An alternating current-direct current hybrid power grid risk assessment method, device, equipment and medium

Abstract

The application discloses an AC-DC hybrid power grid risk assessment method, device, equipment and medium. According to the historical data of the random input variables of a power system, a first optimization model minimizing the difference between the score moments of the input variables and the estimated values is established; by solving the first optimization model, the optimal score moment proportion parameters and weight adjustment parameters are obtained, and a plurality of sample points and their weight parameters are obtained according to the solving result; the score moments of the random output variables are calculated according to the plurality of sample points and their weight parameters; the probability density result when the information entropy of the random output variables is maximized is calculated by taking the score moments of the random output variables as constraints, and the probability density result is reconstructed; according to the probability density result, the probability of the bus voltage and line flow of the power system exceeding the limit is calculated, and the operation state of the power system is evaluated. The application improves the efficiency and accuracy of the probability flow calculation, thereby improving the accuracy of the power grid risk assessment.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, apparatus, equipment and medium for risk assessment of AC / DC hybrid power grids. Background Technology

[0002] In traditional power system analysis, factors such as load fluctuations, changes in grid operation modes, and generator outages already introduce a certain degree of uncertainty into the power system. However, with advancements in the power industry and the incorporation of new energy sources like solar and wind power, the intermittency and randomness of the power grid have significantly increased. Simultaneously, the rise of new concepts such as microgrids, distributed generation, and electric vehicles has greatly enhanced the interaction between power sources, loads, and the grid. This directly leads to a significant increase in power system uncertainty, rendering deterministic power flow-based risk assessment methods infeasible. Therefore, researching probabilistic power flow algorithms for power system analysis has become increasingly crucial and an indispensable tool for understanding and addressing power system risk assessment.

[0003] However, existing technologies still have the following shortcomings when it comes to risk assessment of large AC / DC systems: 1. They cannot complete probabilistic power flow calculations within the required timeframe; 2. The accuracy of probabilistic power flow calculation results is low; 3. They cannot accurately assess the operational risks of the power system using probabilistic power flow calculation results. These shortcomings result in the inability to obtain accurate power system risk assessment results within a reasonable timeframe. Summary of the Invention

[0004] This invention provides a method, apparatus, equipment, and medium for risk assessment of AC / DC hybrid power grids, to solve the technical problem of how to obtain accurate probabilistic power flow calculation results within an effective time frame for assessing power system risks, thereby improving the efficiency and accuracy of power system risk assessment.

[0005] To address the aforementioned technical problems, in a first aspect, embodiments of the present invention provide a risk assessment method for AC / DC hybrid power grids, comprising:

[0006] Historical data of random input variables of the power system are obtained. Based on the historical data, a first optimization model is established with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective.

[0007] Solve the first optimization model to obtain the optimal fractional moment ratio parameter and weight adjustment parameter, and obtain a preset number of sample points of the random input variable and their weight parameters based on the optimal fractional moment ratio parameter and weight adjustment parameter.

[0008] Based on the definition of fractional moments, the fractional moments of the random output variables of the sample points are obtained according to the sample points and their weight parameters of the random input variables.

[0009] Using the fractional moments of the random output variables of the sample points as constraints, calculate the probability density result when the information entropy of the random output variables is maximized, and reconstruct the probability density result;

[0010] Based on the probability density results, the probability of the power system bus voltage and line power flow exceeding the limits is calculated. When the probability exceeds a preset value, the power system is confirmed to be in a dangerous state.

[0011] Compared with existing technologies, the above embodiments have the following beneficial effects: introducing fractional moments into the probabilistic power flow calculation of AC / DC hybrid systems improves the sufficiency of obtaining probabilistic information of random input variables, and since a few fractional moments contain the aforementioned probabilistic information, the calculation time is greatly reduced; in addition, by establishing a first optimization model and adaptively generating high-quality sample points based on the calculation results of the first optimization model, a large amount of probabilistic information is propagated through a small number of sample points, which improves the calculation accuracy of probabilistic power flow while ensuring computational efficiency; by reconstructing the probabilistic power flow calculation results through the fractional moments of random output variables, the accuracy of the probability density results is improved, thereby improving the accuracy of the risk assessment results.

[0012] In one embodiment of the first aspect, historical data of random input variables of the power system are obtained, and based on the historical data, a first optimization model is established with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective, including:

[0013] Calculate the fractional moments of the random input variables based on the historical data of the random input variables;

[0014] Based on the sample points and weight parameters of the random input variable, calculate the estimated value of the fractional moments of the random input variable;

[0015] The first optimization model is established with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective.

[0016] Compared with the prior art, the above embodiments have the following beneficial effects: Since the estimated value of the fractional moments of random input variables is related to the collected sample points and weight parameters, in order to obtain high-quality sampling points, the optimal fractional moment ratio parameter and weight adjustment parameter can be obtained by minimizing the difference between the fractional moments of random input variables and their estimated values.

[0017] In one embodiment of the first aspect, calculating the fractional moment of the random input variable based on historical data of the random input variable includes:

[0018] The formula for calculating the fractional moments of the random input variables is:

[0019]

[0020] Where, n sam The number of historical data points for the random input variable is represented by α; α is the fractional order of the random input variable; x α Let E[·] be the Taylor expansion of the random input variable near its mean; E[·] is the expectation operator; For the i-th random input variable sam Historical data.

[0021] Compared with the prior art, the above embodiments have the following beneficial effects: by calculating the fractional moments of the random input variables using real historical data, the accuracy of the fractional moments of the random input variables can be improved.

[0022] In one embodiment of the first aspect, calculating the estimated value of the fractional moment of the random input variable based on the sample points and weight parameters of the random input variable includes:

[0023] The formula for calculating the fractional moment estimate of the random input variable is as follows:

[0024]

[0025] Where, ε (fm) and β (fm) These are the optimal scaling parameters and the optimal weight adjustment parameters for the fractional moments, respectively. The fractional moment estimate of the random input variable; For the weight parameters of the j-th sample point of the random input variable; Let j be the j-th sample point of the random input variable.

[0026] Compared with the prior art, the above embodiments have the following beneficial effects: by estimating the estimated value of the fractional moments of the random input variable through the sample points and weight parameters of the random input variable, the sample point information is introduced into the optimization model, thereby improving the quality of subsequent sample point acquisition.

