Response-driven dominant oscillation modal energy characteristic influence factor analysis method

Abstract

The invention discloses a response-driven dominant oscillation mode energy characteristic influence factor analysis method, and relates to the technical field of electric power. The method comprises the following steps: extracting a dominant subsynchronous oscillation mode component of a grid-connected point current and a direct-driven wind power plant outlet voltage; fitting the dominant subsynchronous oscillation mode component according to the transient energy function of the grid-connected port of the direct-driven wind power plant to obtain energy flow power; according to a system operation principle, determining influence factors influencing the dominant subsynchronous oscillation energy characteristics; the influence factors serve as independent variables, the energy flow power serves as dependent variables, and the objective weight and the subjective weight of each influence factor on the energy flow power are calculated through a principal component regression method and an improved analytic hierarchy process; according to the objective weight and the subjective weight of each influence factor to the energy flow power, calculating the comprehensive weight of the influence factor to the energy flow power; and according to the comprehensive weight of each influence factor on the energy flow power, determining the dominant influence factor of the dominant oscillation mode energy characteristic.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power, and in particular to an analysis method for influencing factors of the energy characteristics of the dominant oscillation mode in response to driving. Background Art

[0002] In recent years, artificial intelligence algorithms have been widely applied to power systems, mainly focusing on aspects such as load and power prediction, stability assessment and decision-making, fault diagnosis, and early warning. Among them, the research on the oscillation characteristics of wind farms mainly focuses on the detection of oscillation components, the early warning of power oscillation, and the location of disturbance sources. Generally, the oscillation parameter identification is characterized as a regression problem, while the oscillation detection problem is characterized as a classification problem. Some other methods respectively use convolutional neural networks and deep residual networks to achieve the fast blind identification of the oscillation characteristics of measured signals. There are also some methods that combine the fast identification of qualitative features with the quantitative modal identification, and use the random forest algorithm to establish the relationship between the massive data obtained after dimensionality reduction and clustering processing and various oscillation modes, further improving the identification speed. There are also some methods that construct an oscillation evaluation and prediction model based on graph attention neural networks, but the data preprocessing and feature extraction methods used by the two are different, and the latter can only achieve the prediction of key oscillation modes; the former introduces a heterogeneous graph design when constructing the model, differentiates and classifies and aggregates node features and edge information, and designs a multi-task learning framework. While predicting the damping ratio according to the extracted features of effective clustering, the participation degree of the unit in the oscillation model is predicted. There are also some methods that use randomized dynamic mode decomposition to accurately obtain the mode parameters, and then use a multi-layer artificial neural network to obtain the mapping relationship between the response signal and the mode parameters. After sufficient offline training and learning, the real-time online monitoring of oscillation power can be achieved. Although the above-mentioned oscillation analysis methods relying on artificial intelligence algorithms can improve the rapidity and accuracy of oscillation identification, the research models are basically trained from oscillation data in limited scenarios. However, the actual system site conditions are complex, and new oscillation types and different dynamic characteristics may occur, resulting in problems such as insufficient or incorrect selection of model features. Therefore, the applicability of the monitoring and evaluation methods based on intelligent algorithms needs to be further explored.

[0003] To solve the problem that the above data-driven analysis overly relies on data samples, some methods have proposed an online identification method for broadband oscillation by fusing data and impedance models, which equivalently realizes the calculation process of the impedance matrix based on the physical model using a neural network, and then connects the data model and the physical model according to the circuit network topology to achieve the equivalent aggregation of the impedance network and the online evaluation of small-signal stability.

[0004] It is worth mentioning that there are few studies on analyzing the influencing factors of oscillations based on the response-driven method. Some methods use the clustering algorithm to analyze the influencing factors in the oscillation operation data in order to explore the influence of wind speed and the voltage / current fluctuation indexes at the grid connection point of the wind turbine on the oscillation mode, and obtain the cluster characteristics of uncertain factors. Some of these methods further cluster the oscillation modes obtained by the Prony algorithm with the wind speed / voltage fluctuations, establish an Apriori association analysis model, so as to obtain the association rules between the oscillation modes and the research parameters, and realize the real-time monitoring and predictive analysis of oscillations.

[0005] The above data-driven analysis methods have the problem of over-relying on data samples. Moreover, there are few studies on analyzing the influencing factors of oscillations based on the response-driven method. Currently, association rule analysis is commonly used, but its effect is affected by the setting of parameters such as its threshold and support degree, and the above methods are not suitable for analyzing the influence of multiple factors.

[0006] This section aims to provide the background or context for the embodiments of the present invention stated in the claims. The description herein is not admitted to be prior art just because it is included in this section. Summary of the Invention

[0007] To solve the above technical problems, an embodiment of the present invention provides a method for analyzing the influencing factors of the energy characteristics of the dominant oscillation mode driven by response, which can solve at least a part of the above technical problems.

[0008] A method for analyzing the influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by an embodiment of the present invention includes:

[0009] Collect the grid connection point current of the direct-drive wind farm grid connection system and the outlet voltage of the direct-drive wind farm;

[0010] Extract the dominant sub-synchronous oscillation mode components of the voltage and current;

[0011] According to the transient energy function of the grid connection port of the direct-drive wind farm, fit the dominant sub-synchronous oscillation mode components of the voltage and current obtained by extraction to obtain the energy flow power characterizing the energy characteristics of the corresponding oscillation mode components;

[0012] According to the operation principle of the direct-drive wind farm grid connection system, determine the influencing factors affecting the energy characteristics of the dominant sub-synchronous oscillation;

[0013] Taking the influencing factors as independent variables and the energy flow power as the dependent variable, use the principal component regression method to calculate the objective weight of each influencing factor on the energy flow power, and use the improved analytic hierarchy process to calculate the subjective weight of each influencing factor on the energy flow power;

[0014] Calculate the comprehensive weight of each of the influencing factors on the energy flow power according to the objective weight and the subjective weight of each of the influencing factors on the energy flow power;

[0015] Determine the main influencing factors of the dominant oscillation mode energy characteristics of the grid-connected system of the direct-drive wind farm according to the comprehensive weight of each of the influencing factors on the energy flow power.

[0016] In some embodiments, the energy flow power is calculated using the following formula:

[0017] W port,i = ∫Im[(I ij,d - jI ij,q )(dU i,d + jdU i,q )]= ∫(I ij,d dU i,q - I ij,q dU i,d );

[0018] In the formula, I ij,d , I ij,q represent the dominant subsynchronous oscillation mode components of the grid-connected branch L ij of the grid-connected port node i of the grid-connected system of the direct-drive wind farm in the dq coordinate system;

[0019] U i,d , U i,q represent the dominant subsynchronous oscillation mode components of the port voltage of the grid-connected port node i of the grid-connected system of the direct-drive wind farm in the dq coordinate system;

[0020] W port，i represents the port energy of the grid-connected port node i of the grid-connected system of the direct-drive wind farm.

