A model data hybrid driving power system small disturbance stability evaluation method

Abstract

The application discloses a model data hybrid driving power system small disturbance stability evaluation method, obtains the target topology of the power system, carries out the deterministic power flow analysis and the modal analysis according to the target topology, and obtains the static working point under the deterministic scene. The probability distribution modeling of the new energy uncertainty input of the power system is combined with the converter characteristic constraint, the impedance model of the power system is established according to the new energy dynamic characteristic and the static working point under the deterministic scene. Further, the expression of the real part of the dominant characteristic mode of the system is solved, and the influence of the uncertainty on the small disturbance stability of the new energy power system is analyzed based on this. In this way, the data-driven polynomial base is determined for the established model, the small disturbance stability of the system is evaluated in the mode of model-data hybrid driving, and the small disturbance stability of the power system can be determined at one time according to the random characteristics of the renewable energy.

Description

Technical Field

[0001] This invention relates to the field of small-disturbance stability assessment technology for power systems, and in particular to a model-data hybrid-driven method for assessing small-disturbance stability of power systems. Background Technology

[0002] Due to the influence of natural factors such as sunlight and wind speed, renewable energy power generation exhibits significant random fluctuations, which inevitably affect the stable operation of the power system after grid connection. Traditional deterministic methods can only assess system stability under a specific operating state, thus often considering system stability under the worst-case scenario and making scheduling decisions based on this. This approach fails to comprehensively consider system uncertainties for stability assessment, potentially leading to overly conservative evaluation results.

[0003] Probabilistic statistical methods establish an explicit mapping between the uncertainty of renewable energy output and system stability, thus reflecting the true system response and behavior more accurately. Monte Carlo simulation, as a numerical method of probabilistic analysis, primarily generates numerous computational scenarios based on the distribution density of uncertainty sources (such as irradiance and wind speed). It then performs deterministic small-disturbance stability calculations on each scenario, accumulates the results, and ultimately forms the probability distribution density of the system's key eigenvalues, thereby determining the probabilistic stability of the power system. However, for probabilistic stability studies of large-scale power systems, Monte Carlo simulation is an extremely time-consuming method. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a model-data hybrid-driven method for evaluating the small-disturbance stability of a power system, which can determine the small-disturbance stability of the power system in one go, taking into account the stochastic characteristics of renewable energy.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] A model-data hybrid-driven method for evaluating the small-disturbance stability of power systems includes the following steps:

[0007] Obtain the target topology of the power system, and perform deterministic power flow analysis and modal analysis based on the target topology to obtain the static operating point under the deterministic scenario;

[0008] By combining converter characteristic constraints, a probability distribution model is performed on the uncertain input of new energy sources in the power system, and an impedance model of the power system is established based on the dynamic characteristics of new energy sources and the static operating point under the deterministic scenario.

[0009] Based on the impedance model, a data-driven polynomial basis for new energy uncertainty is constructed using the generalized polynomial chaotic expansion method.

[0010] The real part expression of the dominant eigenvalue of the power system is obtained by data-driven polynomial basis fitting.

[0011] Curve fitting is performed based on the real part expression of the dominant eigenvalue, and the small-disturbance probability stability of the power system under the influence of new energy uncertainties is evaluated based on the fitting results.

[0012] The beneficial effects of this invention are as follows: By acquiring the target topology of the power system, deterministic power flow analysis and modal analysis are performed based on the target topology to obtain the static operating point under deterministic scenarios. Probability distribution modeling is performed on the uncertain input of new energy sources to the power system in conjunction with converter characteristic constraints, and an impedance model of the power system is established based on the dynamic characteristics of new energy sources and the static operating point under deterministic scenarios. Then, the expression of the real part of the dominant characteristic mode of the system is obtained, and the impact of uncertainty on the small-disturbance stability of the new energy power system is analyzed based on this. In this way, a data-driven polynomial basis is determined for the established model, and the small-disturbance stability of the system is evaluated through a model-data hybrid driving approach, enabling the determination of the small-disturbance stability of the power system in one go, considering the stochastic characteristics of renewable energy. Attached Figure Description