[0027] In one embodiment of the first aspect, establishing a first optimization model with the difference between the fractional moments of the random input variable and its estimated value as the optimization objective includes:

[0028] The first optimization model is as follows:

[0029]

[0030] Where, ε (fm) and β (fm) These are the optimal scaling parameters and the optimal weight adjustment parameters for the fractional moments, respectively. The fractional moment estimate of the random input variable; For the weight parameters of the j-th sample point of the random input variable; Let be the j-th sample point of the random input variable; n is the upper limit of the value of a, and s is the interval between the values ​​of a.

[0031] Compared with the prior art, the above embodiments have the following beneficial effects: by establishing an optimization model with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective, the optimal sample point generation parameters can be obtained, thereby improving the efficiency and accuracy of subsequent power flow probability calculations.

[0032] In one embodiment of the first aspect, obtaining a preset number of sample points of the random input variable and their weight parameters based on the optimal fractional moment ratio parameter and weight adjustment parameter includes:

[0033] The formula for calculating the sample points of the random input variable is:

[0034]

[0035] The formula for calculating the weight parameters of the sample points of the random input variable is as follows:

[0036]

[0037] in, and These are the weight parameters for the j-th sample point and the j-th sample point of the random input variable, respectively; m X C represents the mean of the sample points of the random input variable; XX is the covariance of the random input variable; w0 is the initial weight parameter; d represents the dimension of the random input variable.

[0038] Compared with the prior art, the above embodiments have the following beneficial effects: by adding optimizable fractional moment ratio parameters and weight adjustment parameters to the sampling points and weight parameter calculation formulas, high-quality sampling points can be adaptively obtained, and only a small number of sample points are needed to propagate a large amount of probability information, thereby improving the calculation accuracy while maintaining the calculation efficiency.

[0039] In one embodiment of the first aspect, calculating the probability of exceeding the limits of the power system bus voltage and line power flow based on the probability density results includes:

[0040] The formula for calculating the probability of the random output variable exceeding the limit is:

[0041]

[0042] Where, p pro (·) represents the probability of exceeding the limits of the power system bus voltage and line power flow; fY (Y _out ) is the probability density function of the random output variable; This represents the allowed range of random output variables.

[0043] Compared with the prior art, the above embodiments have the following beneficial effects: by using the probabilistic power flow calculation results, the risk information of the power system is fully revealed, and the accuracy of power system risk assessment is effectively improved.

[0044] In one embodiment of the first aspect, obtaining the fractional moments of the random output variables of the sample points based on the definition of fractional moments, according to the sample points and weight parameters of the random input variables, includes:

[0045] The sample points and their weight parameters are sequentially input into the preset deterministic power flow calculation model of the DC-DC hybrid system to obtain the random output variable results of the sample points;

[0046] Calculate the fractional moments of the random output variables based on the random output variable results and weight parameters for all the sample points.

[0047] Compared with the prior art, the above embodiments have the following beneficial effects: by calculating the fractional moments of the random output variables using data from all sample points, the reliability of the obtained fractional moments is improved while ensuring high efficiency.

[0048] In one embodiment of the first aspect, the step of calculating the probability density result when the information entropy of the random output variable is maximized, constrained by the fractional moments of the random output variable of the sample points, and reconstructing the probability density result, includes:

[0049] Using the fractional moments of the random output variables as constraints, an information entropy maximization model is established to calculate the probability density results of the random output variables;

[0050] The information entropy maximization model is transformed into an unconstrained term optimization model. The unconstrained term optimization model is solved to obtain the Lagrange multiplier vector and fractional vector of the random output variable.

[0051] Based on the Lagrange multiplier vector and fractional vector of the random output variable, reconstruct the probability density result of the random output variable.

[0052] Compared with the prior art, the above embodiments have the following beneficial effects: based on the fractional moments of the random input variables, the probability flow calculation results are reconstructed, improving the accuracy of the probability density results, thereby improving the accuracy of the risk assessment results.

[0053] Secondly, embodiments of the present invention also provide a risk assessment device for AC / DC hybrid power grids, the risk assessment device for AC / DC hybrid power grids comprising: a first optimization module, a sample point sampling module, a fractional moment calculation module, a probability density function calculation module, and a probability power flow over-limit alarm module;

[0054] The first optimization module is used to acquire historical data of random input variables of the power system, and establish a first optimization model based on the historical data, with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective.

[0055] The sample point sampling module is used to solve the first optimization model, obtain the optimal fractional moment ratio parameter and weight adjustment parameter, and obtain a preset number of sample points and their weight parameters of the random input variable based on the optimal fractional moment ratio parameter and weight adjustment parameter.

[0056] The fractional moment calculation module is used to obtain the fractional moment of the random output variable of the sample point based on the definition of fractional moment and according to the sample points and weight parameters of the random input variable;

[0057] The probability density function calculation module is used to calculate the probability density result when the information entropy of the random output variable is maximized, with the fractional moments of the random output variable of the sample point as constraints, and to reconstruct the probability density result.

[0058] The probability power flow over-limit alarm module is used to calculate the probability of the power system bus voltage and line power flow exceeding the limit based on the probability density result, and to confirm that the power system is in a dangerous state when the probability exceeds a preset value.

[0059] In one embodiment of the second aspect, the first optimization module is configured to acquire historical data of random input variables of the power system, and based on the historical data, establish a first optimization model with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective, including:

[0060] Based on the definition of fractional moments, the fractional moments of the random output variables of the sample points are obtained according to the sample points and their weight parameters of the random input variables.

[0061] Using the fractional moments of the random output variables of the sample points as constraints, calculate the probability density result when the information entropy of the random output variables is maximized, and reconstruct the probability density result;

[0062] Based on the probability density results, the probability of the power system bus voltage and line power flow exceeding the limits is calculated. When the probability exceeds a preset value, the power system is confirmed to be in a dangerous state.

[0063] In one embodiment of the second aspect, calculating the fractional moment of the random input variable based on historical data of the random input variable includes:

[0064] The formula for calculating the fractional moments of the random input variables is:

[0065]

[0066] Where, n sam The number of historical data points for the random input variable is represented by α; α is the fractional order of the random input variable; x α Let E[·] be the Taylor expansion of the random input variable near its mean; E[·] is the expectation operator; For the i-th random input variable sam Historical data.

[0067] In one embodiment of the second aspect, calculating the estimated value of the fractional moment of the random input variable based on its weight parameters from sample points includes:

[0068] The formula for calculating the fractional moment estimate of the random input variable is as follows:

[0069]

[0070] Where, ε (fm) and β (fm) These are the optimal scaling parameters and the optimal weight adjustment parameters for the fractional moments, respectively. The fractional moment estimate of the random input variable; For the weight parameters of the j-th sample point of the random input variable; Let j be the j-th sample point of the random input variable.