[0021] In some embodiments, according to the operating principle of the grid-connected system of the direct-drive wind farm, determine the influencing factors affecting the dominant subsynchronous oscillation energy characteristics, including:

[0022] Obtain at least two system operating parameters of the direct-drive wind farm according to the operating principle of the grid-connected system of the direct-drive wind farm;

[0023] Calculate the correlation coefficients between the system operating parameters;

[0024] Screen the system operating parameters according to the correlation coefficients between the system operating parameters to obtain the influencing factors affecting the dominant subsynchronous oscillation energy characteristics.

[0025] In some embodiments, taking the influencing factors as independent variables and the energy flow power as the dependent variable, calculate the objective weight of each of the influencing factors on the energy flow power using the principal component regression method, including:

[0026] Based on n sets of influencing factors collected, a sample matrix is established, where each set of influencing factors includes at least two influencing factors, and n is a positive integer;

[0027] The influencing factors in the sample matrix are standardized to obtain a standardized matrix;

[0028] Solve the covariance matrix of the standardized matrix;

[0029] Solve the eigenvalues and corresponding eigenvectors of the covariance matrix;

[0030] Based on the eigenvectors and the standardized matrix, construct a principal component matrix;

[0031] Calculate the variance contribution and cumulative contribution of each eigenvalue;

[0032] Based on the variance contribution and cumulative contribution of each eigenvalue, determine the important principal components of the principal component matrix;

[0033] Taking the important principal components as independent variables and the energy flow power as the dependent variable, establish a regression model between the important principal components and the energy flow power;

[0034] Normalize the regression coefficients of the regression model to obtain the objective weight of each influencing factor on the energy flow power.

[0035] In some embodiments, taking the influencing factors as independent variables and the energy flow power as the dependent variable, using the improved analytic hierarchy process to calculate the subjective weight of each influencing factor on the energy flow power, including:

[0036] Sort the influencing factors in descending order according to subjective importance;

[0037] Based on the scale value definition table, determine the scale value between every two influencing factors;

[0038] According to the transitivity of the importance of rating indicators, construct a judgment matrix based on the scale values;

[0039] Based on the judgment matrix, use at least two of the following methods to calculate the weight of each influencing factor: arithmetic mean method, eigenvalue method, geometric mean method;

[0040] For each influencing factor, average the weights of this influencing factor calculated by the at least two methods to obtain the subjective weight of this influencing factor on the energy flow power.

[0041] In some embodiments, calculate the comprehensive weight of each influencing factor on the energy flow power according to the following formula:

[0042]

[0043] Wherein,

[0044] ω i represents the comprehensive weight of the i-th influencing factor on the energy flow power;

[0045] ω i ' represents the objective weight of the i-th influencing factor on the energy flow power;

[0046] ω i " represents the subjective weight of the i-th influencing factor on the energy flow power;

[0047] ω j ' represents the objective weight of the j-th influencing factor on the energy flow power;

[0048] ω j " represents the subjective weight of the j-th influencing factor on the energy flow power;

[0049] p represents the number of influencing factors.

[0050] The embodiment of the present invention further provides an analysis device for influencing factors of the energy characteristics of the dominant oscillation mode driven by response, including:

[0051] An acquisition module, configured to acquire the grid-connected point current of the direct-drive wind farm grid-connected system and the outlet voltage of the direct-drive wind farm;

[0052] An extraction module, configured to extract the dominant subsynchronous oscillation mode components of the voltage and current;

[0053] A fitting module, configured to fit the dominant subsynchronous oscillation mode components of the voltage and current obtained by extraction according to the transient energy function of the grid-connected port of the direct-drive wind farm, so as to obtain the energy flow power characterizing the energy characteristics of the corresponding oscillation mode components;

[0054] A first determination module, configured to determine the influencing factors affecting the energy characteristics of the dominant subsynchronous oscillation according to the operation principle of the direct-drive wind farm grid-connected system;

[0055] A first calculation module, configured to use the influencing factors as independent variables and the energy flow power as the dependent variable, calculate the objective weight of each influencing factor on the energy flow power by using the principal component regression method, and calculate the subjective weight of each influencing factor on the energy flow power by using the improved analytic hierarchy process;

[0056] A second calculation module, configured to calculate the comprehensive weight of each influencing factor on the energy flow power according to the objective weight and subjective weight of each influencing factor on the energy flow power;

[0057] The second determination module is configured to determine the dominant influencing factor of the dominant oscillation mode energy characteristics of the direct-drive wind farm grid-connected system according to the comprehensive weight of each influencing factor on the energy flow power.

[0058] In some embodiments, the fitting module calculates the energy flow power using the following formula:

[0059] W port,i = ∫Im[(I ij,d - jI ij,q )(dU i,d + jdU i,q )] = ∫(I ij,d dU i,q - I ij,q dU i,d );

[0060] In the formula, I ij,d , I ij,q represent the dominant subsynchronous oscillation mode components of the current of the grid-connected branch L ij of the grid-connected port node i of the direct-drive wind farm grid-connected system in the dq coordinate system;

[0061] U i,d , U i,q represent the dominant subsynchronous oscillation mode components of the port voltage of the grid-connected port node i of the direct-drive wind farm grid-connected system in the dq coordinate system;

[0062] W port，i represents the port energy of the grid-connected port node i of the direct-drive wind farm grid-connected system.

[0063] In some embodiments, the first determination module is specifically configured to:

[0064] Obtain at least two system operation parameters of the direct-drive wind farm according to the operation principle of the direct-drive wind farm grid-connected system;

[0065] Calculate the correlation coefficients between the system operation parameters;

[0066] Screen the system operation parameters according to the correlation coefficients between the system operation parameters to obtain the influencing factors that affect the dominant subsynchronous oscillation energy characteristics.

[0067] In some embodiments, the first calculation module uses the principal component regression method to calculate the objective weight of each influencing factor on the energy flow power, with the influencing factor as the independent variable and the energy flow power as the dependent variable, including:

[0068] Establish a sample matrix according to the n groups of influencing factors collected, where each group of influencing factors includes at least two influencing factors, and n is a positive integer;

[0069] Standardize the influencing factors in the sample matrix to obtain a standardized matrix;

[0070] Solve the covariance matrix of the standardized matrix;

[0071] Solve the eigenvalues and corresponding eigenvectors of the covariance matrix;

[0072] Construct a principal component matrix based on the eigenvectors and the standardized matrix;

[0073] Calculate the variance contribution and cumulative contribution of each eigenvalue;

[0074] Determine the important principal components of the principal component matrix according to the variance contribution and cumulative contribution of each eigenvalue;

[0075] Use the important principal components as independent variables and the energy flow power as the dependent variable to establish a regression model between the important principal components and the energy flow power;

[0076] Normalize the regression coefficients of the regression model to obtain the objective weight of each influencing factor on the energy flow power.