[0013] Figure 1 This is a flowchart of a model-data hybrid-driven method for evaluating the stability of a power system under small disturbances, according to an embodiment of the present invention. Detailed Implementation

[0014] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0015] Please refer to Figure 1 This invention provides a model-data hybrid-driven method for evaluating the small-disturbance stability of power systems, comprising the following steps:

[0016] Obtain the target topology of the power system, and perform deterministic power flow analysis and modal analysis based on the target topology to obtain the static operating point under the deterministic scenario;

[0017] By combining converter characteristic constraints, a probability distribution model is performed on the uncertain input of new energy sources in the power system, and an impedance model of the power system is established based on the dynamic characteristics of new energy sources and the static operating point under the deterministic scenario.

[0018] Based on the impedance model, a data-driven polynomial basis for new energy uncertainty is constructed using the generalized polynomial chaotic expansion method.

[0019] The real part expression of the dominant eigenvalue of the power system is obtained by data-driven polynomial basis fitting.

[0020] Curve fitting is performed based on the real part expression of the dominant eigenvalue, and the small-disturbance probability stability of the power system under the influence of new energy uncertainties is evaluated based on the fitting results.

[0021] As described above, the beneficial effects of this invention are as follows: By acquiring the target topology of the power system, deterministic power flow analysis and modal analysis are performed based on the target topology to obtain the static operating point under deterministic scenarios. Probability distribution modeling is performed on the uncertain input of new energy sources to the power system in conjunction with converter characteristic constraints, and an impedance model of the power system is established based on the dynamic characteristics of new energy sources and the static operating point under deterministic scenarios. Furthermore, the expression for the real part of the dominant characteristic mode of the system is obtained, and the impact of uncertainty on the small-disturbance stability of the new energy power system is analyzed based on this. In this way, a data-driven polynomial basis is determined for the established model, and the small-disturbance stability of the system is evaluated through a model-data hybrid driving approach, enabling the determination of the small-disturbance stability of the power system in one go, considering the stochastic characteristics of renewable energy.

[0022] Furthermore, the probability distribution modeling of the uncertain input of new energy sources to the power system, combined with converter characteristic constraints, includes:

[0023] The solar irradiance input is modeled based on the Gaussian distribution to obtain the probability density distribution function of the solar irradiance; the wind speed input is modeled based on the Weibull distribution to obtain the Weibull distribution function of the wind speed.

[0024] Based on the probability density distribution function of solar irradiance and the Weibull distribution function of wind speed, combined with the converter characteristic constraints in the power system, a probabilistic model of wind power output and photovoltaic power output is performed to obtain an uncertain power output model for new energy.

[0025] As described above, by modeling the solar irradiance input and wind speed input, we can obtain a new energy uncertain output model, which facilitates subsequent evaluation of the small-disturbance stability of the new energy power system.

[0026] Furthermore, establishing the impedance model of the power system based on the dynamic characteristics of new energy sources and the static operating point under the deterministic scenario includes:

[0027] Based on the dynamic characteristics of new energy, dynamic models of the filter inductor, phase-locked loop and current inner loop in the motor magnetic field coordinate system are established respectively, and a dynamic model of the DC voltage loop filter capacitor is established.

[0028] The impedance transfer function of the new energy source is obtained by combining all the established dynamic models and the static operating point of the power system under deterministic scenarios.

[0029] The impedance model of the power system is obtained by transforming the coordinate system of the impedance transfer function into a Cartesian coordinate system.

[0030] As described above, an impedance model is established by combining all the established dynamic models and the static operating point of the power system in a deterministic scenario, so as to facilitate the subsequent calculation of the probability distribution of the real part of the dominant characteristic mode of the system.