[0071] In one embodiment of the second aspect, establishing a first optimization model with the difference between the fractional moments of the random input variable and its estimated value as the optimization objective includes:

[0072] The first optimization model is as follows:

[0073]

[0074] Where, ε (fm) and β (fm) These are the optimal scaling parameters and the optimal weight adjustment parameters for the fractional moments, respectively. The fractional moment estimate of the random input variable; For the weight parameters of the j-th sample point of the random input variable; Let be the j-th sample point of the random input variable; n is the upper limit of the value of a, and s is the interval between the values ​​of a.

[0075] In one embodiment of the second aspect, the sample point sampling module is used to obtain a preset number of sample points of the random input variable and their weight parameters according to the optimal fractional moment ratio parameter and weight adjustment parameter, including:

[0076] The formula for calculating the sample points of the random input variable is:

[0077]

[0078] The formula for calculating the weight parameters of the sample points of the random input variable is as follows:

[0079]

[0080] in, and These are the weight parameters for the j-th sample point and the j-th sample point of the random input variable, respectively; m x C represents the mean of the sample points of the random input variable; XX is the covariance of the random input variable; w0 is the initial weight parameter; d represents the dimension of the random input variable.

[0081] In one embodiment of the second aspect, the probabilistic power flow overload alarm module is used to calculate the probability of the power system bus voltage and line power flow exceeding limits based on the probability density result, including:

[0082] The formula for calculating the probability of the random output variable exceeding the limit is:

[0083]

[0084] Where, p pro (·) represents the probability of exceeding the limits of the power system bus voltage and line power flow; f Y (Y _out ) is the probability density function of the random output variable; This represents the allowed range of random output variables.

[0085] In one embodiment of the second aspect, the fractional moment calculation module includes: a power flow calculation unit and a fractional moment calculation unit:

[0086] The power flow calculation unit is used to sequentially input the sample points and their weight parameters into the preset deterministic power flow calculation model of the DC-DC hybrid system, and obtain the random output variable results of the sample points.

[0087] The fractional moment calculation unit is used to calculate the fractional moment of the random output variable based on the random output variable results of all the sample points and the weight parameters.

[0088] In one embodiment of the second aspect, the probability density function calculation module includes: a constrained optimization unit, an unconstrained optimization unit, and a probability density reconstruction unit.

[0089] The constrained optimization unit is used to establish an information entropy maximization model with the fractional moments of the random output variable as constraints, and to calculate the probability density result of the random output variable.

[0090] The unconstrained term optimization unit is used to transform the information entropy maximization model into an unconstrained term optimization model, solve the unconstrained term optimization model, and obtain the Lagrange multiplier vector and fractional vector of the random output variable.

[0091] The probability density reconstruction unit is used to reconstruct the probability density result of the random output variable based on the Lagrange multiplier vector and the fractional vector of the random output variable.

[0092] Thirdly, embodiments of the present invention also provide a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the above-described AC / DC hybrid power grid risk assessment method.

[0093] Fourthly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when running, controls the device containing the computer-readable storage medium to execute the aforementioned AC / DC hybrid power grid risk assessment method. Attached Figure Description

[0094] Figure 1 This is a flowchart illustrating an embodiment of a risk assessment method for AC / DC hybrid power grids provided by the present invention.

[0095] Figure 2 This is another flowchart illustrating an embodiment of the AC / DC hybrid power grid risk assessment method provided by the present invention.

[0096] Figure 3 An improved IEEE 118-node system is an embodiment of the performance evaluation method for a risk assessment method of AC / DC hybrid power grids provided by the present invention.

[0097] Figure 4 The performance evaluation results diagram of the AC / DC hybrid power grid risk assessment method provided by the present invention is shown.

[0098] Figure 5This is a schematic diagram of one embodiment of a risk assessment device for AC / DC hybrid power grids provided by the present invention. Detailed Implementation

[0099] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0100] Example 1

[0101] Please refer to Figure 1 A risk assessment method for AC / DC hybrid power grids provided in this embodiment of the invention includes:

[0102] S101: Obtain historical data of random input variables of the power system, and establish a first optimization model based on the historical data, with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective;

[0103] S102: Solve the first optimization model to obtain the optimal fractional moment ratio parameter and weight adjustment parameter, and obtain a preset number of sample points of the random input variable and their weight parameters based on the optimal fractional moment ratio parameter and weight adjustment parameter.

[0104] S103: Based on the definition of fractional moments, obtain the fractional moments of the random output variables of the sample points according to the sample points and their weight parameters of the random input variables;

[0105] S104: Using the fractional moments of the random output variables of the sample points as constraints, calculate the probability density result when the information entropy of the random output variables is maximized, and reconstruct the probability density result;

[0106] S105: Based on the probability density results, calculate the probability of the power system bus voltage and line power flow exceeding the limits. When the probability exceeds a preset value, confirm that the power system is in a dangerous state.

[0107] Preferred, combined Figure 2 The above embodiments can also be implemented by the following methods:

[0108] Step 1: Collect historical data: Obtain historical data records of random sources (such as wind power, photovoltaic power and load power), and estimate the probability distribution and correlation coefficient of the input random sources;

[0109] Step 2: Perform probabilistic power flow calculation for AC / DC hybrid power grid: Organize the control modes of the DC system of VSC and establish a probabilistic power flow calculation model for the AC / DC hybrid system.

[0110] Step 3: Determine the accurate fractional order: Evaluate the optimal fractional order of the random source, and then calculate the exact fractional moments for each input random source;

[0111] Step 4: Based on the AC / DC hybrid power grid risk assessment method (ASUT) provided in this embodiment of the invention, perform probabilistic power flow (PPF) calculation: select a weighted sample point set, perform probabilistic power flow calculation, and evaluate to obtain the probability density function (PDF);

[0112] Step 5: Power system operation status assessment: Based on the probabilistic power flow settlement results, calculate the probability of node voltage and line power flow exceeding limits. When the probability value exceeds 5%, the power system is determined to be in an unstable operating state.

[0113] In one embodiment, the step of acquiring historical data of random input variables of the power system and establishing a first optimization model based on the historical data, with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective, specifically includes: calculating the fractional moments of the random input variables based on the historical data; calculating the estimated values ​​of the fractional moments of the random input variables based on the sample points and weight parameters of the random input variables; and establishing a first optimization model with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective.

