfying the above constraint.

7. The method of claim 5, wherein, In the step S4, the response Y i The kth orthogonal polynomial approximation is constructed by the following steps: S41, generating k candidate multi-index sets, and using formula cutting them, where n is the number of random variable inputs, p is the total order of orthogonal polynomials, a k is a multiple index, is the order of single-variable orthogonal polynomials under q-norm truncation; evaluating the obtained k-term multivariate polynomial to form a design matrix H k =[H k-1 , ΔH k ], ΔH k is the matrix corresponding to the newly added kth polynomial; S42, select the optimal reserved design matrix H using the least angle regression (LAR) algorithm k least squares to estimate the corresponding coefficients for the random variables corresponding random response computed from the deterministic method The coefficients are computed by solving a least squares minimization problem: S43, calculate leave-one-out error indicator with formula ε cloo = ε loo × T(M C , P) where h = diag[H(H T H) -1 H T ], is empirical variance of response vector y C , h (i) is response value obtained at variable ξ i using proxy model, h The method comprises the following steps: is i-th component of h; diag denotes extracting diagonal elements of matrix; tr denotes trace of matrix; if error reaches specified accuracy, return best orthogonal polynomial approximation, otherwise increase degree of polynomial, return S41.

8. A new energy power grid transmission reliability margin probability evaluation system, characterized in that, It comprises: a data input module configured to input data of the electric network, including probability distribution and parameters of renewable energy and loads; an orthogonal polynomial selection module configured to select an independent standard variable and its corresponding orthogonal polynomial for each random input; an experimental design module configured to generate an experimental design, generate samples in the standard normal space by using the Latin hypercube sampling method, then convert the random variables in the standard normal space to the original input variable space, evaluate the converted samples by using the probabilistic continuous power flow equation PCPF to obtain deterministic response values; a PCE determination module configured to determine the best orthogonal polynomial approximation PCE of each response through an adaptive sparse process; a response value determination module configured to judge whether the accuracy of the orthogonal polynomial approximation of all responses reaches the specified requirement, if yes, call the best orthogonal polynomial approximation obtained to evaluate a large number of samples specified to determine response values, if no, call the experimental design module to generate additional samples, and then proceed to the PCE determination module for processing; a result output module configured to calculate statistical data of each response according to the determined response values and generate a result report.

9. A computer device, comprising: The device comprises one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of the power transmission reliability margin probability evaluation method of the new energy power grid according to any one of claims 1-7.

10. A computer storage medium having stored thereon a computer program, characterized in that The computer program is executed by a processor to implement the steps of the new energy power grid power transmission reliability margin probability evaluation method according to any one of claims 1-7.

Images