[0031] Furthermore, the impedance model of the power system is obtained, and then the process further includes:

[0032] Establish a dynamic model of the AC grid side of the power system and remove the passive nodes of the AC grid side dynamic model;

[0033] The total admittance matrix of the power system is calculated based on the impedance model of the power system and the dynamic model of the AC grid side after removing the passive nodes.

[0034] As described above, the total admittance matrix is ​​calculated based on the impedance model and the AC grid-side dynamic model with passive nodes removed, in order to study the impact of new energy uncertainty on the small-disturbance stability of the power system.

[0035] Furthermore, the construction of a data-driven polynomial basis for new energy uncertainty based on the impedance model using the generalized polynomial chaotic expansion method includes:

[0036] An initial polynomial basis is established based on the linearly independent monomials of the new energy random variables;

[0037] Based on the initial polynomial basis, an orthogonal polynomial basis in the probability space corresponding to the impedance model is constructed using the orthogonal method.

[0038] Construct a data-driven polynomial basis for new energy uncertainty based on the tensor product of the orthogonal polynomial basis.

[0039] As can be seen from the above description, since it is not necessary to know the specific probability distribution information when constructing an orthogonal polynomial basis for the joint probability density function of random variables, the polynomial basis constructed by the above process is based on model-data hybrid driving.

[0040] Furthermore, based on the data-driven polynomial basis fitting of the power system under the influence of new energy uncertainties, the expression for the real part of the dominant eigenvalue of the system includes:

[0041] The random variables for the new energy source include solar irradiance and wind speed;

[0042] The moments of solar irradiance and wind speed are calculated based on the probability distribution model, and the orthogonal polynomial basis of the statistical moments of solar irradiance and wind speed is calculated. All polynomial bases of less than or equal to the preset order are constructed in combination with the probability distribution model.

[0043] Using moments of solar irradiance and wind speed, we perform order reduction and matrix transformation on all polynomial bases of orders less than or equal to a preset order to obtain the real part expression of the dominant characteristic root of the system under the influence of new energy uncertainties.

[0044] Furthermore, curve fitting is performed based on the real part expression of the dominant eigenvalue to evaluate the small-disturbance probability stability of the power system under the influence of new energy uncertainties, including:

[0045] Based on the orthogonal polynomial basis of statistical moments of solar irradiance and wind speed, the expansion of the real part of the dominant eigenvalues ​​of the system is obtained by fitting the real part expression of the dominant eigenvalues ​​of the system.

[0046] The system instability probability is calculated based on the expansion of the real part of the dominant eigenvalue of the system, and the small-disturbance probabilistic stability of the power system under the influence of new energy uncertainties is evaluated based on the probability calculation results.

[0047] As described above, by expanding the theory to obtain the probability distribution of the dominant characteristic roots of the system under the uncertainty of photovoltaic and wind power in a high-proportion new energy system, the stability state of the system under uncertainty can be accurately evaluated.

[0048] The model-data hybrid-driven power system small-disturbance stability assessment method described above is applicable to determining the small-disturbance stability of a power system in a single step, taking into account the stochastic characteristics of renewable energy. The following detailed implementation method illustrates this method:

[0049] Example 1

[0050] Please refer to Figure 1 A model-data hybrid-driven method for evaluating the small-disturbance stability of power systems includes the following steps:

[0051] S1. Obtain the target topology of the power system, and perform deterministic power flow analysis and modal analysis based on the target topology to obtain the static operating point under the deterministic scenario.

[0052] Specifically, the modal analysis of the operating point of a high-proportion renewable energy power system mainly includes the determination of the network topology and parameters of the high-proportion renewable energy power system. The determined parameters include the nodes, lines and initial voltage of the power system.

[0053] Among them, deterministic power flow calculation and deterministic modal analysis are obtained, including determining steady-state voltage and current, system oscillation mode distribution and oscillation mode damping ratio.

[0054] S2. Combine converter characteristic constraints to perform probability distribution modeling of uncertain inputs of new energy sources in the power system, and establish the impedance model of the power system based on the dynamic characteristics of new energy sources and the static operating point under the deterministic scenario.