[0114] Preferably, the step of calculating the fractional moment of the random input variable based on the historical data of the random input variable can be obtained through the following steps:

[0115] Given that x is a positive random variable, the fractional moment of x is defined as follows:

[0116] E[x α ]=∫ x x α f(x)dx (1)

[0117] Where E[·] represents the expectation operator; α is of fractional order and is a real number; x α The Taylor series expansion around its mean x0 is as follows:

[0118]

[0119] The formula for calculating the binomial coefficient is as follows:

[0120]

[0121] Combining formulas (1) to (3), we can obtain:

[0122]

[0123] Where, E[(x-x0)] i ] represents the i-th central moment of variable x.

[0124] If the fractional order *a* is an integer, then the infinite series in equation (2) becomes a finite series. Since the binomial coefficient in equation (3) is 0 when *i* > *a*, that is:

[0125]

[0126] In summary, it can be seen that the i-th integer moment only represents information about a few central moments, but when the fractional order a is not an integer, the binomial coefficient is not zero, that is:

[0127]

[0128] As can be seen from formula (2), theoretically, a single fractional moment encompasses a large amount of information about the central moments. From this perspective, compared with a finite number of integer moments, only one or two fractional moments are sufficient to fully represent the information of the PDF.

[0129] Therefore, in practical power systems, the fractional moments of these random sources can be conveniently calculated based on the collected historical records of these random sources. These fractional moments can be obtained using the following formula:

[0130]

[0131] Where nsam represents the number of historical data records; This represents the isamth historical data point.

[0132] The following beneficial effects can be obtained by combining the above embodiments: First, a small number of fractional moments can provide rich probability information, which is conducive to reproducing accurate probability density functions; second, high-order integer moments with high variability are difficult to estimate accurately, while fractional moments can alleviate this problem to a certain extent, because low-order fractional moments are more stable and easier to estimate accurately.

[0133] Preferably, the step of calculating the estimated value of the fractional moments of the random input variable based on the sample points and weight parameters of the random input variable, and establishing a first optimization model with the difference between the fractional moments of the random input variable and its estimated value as the optimization objective, can be performed through the following steps:

[0134] To achieve higher computational speed, this embodiment uses 2d+1 samples, where d is the dimension of the input variable. The formulas for calculating the sample points and weight parameters used to evaluate the fractional moments are as follows:

[0135]

[0136]

[0137] in, For sample points; ε is the weighting parameter; (fm) and β (fm) These represent the scaling parameter and weight adjustment parameter of the fractional moments, respectively, used to control the position and weight of the sample points; m X C represents the mean of the sample points of the random input variable; XX is the covariance of the random input variable; w0 is the initial weight parameter.

[0138] Fractional moment proportionality parameter ε (fm) and weight adjustment parameter β (fm) This can be obtained by evaluating the fractional moments of each random input variable, which differs from directly assigning so-called empirical parameters in the SUT. Therefore, this method can be called an adaptive scaling unscented variation method.

[0139] Establish the following optimization model:

[0140]

[0141] in, Let x be a random variable of order α. j The exact fractional moment; Let be the estimated value of the fractional moment; where the formula for calculating the estimated value of the fractional moment is Equation (11).

[0142]

[0143] Combining formulas (7) to (11), the optimization model (10) is solved using any solver to obtain the optimal ε. (fm) and β (fm) and obtain the optimal ε (fm) and β (fm) Substitute these values ​​into equations (8) and (9) to obtain high-quality sample points and their weight parameters.

[0144] The following beneficial effects can be obtained by combining the above embodiments: fractional moment proportionality parameter ε (fm) and weight adjustment parameter β (fm) The optimal ε can be obtained by evaluating the fractional moments of each random input variable and by calculation. (fm) and β (fm) This method obtains high-quality sample points, which differs from the direct allocation of so-called empirical parameters in the scaling unscented transformation algorithm. It can adaptively adjust the position of sample points to obtain high-quality sample points and improve the accuracy of probabilistic power flow analysis.

[0145] In one embodiment, the step of obtaining the fractional moments of the random output variables of the sample points based on the definition of fractional moments and according to the sample points and their weight parameters of the random input variables specifically includes: sequentially inputting the sample points and their weight parameters into a preset deterministic power flow calculation model of a DC-DC hybrid system to obtain the random output variable results of the sample points; and calculating the fractional moments of the random output variables based on the random output variable results of all the sample points and the weight parameters.

[0146] Preferably, the above embodiments can be performed through the following steps:

[0147] The sample points obtained through formulas (7) to (11) are input into the following AC / DC hybrid power grid deterministic power flow models (12) to (13) for probabilistic power flow calculation:

[0148]

[0149]

[0150] Among them, X _in With Y _ou t represents the input variable and the output variable, respectively; For a deterministic power flow model of an AC / DC hybrid power grid; P load and Q load These represent the active power and reactive power of the load, respectively; P ren and Q ren These represent the active and reactive power outputs of renewable energy sources, respectively; U ac and U dc These represent the node voltages of the AC and DC systems, respectively; S ac_ij and Q represents the apparent power of the AC branch and the active power of the DC branch, respectively; conver This indicates the reactive power provided by the VSC.

[0151] Input the obtained sample points into the above deterministic power flow model of the DC hybrid power grid to obtain the probabilistic power flow calculation results. Combined with formula (14), obtain the fractional moments of the probabilistic power flow calculation results, specifically:

[0152]

[0153] Among them, y r，j In the probabilistic power flow calculation results, the score ends when the j-th sample point L of the r-th random output variable is reached. It represents the i-th fractional order of the r-th random output variable.

[0154] In combination with the above embodiments, the following beneficial effects can be obtained: The probabilistic power flow calculation method proposed in this invention only requires 2d+1 sample points, and therefore only 2d+1 deterministic power flow calculations are needed to obtain reliable probabilistic power flow response fractional moments, thereby improving the efficiency of probabilistic power flow calculation.

[0155] In one embodiment, the step of calculating the probability density result of maximizing the information entropy of the random output variable using the fractional moments of the random output variable of the sample points as constraints, and reconstructing the probability density result, specifically includes: establishing an information entropy maximization model using the fractional moments of the random output variable as constraints, and calculating the probability density result of the random output variable; transforming the information entropy maximization model into an unconstrained optimization model, solving the unconstrained optimization model, and obtaining the Lagrange multiplier vector and fractional vector of the random output variable; and reconstructing the probability density result of the random output variable based on the Lagrange multiplier vector and fractional vector of the random output variable.

[0156] Preferably, the above embodiments can be performed through the following steps:

[0157] First, the probability density function of the probabilistic power flow calculation result is estimated using formula (15), specifically:

[0158]

[0159] in, This represents the r-th output variable. The value of the fractional moment of order f; Y (y r ) represents the probability density function of the r-th output variable; The nth calculation result y represents the probabilistic power flow. r Information entropy.