[0077] In some embodiments, the first calculation module uses the influencing factors as independent variables and the energy flow power as the dependent variable, and calculates the subjective weight of each influencing factor on the energy flow power by using the improved analytic hierarchy process, including:

[0078] Sort the influencing factors in descending order according to subjective importance;

[0079] Determine the scale value between every two influencing factors based on the scale value definition table;

[0080] Construct a judgment matrix based on the scale value according to the transitivity of the importance of the rating indicators;

[0081] Based on the judgment matrix, calculate the weight of each influencing factor by using at least two of the following methods: arithmetic mean method, eigenvalue method, geometric mean method;

[0082] For each influencing factor, average the weights of the influencing factor calculated by the at least two methods to obtain the subjective weight of the influencing factor on the energy flow power.

[0083] In some embodiments, the second calculation module calculates the comprehensive weight of each influencing factor on the energy flow power according to the following formula:

[0084]

[0085] Where,

[0086] ω i represents the comprehensive weight of the \(i\)-th influencing factor on the energy flow power;

[0087] ω i ′ represents the objective weight of the \(i\)-th influencing factor on the energy flow power;

[0088] ω i ″ represents the subjective weight of the \(i\)-th influencing factor on the energy flow power;

[0089] ω j ′ represents the objective weight of the \(j\)-th influencing factor on the energy flow power;

[0090] ω j ″ represents the subjective weight of the \(j\)-th influencing factor on the energy flow power;

[0091] p represents the number of influencing factors.

[0092] An embodiment of the present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the method for analyzing the influencing factors of the response-driven dominant oscillation mode energy characteristics described in any one of the above embodiments is implemented.

[0093] An embodiment of the present invention also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for analyzing the influencing factors of the response-driven dominant oscillation mode energy characteristics described in any one of the above embodiments is implemented.

[0094] An embodiment of the present invention also provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the method for analyzing the influencing factors of the response-driven dominant oscillation mode energy characteristics described in any one of the above embodiments is implemented.

[0095] The method and device for analyzing the influencing factors of the energy characteristics of the response-driven dominant oscillation mode provided by the embodiments of the present invention use the selected influencing factors as independent variables and the energy flow power as the dependent variable. The objective weight of each influencing factor on the energy flow power is calculated by the principal component regression method, and the subjective weight of each influencing factor on the energy flow power is calculated by the improved analytic hierarchy process. The subjective weight and objective weight information are fused, and the comprehensive weight is used to represent the influence of each influencing factor on the energy flow power. The comprehensive weight can effectively quantitatively and qualitatively characterize the relationship between each influencing factor and the energy flow power, and the result is basically consistent with the small-signal analysis result. This method has strong applicability and is not limited to a certain limited scenario analysis. Only by using measurement information can the dominant influencing factors of the energy characteristics of the system's dominant oscillation mode be identified, providing a new idea for analyzing the oscillation characteristics and suppression from the energy perspective for the new energy grid-connected system, and having certain engineering guiding significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. In the drawings:

[0097] Figure 1 It is a schematic flowchart of a method for analyzing the influencing factors of the energy characteristics of the response-driven dominant oscillation mode provided by the embodiments of the present invention.

[0098] Figure 2 It is an electromagnetic transient simulation model of a direct-drive wind farm connected to the grid through an LCC-HVDC transmission system provided by the embodiments of the present invention.

[0099] Figure 3 It is a partial schematic flowchart of a method for analyzing the influencing factors of the energy characteristics of the response-driven dominant oscillation mode provided by the embodiments of the present invention.

[0100] Figure 4 It is a partial schematic flowchart of a method for analyzing the influencing factors of the energy characteristics of the response-driven dominant oscillation mode provided by the embodiments of the present invention.

[0101] Figure 5 It is a partial schematic flowchart of a method for analyzing the influencing factors of the energy characteristics of the response-driven dominant oscillation mode provided by the embodiments of the present invention.

[0102] Figure 6 It is a schematic diagram of a method for solving the subjective weight provided by the embodiments of the present invention.

[0103] Figure 7It is a schematic flowchart of a method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by an embodiment of the present invention.

[0104] Figure 8 It is a schematic structural diagram of a device for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by an embodiment of the present invention.

[0105] Figure 9 It is a schematic structural diagram of an electronic device entity provided by an embodiment of the present invention. Detailed implementation manners

[0106] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer and more understandable, the following further describes the embodiments of the present invention in detail with reference to the accompanying drawings. Herein, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but do not limit the present invention.

[0107] The information collected in the technical solutions of the present invention is information and data authorized by the user or fully authorized by all parties. Moreover, the processing of relevant data, such as collection, storage, use, processing, transmission, provision, disclosure, and application, complies with the relevant laws, regulations, and standards of relevant countries and regions, takes necessary confidentiality measures, does not violate public order and good customs, and provides a corresponding operation entry for the user to choose to authorize or refuse.

[0108] In the technical solutions of the present invention, the acquisition, transmission, storage, use, processing, etc. of data all comply with the relevant provisions of national laws and regulations.

[0109] It should be noted that in the embodiments of the present invention, some existing industry solutions such as certain software, components, models, etc. may be mentioned. They should be regarded as exemplary, and their purpose is only to illustrate the feasibility in the implementation of the technical solutions of the present invention, but it does not mean that the inventor has used or necessarily used this solution.

[0110] The method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by the embodiments of the present invention, the execution subject includes but is not limited to a computer.

[0111] Figure 1 It is a schematic flowchart of a method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by an embodiment of the present invention. As Figure 1 shown, the method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by the embodiments of the present invention includes:

[0112] S100. Collect the grid connection point current of the direct-drive wind farm grid connection system and the outlet voltage of the direct-drive wind farm;

[0113] In step S100, a current transformer (CT) and a potential transformer (PT) can be used to measure the current and voltage at the grid connection point to obtain multiple sets of data on the subsynchronous oscillation operating state, forming a research data set. Alternatively, an electromagnetic transient simulation model of a direct-drive wind farm grid-connected system can be established, and by adjusting some system operating parameters, multiple sets of data on the subsynchronous oscillation operating state can be obtained, forming a research data set.

[0114] S200. Extract the dominant subsynchronous oscillation mode components of the voltage and current;

[0115] In step S200, the oscillation mode parameters can be obtained through machine learning or an improved adaptive variational mode decomposition method; the improved adaptive variational mode decomposition method (Adaptive Variational Mode Decomposition, VMD) is a technique for signal processing that can decompose a complex signal into a series of intrinsic mode functions (IMFs). This method is particularly suitable for analyzing non-linear and non-stationary signals, such as voltage and current signals in a power system. In a power system, subsynchronous oscillation (Sub-synchronous Oscillation, SSO) is a phenomenon that can cause equipment damage and system instability. Extracting the dominant subsynchronous oscillation mode components in voltage and current signals is crucial for understanding and controlling this phenomenon and can provide raw data for solving the energy flow power.