[0055] S21. Establish a new energy uncertainty output model.

[0056] S211. Model the solar irradiance input based on the Gaussian distribution to obtain the probability density distribution function of the solar irradiance:

[0057]

[0058] h = [h1, h2, ... h i h n ]=σ pv Z+μ pv

[0059]

[0060]

[0061] In the formula, h represents the Gaussian random variable set of solar irradiance, h i Let σ represent the i-th sample in the set of random variables. pv Let μ represent the standard deviation of the random variable set h, Z represent the random variable set that follows a standard normal distribution, and μ represent the standard deviation of the random variable set h. pv f represents the mean of the set of random variables h, n represents the number of samples, and f gau (h) represents the probability density distribution function PDF of solar irradiance.

[0062] S212. Based on the Weibull distribution, model the wind speed input to obtain the Weibull distribution function of the wind speed:

[0063]

[0064] v = [v1, v2, ... v] i v n ]=σ wind Z+μ wind

[0065] μ wind =E[v]=cΓ(1+1 / m)

[0066]

[0067] In the formula, v represents a set of Weibull random variables for wind speed, v i Let σ represent the i-th sample of the random variable. wind μ represents the standard deviation of the random variable v. wind Let f represent the mean of the random variable v, Z represent the random variable following a standard normal distribution, n represent the sample size, m represent the shape parameter, c represent the characteristic time, Γ is the gamma function, and f wei(v) represents a set of random variables with a normalized Weibull distribution.

[0068] S213. Based on the probability density distribution function of solar irradiance and the Weibull distribution function of wind speed, combined with the converter characteristic constraints in the power system, a probabilistic model of wind power output and photovoltaic power output is performed to obtain the uncertain power output model of new energy:

[0069]

[0070]

[0071] In the formula, P pv A represents photovoltaic power generation. p The total area of ​​the solar panels is represented by r, and the production volume of the solar panels is represented by h. i Let PR represent the i-th sample in the set of random variables of solar irradiance, and H represent the performance ratio (loss coefficient). rated P represents the maximum irradiance that generates the rated power of a photovoltaic system. pv,rated Indicates the rated power of photovoltaic power;

[0072] P w P represents wind power generation. w,j Let A represent the wind power generation of the j-th wind turbine, ρ represent the air density, and A represent the wind power output of the j-th wind turbine. w This represents the swept area of ​​the turbine (m²). 2 C represents the power efficiency, which depends on the tip speed ratio and the pitch angle control. j,i P represents the i-th sample of the wind speed of the j-th wind turbine WTG. w,rated This indicates the rated power (WTG) of the wind turbine generator, in V. cut-in V cut-out and V rated These represent the access wind speed, cut-off wind speed, and rated wind speed, respectively.

[0073] S22. Establish an impedance model for the power system based on the dynamic characteristics of new energy sources.

[0074] S221. Construct a dynamic model of a photovoltaic and wind power grid-connected system considering only the grid-side converter. Establish a dynamic model of the filter inductor under the dq axis of the motor magnetic field coordinate system:

[0075]

[0076] In the formula, U sd U sq U represents the d-axis and q-axis components of the output voltage of the grid-connected current transformer for a direct-drive wind turbine, respectively. d U q L represents the d-axis and q-axis components of the device node port voltage, respectively. fIndicates the filter inductance, s represents the Laplace operator, and I represents the filter inductance. d I q These represent the d-axis and q-axis components of the grid-side converter output current, respectively.

[0077] S222. Establish the dynamic model of the phase-locked loop under the dq axis:

[0078]

[0079] In the formula, θ pll The converter angle represented by the phase-locked loop output; k ppll k ipll ω and ω0 represent the proportional constant and integral constant of the phase-locked loop (PLL) PI element, respectively; ω0 represents the synchronous rotation speed.