[0160] Based on the calculation result of formula (15), it can be written in the following form:

[0161]

[0162] Next, the constrained problem (15) is transformed into an unconstrained problem:

[0163]

[0164] Where, λ r Represents the Lagrange multiplier vector; α r This represents a fractional vector.

[0165] The optimization problem (17) is solved using a solver to obtain the optimal λ. r and α rSubstitute it into formula (16) to obtain the result of the reconstructed probability density function.

[0166] Finally, based on the reconstructed density function results, substitute them into formula (18) to assess the probability of exceeding the limits of the power system bus voltage and line power flow:

[0167]

[0168] Where, p pro (·) represents the probability that the probabilistic power flow calculation result exceeds the limit; f Y (Y _ou t) represents the probabilistic power flow density; Represents Y _out The permissible range.

[0169] Combining the above embodiments, the following beneficial effects can be obtained: even if there are a large number of random sources in the historical data, the operational risks of such complex AC / DC hybrid power grids can be fully revealed, effectively improving the accuracy and efficiency of power system risk assessment.

[0170] Example 2

[0171] Based on the AC / DC hybrid power grid risk assessment method described in the above embodiments, this invention also provides a performance evaluation method for verifying the AC / DC hybrid power grid risk assessment method described in the above embodiments. For ease of description, this embodiment uses ASUT to represent the AC / DC hybrid power grid risk assessment method described in the above embodiments, specifically as follows:

[0172] In one embodiment, the experimental parameters required for performance evaluation are obtained as follows:

[0173] refer to Figure 3 Table 1 shows the improved IEEE 118-node system architecture of the VSC-based DC grid used in this embodiment to verify the effectiveness of the AC / DC hybrid power grid risk assessment method described in the above embodiments; Table 2 shows the control method and parameters of VSCS in DC system 1; Table 2 shows the control method and parameters of VSCS in DC system 2.

[0174] Table 1

[0175]

[0176] Table 2

[0177]

[0178] in, Figure 3In this embodiment, WF1, WF2, and WF3 are integrated into the AC system via a three-terminal DC system, while a four-terminal DC system is used for grid connection of wind farms WF4, WF5, WF6, WF7, WF8, and WF9. It is assumed that the capacity of each wind farm in DC systems 1 and 2 is 100MW and 150MW, respectively. The power output data used in this embodiment comes from historical records of actual offshore wind farms in the China Southern Power Grid. The correlation coefficient between the power output of offshore wind farms WF1, WF2, and WF3 is 0.7; the correlation coefficient between the power output of WF4, WF5, and WF6 is 0.55; and the correlation coefficient between the power output of WF7, WF8, and WF9 is 0.5.

[0179] In reality, the entire power grid typically has different power supply zones. Therefore, it is assumed that the test system includes three different power supply zones: Zone A (nodes 1-51), Zone B (nodes 52-96), and Zone C (nodes 97-118). To demonstrate the performance of ASUT, four operating scenarios are defined as follows.

[0180] S1: In the AC / DC power grid, only the power generation from offshore wind farms is considered a random source. Therefore, there are 9 random sources in operational scenario 1.

[0181] S2: The power generation of the offshore wind farm and the load in Area A are considered as random sources. The load follows a normal distribution with a correlation coefficient of 0.2. There are 49 random sources in operational scenario 2.

[0182] S3: Random sources include 9 offshore wind farms and 99 loads in zones A, B, and C. The number of random sources increases to 108. The loads still follow a normal distribution, and the correlation coefficient between loads is set to 0.2.

[0183] S4: The number of random sources remains at 108, including 9 offshore wind farms and 99 loads. Assume that the load in area A follows a Weibull distribution; the load in area B follows a beta distribution; and the load in area C follows a log-normal distribution. Furthermore, the correlation coefficients differ across areas; assuming the correlation coefficients between the loads in areas A, B, and C are 0.45, 0.35, and 0.05, respectively.

[0184] As the operational scenario changes, the dimensionality of the random source increases, the correlation between random variables becomes more complex, and the distribution types become more diverse. The load handling method is as follows: First, the load x... i,raw The original data was generated using the probability density given in Table 3, and then processed through x. i,sam =x i,raw ×10%m i,mat +m i,mat Get the load x i,sam The data, where m i,matThis is the active power value of the load in Matpower 7.0.

[0185] Table 3

[0186]

[0187] In one embodiment, the specific method for evaluating the validity verification of ASUT based on the MCS method is as follows:

[0188] The MCS method is used to obtain a reference calculation of the probabilistic power flow (PPF) to verify the effectiveness of the ASUT. This embodiment uses 10 5 The probability density function (PDF) of the PPF is obtained using the MCS method for each sample. Note that the errors in the tables or figures in this embodiment are expressed as relative errors (percentages) compared to the reference values. The fractional moments of active power loss for the entire AC / DC hybrid system are shown in Table 4. It can be observed that the fractional moments of active power loss provided by ASUT have a considerably smaller relative error compared to MCS. The maximum and average relative errors are 1.7952% and 1.3042%, respectively, indicating that the ASUT method has good accuracy and adaptability in these four different operating scenarios.

[0189] Table 4

[0190]

[0191] refer to Figure 4 The results show the PDF comparison of active power loss of AC / DC hybrid systems under four operating scenarios. Since the PDF of active power loss of AC / DC grid determined by ASUT is consistent with the results obtained by MCS, this method has been proven to be effective and accurate in PPF calculation.

[0192] In one embodiment, the ASUT performance is evaluated based on existing technology and the relative error index of relative entropy and exceeding the limit probability, specifically as follows:

[0193] The formula for calculating the relative entropy is:

[0194]

[0195] Among them, the smaller the relative entropy and the relative error of the probability of exceeding the limit, the better the algorithm performance; Table 5 shows the RE and E-OLP values ​​of the AC bus 38 voltage amplitude, with a maximum limit of 1.06 pu and a minimum limit of 0.94 pu; Table 6 shows the RE value of the DC bus 7 voltage amplitude, with a maximum limit of 1.1 pu and a minimum limit of 0.9 pu; Table 7 shows the calculation time of different methods; Table 8 shows the parameter settings of ASUT.

[0196] Table 5

[0197]

[0198] Table 6

[0199]

[0200] As can be seen from Tables 5 and 6, ASUT has better performance compared to other existing technologies.