[0116] For example, establish an electromagnetic transient simulation model of a direct-drive wind farm transmitted through an LCC-HVDC system as shown in Figure 2 This system includes a direct-drive wind farm, an LCC-HVDC, a sending-end AC system A, and a receiving-end AC system B. Among them, the sending-end AC system A includes an AC filter bank 1, and the receiving-end AC system B includes an AC filter bank 2. The LCC-HVDC part adopts a CIGRE standard test model, and the sending and receiving-end AC systems are established based on the Thevenin equivalent method.

[0117] To smoothly carry out the evaluation of the influencing factors of the oscillation energy characteristics of the wind farm, in this embodiment, by adjusting some system operating parameters, multiple sets of data on the subsynchronous oscillation operating state are obtained, forming a research data set.

[0118] The VMD algorithm is suitable for analyzing complex non-linear signals. Its essence is to determine the center frequency and bandwidth of each mode component by searching for the optimal solution of the variational model, realizing signal decomposition and extracting characteristic information. Suppose the analysis signal X(t) is composed of M mode functions (intrinsic mode function, IMF) u m (t).

[0119]

[0120] Among them, A m (t) ≥ 0 is the envelope function of u m (t), is the phase angle of u m (t), which is a non-decreasing function, and the change rates of the amplitude and the instantaneous angular frequency are lower than the change rate of the phase function.

[0121] The construction and solution process of the variational model are as follows:

[0122] 1) Perform the Hilbert transform on each mode function to obtain its corresponding single-sided spectrum, and mix it with the respective estimated center frequency ω m (t) exponent to obtain the spectral equation as:

[0123]

[0124] In the formula: δ(t) represents the unit impulse function;

[0125] u m (t) represents the mode function, and its expression is as shown in Equation (1).

[0126] * represents the convolution operation symbol.

[0127] 2) To further evaluate the bandwidth of each mode signal, perform Gaussian smoothing on the spectral information, find M modes that minimize the sum of the estimated bandwidths of each mode, and at the same time ensure that the sum of the modes is equal to the input signal, to obtain the model as follows:

[0128]

[0129] In the formula: {u m}, {ω m} respectively represent the components obtained by decomposition and their center frequencies;

[0130] represents the partial derivative of the function with respect to time;

[0131] X(t) represents the input signal.

[0132] 3) To facilitate the solution of the above model, introduce the quadratic penalty term factor α and the Lagrange multiplier λ to obtain the unconstrained augmented Lagrangian function as follows:

[0133]

[0134] 4) Use the multiplier alternating direction algorithm to search for the "saddle point" of the above function, and at the same time update the basis function u m n+1 , the center frequency ωm n+1 When M modal components are obtained, the inner loop solution is stopped; then the gradient descent algorithm is used to update the corresponding λ m n+1 values, and double lifting is performed as the outer loop. The calculation formulas for updating each parameter are as follows:

[0135]

[0136] In the formula:

[0137] α is the quadratic penalty term factor;

[0138] ω m is the center frequency obtained from the previous iteration;

[0139] represents the analytical signal in the frequency domain state;

[0140] respectively represent the IMF component and the Lagrange multiplier in the frequency domain state. The subscript "m" and the superscript "n + 1" represent the m-th mode and the result of the (n + 1)-th iteration respectively;

[0141] τ represents the noise tolerance parameter.

[0142] Given the convergence discrimination accuracy γ > 0, if the following conditions are met at this time:

[0143]

[0144] Then the optimal solution is output. Otherwise, continue to perform cyclic and iterative solutions for the relevant parameters.

[0145] S300. According to the transient energy function of the grid-connected port of the direct-drive wind farm, fit the extracted dominant sub-synchronous oscillation modal components of the voltage and current to obtain the energy flow power characterizing the energy characteristics of the corresponding oscillation modal components;

[0146] In step S300, the transient energy function of the grid-connected port of the direct-drive wind farm can be pre-constructed. After extracting the dominant sub-synchronous oscillation modal components of the voltage and current, by fitting these components into the transient energy function, the first-order coefficient obtained is the energy flow power value characterizing the energy characteristics of the corresponding oscillation component. It should be noted here that in order to obtain a good fitting effect, try to keep the change of the oscillation amplitude within the analyzed time period small. Therefore, the analysis window is selected within the time period when the oscillation is approximately of equal amplitude.

[0147] S400. According to the operating principle of the direct-drive wind farm grid-connected system, determine the influencing factors affecting the energy characteristics of the dominant sub-synchronous oscillation;

[0148] In step S400, according to the system operation principle, select the influencing factors that may affect the subsynchronous oscillation energy characteristics; for example, wind speed, number of wind turbines, stabilized DC capacitor, grid-side converter control parameters, phase-locked loop control parameters, transmission line impedance, etc. can be selected.

[0149] S500. Take the influencing factors as independent variables and the energy flow power as the dependent variable, and use the principal component regression method to calculate the objective weight of each influencing factor on the energy flow power, and use the improved analytic hierarchy process to calculate the subjective weight of each influencing factor on the energy flow power;

[0150] In step S500, take the influencing factors as independent variables, and the dependent variable is the energy flow power corresponding to the dominant subsynchronous oscillation under the corresponding working conditions. The principal component analysis method reconstructs and combines the input features. Using orthogonal transformation technology, a set of n original variables that may be linearly correlated are transformed into m new uncorrelated variables linearly combined by n original variables (n>m). By calculating the contribution percentage of the cumulative variance of the new features, the new features that can reflect most of the original information are determined as the principal components.

[0151] The core of the traditional analytic hierarchy process lies in constructing a reasonable and consistent judgment matrix R. To determine R, it is necessary to first determine the factors to be evaluated, then establish the corresponding square matrix, then make pairwise comparisons of adjacent factors and assign a certain scale value, and finally fill it according to the consistent matrix method to obtain a complete matrix. To ensure that the judgment matrix and the solved weights are reasonable and scientific, the analytic hierarchy process is improved, and the improved analytic hierarchy process is used to calculate the subjective weight of each influencing factor on the energy flow power.

[0152] S600. Calculate the comprehensive weight of each influencing factor on the energy flow power according to the objective weight and subjective weight of each influencing factor on the energy flow power;

[0153] In step S600, to fully integrate the subjective weight and objective weight information, the comprehensive weight is used to represent the influence of each influencing factor on the energy flow power.

[0154] S700. Determine the dominant influencing factors of the dominant oscillation mode energy characteristics of the direct-drive wind farm grid-connected system according to the comprehensive weight of each influencing factor on the energy flow power.

[0155] In step S700, the dominant influencing factors of the dominant oscillation mode energy characteristics of the direct-drive wind farm grid-connected system can be determined according to the magnitude of the comprehensive weight. For example, if the comprehensive weight of the influencing factor on the energy flow power is larger, it means that the possibility of this influencing factor being the dominant influencing factor of the dominant oscillation mode energy characteristics of the direct-drive wind farm grid-connected system is higher.