[0080] S223. Establish the dynamic model of the inner current loop under the dq axis:

[0081]

[0082] In the formula, U sdref U sqref These represent the reference values ​​for the d-axis and q-axis components of the output voltage of the grid-connected current transformer for direct-drive fans, respectively; I dref I represents the d-axis reference value of the current flowing through the filter inductor. qref This represents the q-axis reference value of the filter inductor current, I in the DC voltage outer loop control. qref Take 0; k pc k ic T represents the proportional constant and integral constant of the PI element in the inner current loop, respectively; FF This represents the time constant of the voltage feedforward low-pass filter.

[0083] S224. Establish a dynamic model of the DC voltage loop filter capacitor:

[0084]

[0085] In the formula, U dref U dc These represent the reference and actual values ​​of the DC capacitor voltage, respectively; C dc Indicates DC capacitance; P, P e These represent the input and output power across the DC capacitor, respectively; k pdc k idc These represent the proportional constant and integral constant of the PI link in the outer loop of the DC voltage, respectively.

[0086] S225. By combining the four dynamic models from steps S221 to S224, we can obtain the second-order impedance transfer function of the new energy source in the dq local coordinate system:

[0087]

[0088] In the formula, G dd (s), G qq (s) are matrix elements, which contain DC capacitors and grid-side converter control parameters, characterizing the internal dynamics of the wind turbine. d0 U dc0 U d0 These represent the static operating point values ​​corresponding to the d-axis components of the converter output current, the DC-side capacitor voltage, and the grid-connected voltage, respectively.

[0089] S226. Transform the new energy impedance model to the xy coordinate system:

[0090]

[0091] S23. Calculate the total admittance matrix of the new energy power system.

[0092] S231. Suppose that the AC network contains (n+m+1) nodes, and is divided into two parts according to whether it is connected to photovoltaic, wind turbines and external power grid. Then the dynamic model of the AC grid side can be expressed as Y. net (s) are as follows:

[0093]

[0094] In the formula, ΔI xy_net and ΔU xy_net The static operating point column vectors, which represent the x and y axis deviation components of the AC grid node input current and node voltage after linearization in the global xy coordinate system, can be calculated based on step 22. Let Ynet(s) represent the Kronecker product, and ω0 be the synchronous rotational speed. Ynet(s) is the AC network transfer function matrix in the frequency domain, and B is the network admittance matrix, divided into four parts according to whether it is connected to the wind turbine or an external power grid node. net11 ∈R (n+1)×(n+1) B net12 ∈R (n+1)×m B net21 ∈R m×(n+1) B net22 ∈R m×m This is the admittance submatrix.

[0095] S232. Based on Schur elimination of passive nodes in the dynamic model of the AC power grid side, we can obtain:

[0096]

[0097] In the formula, ΔI xy_red and ΔU xy_redIt only contains the static operating point column vectors corresponding to the linearized x and y axis input current deviations of the wind turbine nodes and external grid nodes.

[0098] S233. Obtain the total admittance matrix Y of the power system. total Y total (s)=-Y g (s)+Y sys (s).

[0099] S3. Based on the impedance model, a data-driven polynomial basis for new energy uncertainty is constructed using the generalized polynomial chaotic expansion method.

[0100] S31. Select linearly independent monomials l with respect to d-dimensional random variables to form an initial polynomial basis, and use the set e j (l) indicates:

[0101]

[0102] In the formula, e j (l) denotes the j-th order polynomial basis, where m is the maximum order of the polynomial basis. Let l represent the tensor product; l is any random variable, which is substituted into irradiance h and wind speed v respectively when applying uncertain basis for photovoltaic and wind power.

[0103] S32. Construct an orthogonal polynomial basis in the probability space using the Schmitt orthogonalization algorithm:

[0104]

[0105] In the formula, <·> represents the inner product function, and L is an orthogonal basis for any random variable.

[0106] S33. Constructing orthogonal bases in a multidimensional probability space based on the tensor product of orthogonal bases in independent univariate probability spaces:

[0107]

[0108] In the formula, Represents a 1-dimensional random variable h i Related k i Polynomial basis of order.