[0201] Table 7

[0202]

[0203] As shown in Table 7, ASUT can achieve higher computational accuracy while ensuring computational timeliness.

[0204] Based on the above embodiments, it can be seen that the AC / DC hybrid power grid risk assessment method provided by the present invention has the following beneficial effects: it effectively improves the efficiency of probabilistic power flow calculation while ensuring calculation accuracy, thereby improving the accuracy and timeliness of power system risk assessment.

[0205] Example 3

[0206] See Figure 5 , Figure 5 This is a schematic diagram of one embodiment of a risk assessment device for AC / DC hybrid power grids provided by the present invention, as shown below. Figure 5 As shown, the device includes: a first optimization module 501, a sample point sampling module 502, a fractional moment calculation module 503, a probability density function calculation module 504, and a probability power flow over-limit alarm module 505. The fractional moment calculation module 503 further includes a power flow calculation unit 5031 and a fractional moment calculation unit 5032. The probability density function calculation module 504 further includes a constrained optimization unit 5041, an unconstrained optimization unit 5042, and a probability density reconstruction unit 5043. Specifically:

[0207] The first optimization module is used to acquire historical data of random input variables of the power system, and establish a first optimization model based on the historical data, with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective.

[0208] The sample point sampling module is used to solve the first optimization model, obtain the optimal fractional moment ratio parameter and weight adjustment parameter, and obtain a preset number of sample points and their weight parameters of the random input variable based on the optimal fractional moment ratio parameter and weight adjustment parameter.

[0209] The fractional moment calculation module is used to obtain the fractional moment of the random output variable of the sample point based on the definition of fractional moment and according to the sample points and weight parameters of the random input variable;

[0210] The probability density function calculation module is used to calculate the probability density result when the information entropy of the random output variable is maximized, with the fractional moments of the random output variable of the sample point as constraints, and to reconstruct the probability density result.

[0211] The probability power flow over-limit alarm module is used to calculate the probability of the power system bus voltage and line power flow exceeding the limit based on the probability density result, and to confirm that the power system is in a dangerous state when the probability exceeds a preset value.

[0212] In one embodiment, the first optimization module is used to acquire historical data of random input variables of the power system, and establish a first optimization model based on the historical data, with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective. The optimization model includes: based on the definition of fractional moments, acquiring the fractional moments of the random output variables of the sample points according to the sample points and their weight parameters; using the fractional moments of the random output variables of the sample points as constraints, calculating the probability density result when the information entropy of the random output variables is maximized, and reconstructing the probability density result; calculating the probability of the power system bus voltage and line power flow exceeding limits based on the probability density result; and confirming that the power system is in a dangerous state when the probability exceeds a preset value.

[0213] Preferably, calculating the fractional moment of the random input variable based on its historical data includes:

[0214] The formula for calculating the fractional moments of the random input variables is:

[0215]

[0216] Where, n sam The number of historical data points for the random input variable is represented by α; α is the fractional order of the random input variable; x α Let E[·] be the Taylor expansion of the random input variable near its mean; E[·] is the expectation operator; For the i-th random input variable sam Historical data.

[0217] Preferably, the step of calculating the estimated value of the fractional moments of the random input variable based on the sample points and weight parameters of the random input variable includes:

[0218] The formula for calculating the fractional moment estimate of the random input variable is as follows:

[0219]

[0220] Where, ε (fm) and β (fm) These are the optimal scaling parameters and the optimal weight adjustment parameters for the fractional moments, respectively. The fractional moment estimate of the random input variable; For the weight parameters of the j-th sample point of the random input variable; Let j be the j-th sample point of the random input variable.

[0221] Preferably, the step of establishing a first optimization model with the difference between the fractional moments of the random input variable and its estimated value as the optimization objective includes:

[0222] The first optimization model is as follows:

[0223]

[0224] Where, ε (fm) and β (fm) These are the optimal scaling parameters and the optimal weight adjustment parameters for the fractional moments, respectively. The fractional moment estimate of the random input variable; For the weight parameters of the j-th sample point of the random input variable; Let j be the j-th sample point of the random input variable.

[0225] Preferably, the sample point sampling module is used to obtain a preset number of sample points and their weight parameters of the random input variable according to the optimal fractional moment ratio parameter and weight adjustment parameter, including:

[0226] The formula for calculating the sample points of the random input variable is:

[0227]

[0228] The formula for calculating the weight parameters of the sample points of the random input variable is as follows:

[0229]

[0230] Where, ε (fm) and β (fm) These are the optimal scaling parameters and the optimal weight adjustment parameters for the fractional moments, respectively. The fractional moment estimate of the random input variable; For the weight parameters of the j-th sample point of the random input variable; Let j be the j-th sample point of the random input variable.

[0231] Preferably, the probabilistic power flow overload alarm module is used to calculate the probability of the power system bus voltage and line power flow exceeding limits based on the probability density results, including:

[0232] The formula for calculating the probability of the random output variable exceeding the limit is:

[0233]

[0234] Where, p pro (·) represents the probability of exceeding the limits of the power system bus voltage and line power flow; f Y (Y _out ) is the probability density function of the random output variable; This represents the allowed range of random output variables.

[0235] In one embodiment, the fractional moment calculation module includes: a power flow calculation unit and a fractional moment calculation unit; wherein, the power flow calculation unit is used to sequentially input the sample points and their weight parameters into a preset deterministic power flow calculation model of a DC-DC hybrid system to obtain the random output variable results of the sample points; the fractional moment calculation unit is used to calculate the fractional moments of the random output variables based on the random output variable results of all the sample points and the weight parameters.

[0236] In one embodiment, the probability density function calculation module includes: a constrained optimization unit, an unconstrained optimization unit, and a probability density reconstruction unit. The constrained optimization unit is used to establish an information entropy maximization model with the fractional moments of the random output variable as constraints, and calculate the probability density result of the random output variable. The unconstrained optimization unit is used to transform the information entropy maximization model into an unconstrained optimization model, solve the unconstrained optimization model, and obtain the Lagrange multiplier vector and fractional vector of the random output variable. The probability density reconstruction unit is used to reconstruct the probability density result of the random output variable based on the Lagrange multiplier vector and fractional vector of the random output variable.

[0237] Example 4

[0238] Based on the above-described embodiments of the AC / DC hybrid power grid risk assessment method, another embodiment of the present invention provides an AC / DC hybrid power grid risk assessment device. The AC / DC hybrid power grid risk assessment terminal device includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the AC / DC hybrid power grid risk assessment method of any embodiment of the present invention.

[0239] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the AC / DC hybrid power grid risk assessment device.