[0156] The method for analyzing the influencing factors of the energy characteristics of the response-driven dominant oscillation mode provided by the embodiments of the present invention takes the screened influencing factors as independent variables and the energy flow power as the dependent variable, calculates the objective weight of each influencing factor on the energy flow power using the principal component regression method, calculates the subjective weight of each influencing factor on the energy flow power using the improved analytic hierarchy process, and fuses the subjective weight and objective weight information. The comprehensive weight is used to represent the influence of each influencing factor on the energy flow power. The comprehensive weight can effectively quantitatively and qualitatively characterize the relationship between each influencing factor and the energy flow power, and its result is basically consistent with the small-signal analysis result. This method has strong applicability and is not limited to a certain limited scenario analysis. It can identify the dominant influencing factors of the energy characteristics of the system's dominant oscillation mode only by using measurement information, providing a new idea for analyzing the oscillation characteristics and suppression from the energy perspective for the new energy grid-connected system, and having certain engineering guiding significance.

[0157] The transient energy of the components in the direct-drive wind farm grid-connected system generally includes two parts. One part is the conservative term that has nothing to do with the energy path, which shows the change of the transient energy of the component and is not conducive to analyzing the component characteristics. The other part is the non-conservative term related to the path, which represents the energy consumed or generated by the component during the change process, and it has a good corresponding relationship with the damping characteristics. This method has strong scalability and versatility and is often used for disturbance source localization.

[0158] If the grid-connected port node of a certain system is i and the grid-connected branch is Li j and the current is I ij , the port power is P i +jQ i , the terminal voltage amplitude and phase angle are U i and θ i respectively, then the port energy W port，i can be expressed as:

[0159]

[0160] In the formula: x represents the stable equilibrium point before the system fault;

[0161] x s represents the stable equilibrium point after the system fault.

[0162] And the wind farm modeling is generally based on the dq coordinate system. At this time, the port energy W port，i can be expressed as:

[0163] W port，i =∫Im(I ij，d -jI ij，q )(dU i，d +jdU i，q )]=∫(I ij，d dU i,q -I ij，qdU i，d ) (8)

[0164] wherein, Ii j,d , I ij,q represents the dominant sub-synchronous oscillation mode component of the grid-connected branch current of grid-connected port node i of the direct-drive wind farm grid-connected system in the dq coordinate system; ij in the dq coordinate system;

[0165] U i,d , U i,q represents the dominant sub-synchronous oscillation mode component of the port voltage of grid-connected port node i of the direct-drive wind farm grid-connected system in the dq coordinate system;

[0166] W port，i represents the port energy of grid-connected port node i of the direct-drive wind farm grid-connected system.

[0167] Therefore, the above formula (8) can be used to calculate the energy flow power. Specifically, if the integration interval on the right side of the equation in the above formula (8) is set to the unit time, the obtained port，i W value is the energy flow power value.

[0168] As Figure 3 shown, in some embodiments, the above step S400 may include:

[0169] S401. Obtain at least two system operation parameters of the direct-drive wind farm according to the operation principle of the direct-drive wind farm grid-connected system;

[0170] S402. Calculate the correlation coefficients between the system operation parameters;

[0171] S403. Screen the system operation parameters according to the correlation coefficients between the system operation parameters to obtain the influencing factors affecting the dominant sub-synchronous oscillation energy characteristics.

[0172] Specifically, according to the operation principle of the system, select the influencing factors that may affect the sub-synchronous oscillation energy characteristics; according to the solution results of the Pearson correlation coefficient method, conduct a preliminary screening to remove the independent variables with too strong correlation, and try to reduce the correlation and information overlap between the influencing factors.

[0173] Select the wind speed, the number of wind turbines, the stabilized DC capacitor, the grid-side converter control parameters, the phase-locked loop control parameters, and the transmission line impedance as the independent variables of the evaluation model; the dependent variable of the evaluation model is the dominant sub-synchronous oscillation corresponding energy flow power under the corresponding working conditions obtained by the improved VMD decomposition and calculation of the wind farm outlet voltage and current components under the corresponding working conditions.

[0174] Calculate the correlation coefficients between the initially selected influencing parameters of the grid-connected system of the wind farm, and reasonably screen the above initially selected parameters based on the correlation coefficients. The calculation formula of the correlation coefficient is as follows:

[0175]

[0176] In the formula: x, respectively represent the independent variable to be analyzed and its average value, and y has the same meaning.

[0177] The value of R is generally in [-1, 1]. When R is positive, it means that the two variables are positively correlated; when it is negative, it is negatively correlated. Generally, it is considered that |R| = [0, 0.2) indicates extremely weak correlation or no correlation, |R| = (0.2, 0.4) indicates weak correlation, |R| = (0.4, 0.6) indicates moderate correlation, |R| = (0.6, 0.8) indicates strong correlation, and |R| = (0.8, 1] indicates extremely strong correlation. In this embodiment, the threshold of R is set to 0.7. When the correlation coefficient between two variables is greater than 0.7, it is considered that the collinearity between the variables is too strong. In order to improve the stability of the regression analysis, it is necessary to screen the strongly correlated variables or perform principal component analysis for dimensionality reduction, so as to eliminate the redundancy and correlation between the independent variable data.

[0178] As Figure 4 shown, in some embodiments, the above step S500 may include:

[0179] S501. Establish a sample matrix according to the n groups of influencing factors collected, where each group of influencing factors includes at least two influencing factors, and n is a positive integer;

[0180] S502. Standardize the influencing factors in the sample matrix to obtain a standardized matrix;

[0181] S503. Solve the covariance matrix of the standardized matrix;

[0182] S504. Solve the eigenvalues and corresponding eigenvectors of the covariance matrix;

[0183] S505. Construct a principal component matrix according to the eigenvectors and the standardized matrix;

[0184] S506. Calculate the variance contribution and cumulative contribution of each eigenvalue;

[0185] S507. Determine the important principal components of the principal component matrix according to the variance contribution and cumulative contribution of each eigenvalue;

[0186] S508. Use the important principal components as independent variables and the energy flow power as the dependent variable to establish a regression model between the important principal components and the energy flow power;

[0187] S509. Normalize the regression coefficients of the regression model to obtain the objective weights of each influencing factor on the energy flow power.

[0188] Specifically, the steps to calculate the objective weights of each influencing factor on the energy flow power using the principal component regression method can be as follows:

[0189] 1) Data standardization. Assume that n groups of measured data are collected, with p variables (influencing factors) in each group, and an n×p sample X (sample matrix) can be constructed:

[0190]

[0191] Considering that the influencing factors to be analyzed have different dimensions and units, to ensure the reliability of data analysis, the data is standardized to obtain a standardized matrix Z, and its constituent elements are as follows:

[0192]

[0193] In the formula: σ j respectively represent the average value and standard deviation of the jth variable.