[0109] S34. Calculate the first 2m-1 moments of a random variable:

[0110]

[0111] In the formula, The moment of order j can be directly obtained from statistical data, and j∈{1,2,…,2m-1}. Since the construction of the orthogonal polynomial basis of the joint probability density function of random variables only utilizes the moment information of a certain order from the statistical data, and does not require knowledge of its specific probability distribution information, the polynomial basis constructed by the above process is data-driven rather than probability distribution model-driven.

[0112] S4. Obtain the real part expression of the dominant eigenvalue of the power system by performing polynomial basis fitting based on the data.

[0113] S41. Calculate the moments for the random variables irradiance and wind speed:

[0114]

[0115] In the formula, E(·) is the mean function, and α ν and β ν Let v = 1, ..., n represent the moments corresponding to the random variables of photovoltaic and wind speed, respectively, where v = 1, ..., n represents the order of the corresponding moments.

[0116] S42. Calculate the first (2m-1)th moment of the input illumination and wind power statistics based on step S34.

[0117] S43. Orthogonal polynomial basis for calculating the statistical moments of light intensity and wind speed:

[0118]

[0119] In the formula, H (m) (h) and V (m) (v) represents the m-th order polynomial basis for irradiance and wind speed, respectively; and These are the coefficients of the basis of the corresponding order polynomial.

[0120] S44. Construct a polynomial basis of order m and a polynomial basis of all orders less than m:

[0121]

[0122] In the formula, and denoted by and respectively, representing the coefficients of the m-th order polynomial basis for irradiance and wind speed.

[0123] S45. By reducing the order of the polynomial basis in step S44 using the orthogonal polynomial basis H and V, we obtain:

[0124]

[0125] In the formula, α i and β iLet be the i-th moments of the random variables photovoltaic and wind speed, respectively.

[0126] S46. Denote the reduced-order polynomial basis in matrix form:

[0127]

[0128] In the formula, we can obtain the following from the multivariate linear equation: and These are the coefficients of the basis of the corresponding order polynomial.

[0129] S5. Based on the real part expression of the dominant eigenvalue, perform curve fitting, and evaluate the small-disturbance probability stability of the power system under the influence of new energy uncertainties based on the fitting results.

[0130] S51. Establish the real part expansion of the dominant eigenvalue of the system considering uncertainties in wind speed and irradiance:

[0131]

[0132] In the formula, c0, c1, ..., c n-1 The coefficients are undetermined. The real part of the dominant eigenvalues ​​of the system is approximated as a polynomial function, which is a weighted sum of the polynomial basis.

[0133] S52. Locate probability points h0, h1, ..., h n-1 and v0, v1, ..., v n-1 Substituting into the polynomial basis H respectively (m) (h) and V (m) (v) in.

[0134] S53. Substitute the irradiance and wind speed corresponding to the probability collocation points into the small-signal model of the system, and calculate the real part of the dominant eigenvalue σ(x) of the system.

[0135] S54. Calculate c0, c1, ..., c n-1 :

[0136]

[0137] S55. Obtain the real part expansion of the dominant eigenvalues ​​of the system based on the generalized chaotic polynomial fitting by obtaining the coefficients.

[0138] S56. Calculate the system instability probability and determine the system's small-disturbance probability stability state under the uncertainties of wind power and photovoltaic power:

[0139]

[0140] In summary, this invention provides a model-data hybrid-driven method for assessing the small-disturbance stability of power systems. It obtains the target topology of the power system, performs deterministic power flow analysis and modal analysis based on the target topology, and obtains the static operating point under deterministic scenarios. Combining converter characteristic constraints, it models the probability distribution of uncertain inputs from renewable energy sources in the power system, and establishes an impedance model of the power system based on the dynamic characteristics of renewable energy sources and the static operating point under deterministic scenarios. Furthermore, it derives the expression for the real part of the dominant characteristic mode of the system, and analyzes the impact of uncertainty on the small-disturbance stability of the renewable energy power system. In this way, a data-driven polynomial basis is determined for the established model, and the small-disturbance stability of the system is assessed through a model-data hybrid approach. This method can determine the small-disturbance stability of the power system in one step, considering the stochastic characteristics of renewable energy.