[0240] The AC / DC hybrid power grid risk assessment equipment can be a desktop computer, laptop, handheld computer, or cloud server, etc. The AC / DC hybrid power grid risk assessment terminal equipment may include, but is not limited to, processors and memory.

[0241] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the AC / DC hybrid power grid risk assessment equipment, connecting various parts of the equipment via various interfaces and lines. The memory can be used to store the computer programs and / or modules. The processor implements various functions of the AC / DC hybrid power grid risk assessment equipment by running or executing the computer programs and / or modules stored in the memory, and by calling data stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, at least one application program required for a function, etc.; the data storage area can store data created based on the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0242] Example 5

[0243] Based on the above embodiments of the AC / DC hybrid power grid risk assessment method, another embodiment of the present invention provides a storage medium, the storage medium including a stored computer program, wherein, when the computer program is running, the device where the storage medium is located controls the execution of the AC / DC hybrid power grid risk assessment method of any embodiment of the present invention.

[0244] In this embodiment, the storage medium is a computer-readable storage medium, and the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium can include any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0245] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A risk assessment method for AC / DC hybrid power grids, characterized in that, include: Historical data of random input variables of the power system are obtained. Based on the historical data, a first optimization model is established with the optimization objective of minimizing the difference between the fractional moments of the random input variables and their estimated values. The random input variables include: the active and reactive power of the load and the active and reactive power output of renewable energy. Solve the first optimization model to obtain the optimal fractional moment ratio parameter and weight adjustment parameter, and obtain a preset number of sample points of the random input variable and their weight parameters based on the optimal fractional moment ratio parameter and weight adjustment parameter. Based on the definition of fractional moments, the fractional moments of the random output variables of the sample points are obtained according to the sample points and weight parameters of the random input variables; wherein, the random output variables include: node voltages of AC and DC systems, apparent power of AC branches, active power of DC branches, and reactive power provided by VSC. Using the fractional moments of the random output variables of the sample points as constraints, calculate the probability density result when the information entropy of the random output variables is maximized, and reconstruct the probability density result; Based on the probability density results, the probability of the power system bus voltage and line power flow exceeding the limits is calculated. When the probability exceeds a preset value, the power system is confirmed to be in a dangerous state.

2. The risk assessment method for AC / DC hybrid power grids as described in claim 1, characterized in that, The process involves acquiring historical data of random input variables of the power system, and based on this historical data, establishing a first optimization model with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective. This model includes: Based on the historical data of the random input variable, construct a formula for calculating the fractional moments of the random input variable; Based on the sample points and weight parameters of the random input variable, construct a formula for calculating the estimated value of the fractional moments of the random input variable; Based on the calculation formulas corresponding to the fractional moments and their estimated values ​​of the random input variables, a first optimization model is constructed with the optimization objective of minimizing the difference between the fractional moments and their estimated values ​​of the random input variables.

3. The risk assessment method for AC / DC hybrid power grids as described in claim 2, characterized in that, The step of constructing the formula for calculating the fractional moments of the random input variable based on the historical data of the random input variable includes: The formula for calculating the fractional moments of the random input variables is: in, This indicates the number of historical data points for the random input variable; The fractional order of the random input variable; The Taylor expansion of the random input variable around its mean; For expectation operators; For the first random input variable Historical data; For the first random input variable Historical data Power of 1.

4. The risk assessment method for AC / DC hybrid power grids as described in claim 2, characterized in that, The calculation formula for constructing the estimate of the fractional moment of the random input variable based on the sample points and their weight parameters includes: The formula for calculating the fractional moment estimate of the random input variable is as follows: in, and These are the proportional parameter and the weight adjustment parameter of the fractional moment, respectively; The fractional moment estimate of the random input variable; For the first random input variable Weight parameters for each sample point; For the first random input variable 100 sample points; This represents the number of sample points.

5. The method for risk assessment of AC / DC hybrid power grids as described in claim 2, characterized in that, The step of constructing a first optimization model based on the calculation formulas corresponding to the fractional moments and estimated values ​​of the random input variables, with the optimization objective being to minimize the difference between the fractional moments and estimated values ​​of the random input variables, includes: The first optimization model is as follows: in, For the random input variable The fractional moment estimate; and These are the proportional parameter and the weight adjustment parameter of the fractional moment, respectively; Representative with and The objective function for the decision variables; Represents the number of sample points; Representing the One sample point; To represent the order as random input variables fractional moments, for The upper limit of the value, for The interval between values.

6. The method for risk assessment of AC / DC hybrid power grids as described in claim 1, characterized in that, The step of obtaining a preset number of sample points and their weight parameters of the random input variables based on the optimal fractional moment ratio parameter and weight adjustment parameter includes: The formula for calculating the sample points of the random input variable is: The formula for calculating the weight parameters of the sample points of the random input variable is as follows: in, and The first and second random input variables are respectively The sample points and the first The weight parameters of each sample point; The mean of the sample points representing the random input variable; For the random input variable covariance; These are the initial weight parameters; The dimension representing the random input variable; and These are the optimal fractional moment ratio parameter and the weight adjustment parameter, respectively; It is the square of the optimal fractional moment proportionality parameter.

7. The risk assessment method for AC / DC hybrid power grids as described in claim 1, characterized in that, The step of calculating the probability of exceeding the limits of the power system bus voltage and line power flow based on the probability density results includes: The formula for calculating the probability of the random output variable exceeding the limit is: in, This represents the probability of exceeding the limits of the power system bus voltage and line power flow; Let be the probability density function of the random output variable; This represents the allowed range of random output variables; This is a random output variable.

8. The method for risk assessment of AC / DC hybrid power grids as described in claim 1, characterized in that, The step of obtaining the fractional moments of the random output variables based on the definition of fractional moments, according to the sample points and weight parameters of the random input variables, includes: The sample points and their weight parameters are sequentially input into the preset deterministic power flow calculation model of the DC-DC hybrid system to obtain the random output variable results of the sample points; Calculate the fractional moments of the random output variables based on the random output variable results and weight parameters for all the sample points.

9. The risk assessment method for AC / DC hybrid power grids as described in claim 1, characterized in that, The step of calculating the probability density result when the information entropy of the random output variable is maximized, constrained by the fractional moments of the random output variable of the sample points, and reconstructing the probability density result, includes: Using the fractional moments of the random output variables as constraints, an information entropy maximization model is established to calculate the probability density results of the random output variables; The information entropy maximization model is transformed into an unconstrained term optimization model. The unconstrained term optimization model is solved to obtain the Lagrange multiplier vector and fractional vector of the random output variable. Based on the Lagrange multiplier vector and fractional vector of the random output variable, reconstruct the probability density result of the random output variable.