[0194]

[0195] 2) Obtain the covariance matrix R of matrix Z. The calculation formula of R is as follows:

[0196]

[0197] 3) Obtain the eigenvalues and corresponding eigenvectors of matrix R, and calculate the matrix composed of the principal components. Calculate the eigenvalues λ of R according to Ra = λa 1 ≥λ 2 ≥…≥λ p ≥0, and the corresponding eigenvectors a = [a 1 ,a 2 ,…,a p T , where a i = [a 1i ,a 2i ,…,a pi T . According to the above eigenvectors, p new index variables can be formed, and the principal component F (principal component matrix) formed is:

[0198]

[0199] ​​4) Determine the number of important principal components. For the convenience of subsequent analysis, it is necessary to comprehensively evaluate multiple new variables and select the principal components that can contain as much original information as possible. Since the variances of the principal components are decreasing, the variance contribution η j and the cumulative contribution α c are mainly calculated. Generally, it is required that the cumulative contribution rate reaches more than 85%, so as to determine the number c of important principal components and realize the dimensionality reduction of the data.

[0200]

[0201] According to the above content, c principal components F of the sample X are obtained, and a regression model between the principal components and the transient energy function is established, so as to quantitatively evaluate the influence of the corresponding influencing factors on the energy characteristics. The basic idea of stepwise regression is to introduce variables into the regression model in turn according to their contribution sizes, introduce the independent variables that pass the significance test, and eliminate those that are not significant. It should be noted that the independent variables that have been selected may lose their significance due to the introduction of new variable indicators. At this time, an F test is required to determine whether to eliminate the independent variable, so as to ensure that each fitted regression equation only contains variables that have a significant effect on the dependent variable until no new variables need to be introduced or eliminated are detected, and the regression model is obtained as follows:

[0202] y = β 0 + β 1 z 1 + … + β i z i + ε (16)

[0203] In the formula: y represents the dependent variable, which is the energy flow power calculated in this embodiment and can characterize the energy characteristics corresponding to the subsynchronous oscillation of the wind farm grid-connected system;

[0204] β i represents the regression coefficient. The larger it is, the more significant the corresponding variable's effect on the dependent variable is, and its value can be obtained by the least squares method;

[0205] ε represents a random variable with a mean of 0, representing the influence of random factors on y.

[0206] After normalizing the regression coefficients obtained by regression fitting, the objective weight ω′ j of the jth influencing factor on the energy flow power can be obtained, as shown in formula (17).

[0207]

[0208] As Figure 5 shown, in some embodiments, the above step S500 may include:

[0209] S510. Sort each of the influencing factors in a decreasing order according to the subjective importance degree.

[0210] S511. Determine the scale value between every two of the influencing factors based on the scale value definition table.

[0211] S512. Construct a judgment matrix based on the scale value according to the transitivity of the importance of the rating indicators.

[0212] S513. Based on the judgment matrix, calculate the weight of each of the influencing factors by using at least two of the following methods: arithmetic mean method, eigenvalue method, geometric mean method.

[0213] S514. For each of the influencing factors, average the weights of the influencing factor calculated by the at least two methods to obtain the subjective weight of the influencing factor on the energy flow power.

[0214] Specifically, the core of the traditional analytic hierarchy process lies in constructing a reasonable and consistent judgment matrix R. To determine R, it is necessary to first determine the factors to be evaluated, then establish the corresponding square matrix, then make pairwise comparisons of adjacent factors and assign a certain scale value, and finally fill the complete matrix according to the consistent matrix method.

[0215] Table 1. Scale value t i and its meaning

[0216]

[0217] To ensure that the judgment matrix and the obtained weights are reasonable and scientific, the analytic hierarchy process is improved. In this embodiment, according to expert opinions or user subjective requirements, the p indicators are re - sorted in a decreasing order of importance. At this time, the importance ranking of the evaluation indicators is x 1 ≥x 2 ≥…≥x p ; then make pairwise comparisons of the indicators, and determine the scale value t i of adjacent two indicators x i+1 and x i according to Table 1.

[0218] According to the transitivity of the importance of the evaluation indicators, a judgment matrix R is constructed based on the ratio scale, as shown in Equation (18). And the element r ij in the matrix R satisfies r ij = 1 / r ji = r ik r kj > 0, so as to ensure that R all satisfies consistency, and there is no need to perform consistency test and adjust the judgment matrix.

[0219]

[0220] Based on the matrix R, the weights of each factor can be directly calculated. The calculation methods mainly include the arithmetic mean method, the eigenvalue method, and the geometric mean method. To improve the robustness of the subjective analysis results, in this embodiment, three methods are used to solve the corresponding weight values, and the average value is taken as the final subjective weight ω″. The solution process is shown in Figure 6 the following.

[0221] In some embodiments, to fully integrate the subjective weight and objective weight information without favoring either side, the comprehensive weight ω can be obtained by solving the equation in formula (19) based on the principle of minimum discrimination information. i .

[0222]

[0223] The least squares method is used to solve this optimization model, and a Lagrangian function as shown in formula (20) is established, and its parameters are taken for partial differentiation.

[0224]

[0225] Thus, the comprehensive weight vector of p evaluation indicators is ω = [ω 1 , ω 2 , …, ω p T , where ω i is:

[0226]

[0227] In the formula, ω i represents the comprehensive weight of the i-th influencing factor on the energy flow power;

[0228] ω i ′ represents the objective weight of the i-th influencing factor on the energy flow power;

[0229] ω i ″ represents the subjective weight of the i-th influencing factor on the energy flow power;

[0230] ω j ′ represents the objective weight of the j-th influencing factor on the energy flow power;

[0231] ω j ″ represents the subjective weight of the j-th influencing factor on the energy flow power;

[0232] p represents the number of influencing factors.

[0233] Figure 7 ​It is a schematic flow chart of a method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by a specific embodiment of the present invention. For the detailed implementation process of each step, reference can be made to the description of the relevant part of the above embodiment.

[0234] The method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by the embodiment of the present invention aims to quantitatively evaluate the influence of wind farm parameters on subsynchronous oscillation based on response data, and establishes a comprehensive evaluation model of subsynchronous oscillation mode energy flow power - influencing parameters that combines subjective and objective factors. The primary fitting coefficient of the energy at the wind farm port - energy flow power can characterize the energy characteristics of the oscillation mode of the wind farm. The stepwise regression standard coefficients of each influencing factor obtained by solving the evaluation model can quantitatively and qualitatively characterize the relationship between each influencing factor and the energy flow power, and the results are basically consistent with the small - signal analysis results. According to the obtained comprehensive weights, it can be known that wind speed, the integral coefficient of the d - axis voltage outer - loop control of the grid - side converter, the proportional coefficient of the current inner - loop control, the number of wind turbines, the strength of the AC system, and the DC capacitor have a greater impact on the oscillation stability of the direct - drive wind farm. This method has strong applicability, is not limited to the analysis of a certain limited scenario, and can identify the dominant influencing factors of the energy characteristics of the system's dominant oscillation mode only by using measurement information, realizing the identification of the dominant factors of the system's subsynchronous oscillation (generally, the oscillation frequency of the system is expressed by the oscillation frequency of the power component), providing a new idea for analyzing the oscillation characteristics and suppression from the perspective of energy for the new - energy grid - connected system, and having certain engineering guiding significance.