[0141] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A model-data hybrid-driven method for evaluating the small-disturbance stability of power systems, characterized in that, Including the following steps: Obtain the target topology of the power system, and perform deterministic power flow analysis and modal analysis based on the target topology to obtain the static operating point under the deterministic scenario; By combining converter characteristic constraints, a probability distribution model is performed on the uncertain input of new energy sources in the power system: the solar irradiance input is modeled based on the Gaussian distribution to obtain the probability density distribution function of the solar irradiance; the wind speed input is modeled based on the Weibull distribution to obtain the Weibull distribution function of the wind speed; based on the probability density distribution function of the solar irradiance and the Weibull distribution function of the wind speed, combined with the converter characteristic constraints in the power system, a probability model is performed on the wind power output and photovoltaic power output to obtain the uncertain output model of new energy sources. Based on the dynamic characteristics of new energy sources and the static operating point under the deterministic scenario, an impedance model of the power system is established: dynamic models of the filter inductor, phase-locked loop, and inner current loop in the motor magnetic field coordinate system are established according to the dynamic characteristics of new energy sources, and a dynamic model of the DC voltage loop filter capacitor is established; combining all the established dynamic models and the static operating point of the power system under the deterministic scenario, the impedance transfer function of the new energy sources is obtained; the coordinate system of the impedance transfer function is transformed into a plane rectangular coordinate system to obtain the impedance model of the power system. Establish a dynamic model of the AC grid side of the power system and remove the passive nodes of the AC grid side dynamic model; The total admittance matrix of the power system is calculated based on the impedance model of the power system and the dynamic model of the AC grid side after removing the passive nodes. Based on the impedance model, a data-driven polynomial basis for new energy uncertainty is constructed using the generalized polynomial chaotic expansion method: an initial polynomial basis is established based on the linearly independent monomials of the new energy random variables; based on the initial polynomial basis, an orthogonal polynomial basis in the probability space corresponding to the impedance model is constructed using the orthogonal method; and a data-driven polynomial basis for new energy uncertainty is constructed based on the tensor product of the orthogonal polynomial basis. The real part expression of the dominant characteristic root of the power system is obtained by data-driven polynomial basis fitting: the new energy random variables include solar irradiance and wind speed; The moments of solar irradiance and wind speed are calculated based on the probability distribution model, and the orthogonal polynomial basis of the statistical moments of solar irradiance and wind speed is calculated. All polynomial bases of less than or equal to the preset order are constructed in combination with the probability distribution model. The moments of solar irradiance and wind speed are used to reduce the order and perform matrix transformation on all polynomial bases of less than or equal to the preset order to obtain the real part expression of the dominant characteristic root of the system under the influence of new energy uncertainty. Curve fitting is performed based on the real part expression of the dominant eigenvalue, and the small-disturbance probability stability of the power system under the influence of new energy uncertainties is evaluated based on the fitting results.

2. The method for evaluating the small-disturbance stability of a power system driven by hybrid model data according to claim 1, characterized in that, Curve fitting is performed based on the real part expression of the dominant eigenvalue to evaluate the small-disturbance probability stability of the power system under the influence of new energy uncertainties, including: Based on the orthogonal polynomial basis of statistical moments of solar irradiance and wind speed, the expansion of the real part of the dominant eigenvalues ​​of the system is obtained by fitting the real part expression of the dominant eigenvalues ​​of the system. The system instability probability is calculated based on the expansion of the real part of the dominant eigenvalue of the system, and the small-disturbance probabilistic stability of the power system under the influence of new energy uncertainties is evaluated based on the probability calculation results.

Images