10. A risk assessment device for AC / DC hybrid power grids, characterized in that, The AC / DC hybrid power grid risk assessment device includes: a first optimization module, a sample point sampling module, a fractional moment calculation module, a probability density function calculation module, and a probability power flow over-limit alarm module. The first optimization module is used to acquire historical data of random input variables of the power system, and establish a first optimization model based on the historical data, with the optimization objective being to minimize the difference between the fractional moments of the random input variables and their estimated values; wherein the random input variables include: the active and reactive power of the load and the active and reactive power output of renewable energy. The sample point sampling module is used to solve the first optimization model, obtain the optimal fractional moment ratio parameter and weight adjustment parameter, and obtain a preset number of sample points and their weight parameters of the random input variable based on the optimal fractional moment ratio parameter and weight adjustment parameter. The fractional moment calculation module is used to obtain the fractional moments of the random output variables of the sample points based on the definition of fractional moments and according to the sample points and weight parameters of the random input variables; wherein, the random output variables include: node voltages of AC and DC systems, apparent power of AC branches, active power of DC branches, and reactive power provided by VSC. The probability density function calculation module is used to calculate the probability density result when the information entropy of the random output variable is maximized, with the fractional moments of the random output variable of the sample point as constraints, and to reconstruct the probability density result. The probability power flow over-limit alarm module is used to calculate the probability of the power system bus voltage and line power flow exceeding the limit based on the probability density result, and to confirm that the power system is in a dangerous state when the probability exceeds a preset value.

11. The AC / DC hybrid power grid risk assessment device as described in claim 10, characterized in that, The first optimization module is used to acquire historical data of random input variables of the power system, and based on the historical data, establish a first optimization model with the difference between the fractional moments of the random input variables and their estimated values ​​as the optimization objective, including: Based on the historical data of the random input variable, construct a formula for calculating the fractional moments of the random input variable; Based on the sample points and weight parameters of the random input variable, construct a formula for calculating the estimated value of the fractional moments of the random input variable; Based on the calculation formulas corresponding to the fractional moments and their estimated values ​​of the random input variables, a first optimization model is constructed with the optimization objective of minimizing the difference between the fractional moments and their estimated values ​​of the random input variables.

12. The AC / DC hybrid power grid risk assessment device as described in claim 11, characterized in that, The step of constructing the formula for calculating the fractional moments of the random input variable based on the historical data of the random input variable includes: The formula for calculating the fractional moments of the random input variables is: in, This indicates the number of historical data points for the random input variable; The fractional order of the random input variable; The Taylor expansion of the random input variable around its mean; For expectation operators; For the first random input variable Historical data; For the first random input variable Historical data Power of 1.

13. The AC / DC hybrid power grid risk assessment device as described in claim 11, characterized in that, The calculation formula for constructing the estimate of the fractional moment of the random input variable based on the sample points and their weight parameters includes: The formula for calculating the fractional moment estimate of the random input variable is as follows: in, and These are the proportional parameter and the weight adjustment parameter of the fractional moment, respectively; The fractional moment estimate of the random input variable; For the first random input variable Weight parameters for each sample point; For the first random input variable 100 sample points; This represents the number of sample points.

14. The AC / DC hybrid power grid risk assessment device as described in claim 11, characterized in that, The step of constructing a first optimization model based on the calculation formulas corresponding to the fractional moments and estimated values ​​of the random input variables, with the optimization objective being to minimize the difference between the fractional moments and estimated values ​​of the random input variables, includes: The first optimization model is as follows: in, and These are the proportional parameter and the weight adjustment parameter of the fractional moment, respectively; Representative with and The objective function for the decision variables; Represents the number of sample points; Representing the One sample point; For the random input variable The fractional moment estimate; To represent the order as random input variables fractional moments, for The upper limit of the value, for The interval between values.

15. The AC / DC hybrid power grid risk assessment device as described in claim 10, characterized in that, The sample point sampling module is used to obtain a preset number of sample points and their weight parameters of the random input variable according to the optimal fractional moment ratio parameter and weight adjustment parameter, including: The formula for calculating the sample points of the random input variable is: The formula for calculating the weight parameters of the sample points of the random input variable is as follows: in, and The first and second random input variables are respectively The sample points and the first The weight parameters of each sample point; The mean of the sample points representing the random input variable; For the random input variable covariance; These are the initial weight parameters; The dimension representing the random input variable; and These are the optimal fractional moment ratio parameter and the weight adjustment parameter, respectively; It is the square of the optimal fractional moment proportionality parameter.

16. The AC / DC hybrid power grid risk assessment device as described in claim 10, characterized in that, The probabilistic power flow overload alarm module is used to calculate the probability of the power system bus voltage and line power flow exceeding limits based on the probability density results, including: The formula for calculating the probability of the random output variable exceeding the limit is: in, This represents the probability of exceeding the limits of the power system bus voltage and line power flow; Let be the probability density function of the random output variable; This represents the allowed range of random output variables; This is a random output variable.

17. The AC / DC hybrid power grid risk assessment device as described in claim 10, characterized in that, The fractional moment calculation module includes: a power flow calculation unit and a fractional moment calculation unit. The power flow calculation unit is used to sequentially input the sample points and their weight parameters into the preset deterministic power flow calculation model of the DC-DC hybrid system, and obtain the random output variable results of the sample points. The fractional moment calculation unit is used to calculate the fractional moment of the random output variable based on the random output variable results of all the sample points and the weight parameters.

18. The AC / DC hybrid power grid risk assessment device as described in claim 10, characterized in that... The probability density function calculation module includes: a constrained optimization unit, an unconstrained optimization unit, and a probability density reconstruction unit. The constrained optimization unit is used to establish an information entropy maximization model with the fractional moments of the random output variable as constraints, and to calculate the probability density result of the random output variable. The unconstrained term optimization unit is used to transform the information entropy maximization model into an unconstrained term optimization model, solve the unconstrained term optimization model, and obtain the Lagrange multiplier vector and fractional vector of the random output variable. The probability density reconstruction unit is used to reconstruct the probability density result of the random output variable based on the Lagrange multiplier vector and the fractional vector of the random output variable.

19. A terminal device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements a risk assessment method for an AC / DC hybrid power grid as described in any one of claims 1 to 9.

20. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform a risk assessment method for AC / DC hybrid power grids as described in any one of claims 1 to 9.

Images