[0235] Based on the same inventive concept, an apparatus for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response is also provided in the embodiment of the present invention, as described in the following embodiment. Since the principle of solving problems by this apparatus is similar to that of the method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response, the implementation of this apparatus can refer to the implementation of the method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response, and the repeated parts will not be elaborated.

[0236] As Figure 8 shown, an apparatus for a method for analyzing influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by the embodiment of the present invention includes:

[0237] A collection module 21, configured to collect the grid - connection point current of the direct - drive wind farm grid - connected system and the outlet voltage of the direct - drive wind farm;

[0238] An extraction module 22, configured to extract the dominant subsynchronous oscillation mode components of the voltage and current;

[0239] A fitting module 23, configured to fit the dominant subsynchronous oscillation mode components of the voltage and current obtained by extraction according to the transient energy function of the direct - drive wind farm grid - connection port, so as to obtain the energy flow power characterizing the energy characteristics of the corresponding oscillation mode components;

[0240] The first determination module 24 is configured to determine the influencing factors that affect the energy characteristics of the dominant subsynchronous oscillation according to the operation principle of the direct-drive wind farm grid-connected system;

[0241] The first calculation module 25 is configured to use the principal component regression method to calculate the objective weight of each influencing factor on the energy flow power, taking the influencing factors as independent variables and the energy flow power as the dependent variable, and calculate the subjective weight of each influencing factor on the energy flow power by using the improved analytic hierarchy process;

[0242] The second calculation module 26 is configured to calculate the comprehensive weight of each influencing factor on the energy flow power according to the objective weight and the subjective weight of each influencing factor on the energy flow power;

[0243] The second determination module 27 is configured to determine the dominant influencing factors of the energy characteristics of the dominant oscillation mode of the direct-drive wind farm grid-connected system according to the comprehensive weight of each influencing factor on the energy flow power.

[0244] In the analysis device for influencing factors of the energy characteristics of the dominant oscillation mode driven by response provided by the embodiment of the present invention, the selected influencing factors are used as independent variables and the energy flow power is used as the dependent variable. The principal component regression method is used to calculate the objective weight of each influencing factor on the energy flow power, and the improved analytic hierarchy process is used to calculate the subjective weight of each influencing factor on the energy flow power. The subjective weight and the objective weight information are fused, and the comprehensive weight is used to represent the influence of each influencing factor on the energy flow power. The comprehensive weight can effectively quantitatively and qualitatively characterize the relationship between each influencing factor and the energy flow power, and its result is basically consistent with the small-signal analysis result. This method has strong applicability and is not limited to a certain limited scenario analysis. Only by using the measurement information, the dominant influencing factors of the energy characteristics of the dominant oscillation mode of the system can be identified, providing a new idea for analyzing the oscillation characteristics and suppression based on the energy perspective for the new energy grid-connected system, and having certain engineering guiding significance.

[0245] In some embodiments, the fitting module calculates the energy flow power by using the following formula:

[0246] W port，i =∫Im[(I ij，d -jI ij，q )(dU i，d +jdU i，q )]=∫(I ij，d dU i，q -I ij，q dU i，d );

[0247] In the formula, I ij,d , I ij,q represent the grid-connected branch L of the grid-connected port node i of the direct-drive wind farm grid-connected systemij The dominant sub-synchronous oscillation mode component of the current in the dq coordinate system;

[0248] U i,d ,U i,q represents the dominant sub-synchronous oscillation mode component of the port voltage of the grid-connected port node i of the direct-drive wind farm grid-connected system in the dq coordinate system;

[0249] W port，i represents the port energy of the grid-connected port node i of the direct-drive wind farm grid-connected system.

[0250] In some embodiments, the first determination module is specifically configured to:

[0251] Obtain at least two system operation parameters of the direct-drive wind farm according to the operation principle of the direct-drive wind farm grid-connected system;

[0252] Calculate the correlation coefficients between the system operation parameters;

[0253] Screen the system operation parameters according to the correlation coefficients between the system operation parameters to obtain the influencing factors that affect the dominant sub-synchronous oscillation energy characteristics.

[0254] In some embodiments, the first calculation module uses the influencing factors as independent variables and the energy flow power as the dependent variable, and calculates the objective weight of each influencing factor on the energy flow power by using the principal component regression method, including:

[0255] Establish a sample matrix according to the n groups of influencing factors collected, where each group of influencing factors includes at least two influencing factors, and n is a positive integer;

[0256] Perform standardization processing on the influencing factors in the sample matrix to obtain a standardized matrix;

[0257] Solve the covariance matrix of the standardized matrix;

[0258] Solve the eigenvalues and corresponding eigenvectors of the covariance matrix;

[0259] Construct a principal component matrix according to the eigenvectors and the standardized matrix;

[0260] Calculate the variance contribution and cumulative contribution of each eigenvalue;

[0261] Determine the important principal components of the principal component matrix according to the variance contribution and cumulative contribution of each eigenvalue;

[0262] Establish a regression model between the important principal components and the energy flow power with the important principal components as independent variables and the energy flow power as the dependent variable;

[0263] Normalize the regression coefficients of the regression model to obtain the objective weights of each influencing factor on the energy flow power.

[0264] In some embodiments, the first calculation module uses the influencing factors as independent variables and the energy flow power as the dependent variable, and calculates the subjective weights of each influencing factor on the energy flow power by using the improved analytic hierarchy process, including:

[0265] Sort the influencing factors in descending order according to subjective importance;

[0266] Determine the scale value between every two influencing factors based on the scale value definition table;

[0267] Construct a judgment matrix based on the scale value according to the transitivity of the importance of rating indicators;

[0268] Based on the judgment matrix, calculate the weights of each influencing factor by using at least two of the following methods: arithmetic mean method, eigenvalue method, geometric mean method;

[0269] For each influencing factor, average the weights of this influencing factor calculated by the at least two methods to obtain the subjective weight of this influencing factor on the energy flow power.

[0270] In some embodiments, the second calculation module calculates the comprehensive weight of each influencing factor on the energy flow power according to the following formula:

[0271]

[0272] In the formula,

[0273] ω i represents the comprehensive weight of the i-th influencing factor on the energy flow power;

[0274] ω i ′ represents the objective weight of the i-th influencing factor on the energy flow power;

[0275] ω i ″ represents the subjective weight of the i-th influencing factor on the energy flow power;

[0276] ω j ′ represents the objective weight of the j-th influencing factor on the energy flow power;

[0277] ω j ″ represents the subjective weight of the j-th influencing factor on the energy flow power;

[0278] p represents the number of influencing factors.

[0279] Figure 9Schematic diagram of the physical structure of the electronic device provided by the embodiments of the present invention, as Figure 9 shown, the electronic device 003 includes a processor 301, a memory 302, and a bus 303. Among them, the processor 301 and the memory 302 communicate with each other through the bus 303.

[0280] The processor 301 is used to call program instructions in the memory 302 to execute the methods provided by the above method embodiments.

[0281] The embodiments of the present invention also provide a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above method for analyzing influencing factors of the energy characteristics of the response-driven dominant oscillation mode is implemented.

[0282] The embodiments of the present invention also provide a computer program product, which includes a computer program. When the computer program is executed by a processor, the above method for analyzing influencing factors of the energy characteristics of the response-driven dominant oscillation mode is implemented.

[0283] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0284] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0285] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the process in Figure 1one or more processes and / or blocks Figure 1 functions specified in one or more blocks.

[0286] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more processes and / or blocks Figure 1 one or more blocks.

[0287] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for analyzing the influencing factors of the energy characteristics of the dominant oscillation mode driven by response, characterized in that: include: Collect the grid connection point current of the direct-drive wind farm grid-connected system and the direct-drive wind farm outlet voltage; extracting the dominant subsynchronous oscillation modal components of the voltage and current; According to the transient energy function of the grid-connected port of the direct-drive wind farm, the extracted dominant subsynchronous oscillation modal components of the voltage and current are fitted to obtain the energy flow power that characterizes the energy characteristics of the corresponding oscillation modal components; According to the operation principle of the direct-drive wind farm grid-connected system, determining the influencing factors affecting the energy characteristics of the dominant subsynchronous oscillation; Taking the influencing factors as independent variables and the energy flow power as dependent variables, the objective weight of each influencing factor on the energy flow power is calculated using the principal component regression method, and the subjective weight of each influencing factor on the energy flow power is calculated using the improved hierarchical analysis method; Calculate the comprehensive weight of each influencing factor on the energy flow power according to the objective weight and subjective weight of each influencing factor on the energy flow power; According to the comprehensive weight of each of the influencing factors on the energy flow power, the dominant influencing factors of the dominant oscillation mode energy characteristics of the direct-drive wind farm grid-connected system are determined.

2. The method according to claim 1, characterized in that The energy flow power is calculated using the following formula: W port,i =∫Im[(I ij,d -her ij,q )(dU i,d +jdU i,q )]=∫(I ij,d dU i,q -AND ij,q dU i,d ); In the formula, I ij,d , I ij,q The grid-connected branch L of the grid-connected port node i of the direct-drive wind farm grid-connected system is represented by ij The dominant subsynchronous oscillation modal component of the current in the dq coordinate system; U i,d , U i,q represents the dominant subsynchronous oscillation modal component of the port voltage of the grid-connected port node i of the direct-drive wind farm grid-connected system in the dq coordinate system; W port，i Represents the port energy of the grid-connected port node i of the direct-drive wind farm grid-connected system.

3. The method according to claim 1, characterized in that According to the operation principle of the direct-drive wind farm grid-connected system, the factors affecting the energy characteristics of the dominant subsynchronous oscillation are determined, including: According to the operation principle of the direct-drive wind farm grid-connected system, obtaining at least two system operation parameters of the direct-drive wind farm; Calculating the correlation coefficients between the system operating parameters; According to the correlation coefficients between the system operation parameters, the system operation parameters are screened to obtain the influencing factors affecting the energy characteristics of the dominant sub-synchronous oscillation.

4. The method according to claim 1, characterized in that: The influencing factors are used as independent variables and the energy flow power is used as the dependent variable, and the objective weight of each influencing factor on the energy flow power is calculated using the principal component regression method, including: A sample matrix is ​​established according to the collected n groups of influencing factors, wherein each group of influencing factors includes at least two influencing factors, and n is a positive integer; Standardizing the influencing factors in the sample matrix to obtain a standardized matrix; Solving the covariance matrix of the standardized matrix; Solving the eigenvalues ​​and corresponding eigenvectors of the covariance matrix; Constructing a principal component matrix according to the eigenvectors and the standardized matrix; Calculate the variance contribution and cumulative contribution of each eigenvalue; Determine the important principal components of the principal component matrix according to the variance contribution and cumulative contribution of each eigenvalue; Taking the important principal component as an independent variable and the energy flow power as a dependent variable, a regression model of the important principal component and the energy flow power is established; The regression coefficient of the regression model is normalized to obtain the objective weight of each influencing factor on the energy flow power.

5. The method according to claim 1, characterized in that: Taking the influencing factors as independent variables and the energy flow power as dependent variables, the subjective weight of each influencing factor on the energy flow power is calculated using the improved hierarchical analysis method, including: Arrange the influencing factors in descending order according to their subjective importance; Determine the scale value between every two of the influencing factors based on the scale value definition table; According to the transitivity of the importance of the rating indicators, a judgment matrix is ​​constructed based on the scale values; Based on the judgment matrix, at least two of the following methods are used to calculate the weight of each of the influencing factors: arithmetic mean method, eigenvalue method, and geometric mean method; For each of the influencing factors, the weights of the influencing factor calculated by the at least two methods are averaged to obtain a subjective weight of the influencing factor on the energy flow power.

6. The method according to claim 1, characterized in that The comprehensive weight of each influencing factor on the energy flow power is calculated according to the following formula: In the formula, ω i It represents the comprehensive weight of the i-th influencing factor on the energy flow power; ω i ′ represents the objective weight of the i-th influencing factor on the energy flow power; ω i ″ represents the subjective weight of the i-th influencing factor on the energy flow power; ω j ′ represents the objective weight of the jth influencing factor on the energy flow power; ω j ″ represents the subjective weight of the jth influencing factor on the energy flow power; p represents the number of influencing factors.

7. A response-driven dominant oscillation mode energy characteristic influencing factor analysis device, characterized in that: include: A collection module, used to collect the grid connection point current of the direct-drive wind farm grid connection system and the direct-drive wind farm outlet voltage; An extraction module, used for extracting the dominant subsynchronous oscillation modal components of the voltage and current; A fitting module is used to fit the extracted dominant subsynchronous oscillation modal components of the voltage and current according to the transient energy function of the grid-connected port of the direct-drive wind farm, so as to obtain the energy flow power representing the energy characteristics of the corresponding oscillation modal components; A first determination module is used to determine the influencing factors affecting the energy characteristics of the dominant secondary synchronous oscillation according to the operation principle of the direct-drive wind farm grid-connected system; A first calculation module is used to use the influencing factors as independent variables and the energy flow power as dependent variables, calculate the objective weight of each influencing factor on the energy flow power by using the principal component regression method, and calculate the subjective weight of each influencing factor on the energy flow power by using the improved hierarchical analysis method; A second calculation module is used to calculate the comprehensive weight of each of the influencing factors on the energy flow power according to the objective weight and subjective weight of each of the influencing factors on the energy flow power; The second determination module is used to determine the dominant influencing factors of the dominant oscillation mode energy characteristics of the direct-drive wind farm grid-connected system according to the comprehensive weight of each influencing factor on the energy flow power.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images