Power grid steady-state and transient-state characteristic evaluation method and system under large-scale new energy grid connection

Abstract

The application provides a power grid steady-state and transient-state characteristic evaluation method and system under large-scale new energy grid connection, relates to the technical field of power systems, and comprises the following steps: acquiring power grid topological relations and new energy output data, performing impedance equivalent transformation to obtain a simplified topology; performing singular value decomposition on the basis of a piecewise admittance function to determine a steady-state weak node; selecting the weak node closest in electrical distance as a fault point, calculating a transient-state response, and obtaining system transient-state characteristics through Hilbert transformation. The application can accurately evaluate power grid steady-state and transient-state characteristics and provide stability guidance for large-scale new energy grid connection.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method and system for evaluating the steady-state and transient characteristics of a power grid under large-scale new energy grid integration. Background Technology

[0002] With the advancement of the global energy transition, the proportion of new energy power generation, such as wind power and photovoltaic power generation, in the power grid is constantly increasing. The large-scale grid connection of new energy sources brings new challenges to the steady-state and transient characteristics of the power grid. Compared with traditional synchronous generators, new energy power generation units have completely different dynamic characteristics, and the power control characteristics based on power electronic devices make the assessment of the operational stability of the power grid more complex.

[0003] Traditional methods for assessing the steady-state and transient characteristics of power grids are mainly based on the dynamic characteristics of synchronous generators. They typically employ power flow calculations and transient stability analysis. However, with the increasing proportion of renewable energy connected to the grid, existing technologies have limitations. They cannot accurately reflect the power and voltage characteristics of renewable energy generation units, cannot fully consider the dynamic impact of renewable energy power output fluctuations on weak links in the system, resulting in significant errors in the identification results. Furthermore, they lack in-depth analysis of the interaction between the power control loop and the current control loop of renewable energy generation units, making it difficult to accurately assess the dynamic response characteristics and stability margin of the system under disturbances. Summary of the Invention

[0004] This invention provides a method and system for evaluating the steady-state and transient characteristics of a power grid under large-scale renewable energy grid integration, which can at least solve some of the problems existing in the prior art.

[0005] A first aspect of this invention provides a method for evaluating the steady-state and transient characteristics of a power grid under large-scale renewable energy grid integration, comprising: The topology of the power grid system, the access nodes of new energy power generation units and their corresponding output time series are obtained. Non-critical nodes in the topology are identified and impedance equivalent transformation is performed to obtain a simplified topology. Extract the power-voltage response curve corresponding to the output time series and perform piecewise linearization to obtain the piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, calculate the minimum singular value and the corresponding singular vector through singular value decomposition. For each new energy power generation unit access node, extract the node corresponding to the element with the largest amplitude in the singular vector as the steady-state weak node. The steady-state weak node with the closest electrical distance to the access node of the new energy power generation unit is taken as the fault application point. A three-phase short-circuit disturbance is set at the fault application point, and the transient response of the power control loop and the current control loop in the new energy power generation unit is solved to obtain the transient trajectory corresponding to the port of the new energy power generation unit. The transient trajectory is subjected to Hilbert transform to obtain the instantaneous amplitude envelope and instantaneous frequency curve. The maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve are calculated. The transient stability margin is determined based on the maximum descent depth and the attenuation coefficient, and an evaluation result is generated.

[0006] In one alternative implementation, Identifying non-critical nodes in the topological connections and performing impedance equivalent transformations to obtain simplified topological connections includes: Based on the topological connection relationship, a node association matrix is ​​constructed. The set of shortest electrical paths from the access node of the new energy power generation unit to the system balance node is extracted. The frequency of occurrence of each access node of the new energy power generation unit in the set of shortest electrical paths is counted to obtain the path frequency statistics. After performing the node deletion operation, the algebraic connectivity change rate of the node association matrix is ​​calculated. Non-critical nodes are selected based on the path frequency statistics and the algebraic connectivity change rate. The non-critical nodes are sorted by degree, and the set of adjacent nodes and branch impedance values ​​of each non-critical node are extracted in turn. The equivalent impedance value is obtained by selecting the impedance equivalent calculation method according to the number of nodes in the set of adjacent nodes. When there are two nodes, the series equivalent calculation method is used. When there are three nodes, the star-triangle equivalent transformation calculation method is used. When there are more than three nodes, the admittance matrix block elimination calculation method is used. The equivalent impedance value is updated to the node correlation matrix and the non-critical nodes are deleted. After processing, the simplified topology connection relationship is formed.

[0007] In one alternative implementation, The power-voltage response curve corresponding to the output time series is extracted and piecewise linearized to obtain a piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, the minimum singular value and the corresponding singular vector are calculated through singular value decomposition. For each new energy power generation unit access node, the node corresponding to the element with the largest amplitude in the singular vector is extracted as a steady-state weak node, including: The power output value and the corresponding voltage measurement value are extracted from the output time series and curve fitting is performed to obtain the power-voltage response curve. Based on the slope change characteristics of the power-voltage response curve in different power ranges, multiple power range segments are obtained by identifying segmentation points through an adaptive curvature detection algorithm. Linear regression fitting is performed on the power range segments to obtain linearized slope coefficients, and the piecewise admittance value corresponding to the power range segment is determined according to the linearized slope coefficients to obtain the piecewise admittance function. Based on the simplified topology connection relationship, a node admittance matrix is ​​constructed, and the self-admittance element corresponding to the access node of the new energy power generation unit is replaced with the piecewise admittance value corresponding to the current operating power in the piecewise admittance function to obtain the modified node admittance matrix. Perform singular value decomposition on the modified node admittance matrix to obtain a singular value sequence and a right singular vector matrix. Extract the singular value with the smallest value in the singular value sequence as the minimum singular value. Extract the singular vector in the right singular vector matrix that corresponds to the minimum singular value. Calculate the magnitude of each element in the singular vector and establish a mapping relationship with the node number in the simplified topology connection relationship. For each new energy power generation unit access node, after excluding itself from the mapping relationship, the node corresponding to the element with the largest amplitude among the remaining elements is extracted as the steady-state weak node.

[0008] In one alternative implementation, Based on the slope variation characteristics of the power-voltage response curve in different power ranges, multiple power range segments are obtained by identifying segmentation points using an adaptive curvature detection algorithm, including: The power-voltage response curve is subjected to adaptive density sampling along the power axis to obtain multiple sampling points. For each sampling point, adjacent sampling points are selected and the first derivative value is calculated. The first derivative value is subjected to second difference and a filter is applied to remove high-frequency noise to obtain the filtered second derivative value. The absolute value of the second derivative of the filter is extracted to form a curvature amplitude sequence. Statistical operations are performed on the curvature amplitude sequence to obtain the mean and standard deviation, and a weighted sum is obtained to obtain the curvature threshold. The difference between adjacent elements of the curvature amplitude sequence is calculated and the abrupt change position is identified as the curvature amplitude abrupt change position. Traverse the sampling points and remove sampling points whose absolute value of the second derivative of the filter is less than the curvature threshold and are not located at the position of curvature amplitude change. Add the retained sampling points as candidate segmentation points to the candidate segmentation point list. Temporary power intervals are divided based on candidate segmentation points, and linear regression fitting is performed to obtain linearized fitting error. If the linearized fitting error exceeds a preset error threshold, segmentation points are added. If the linearized fitting error of adjacent temporary power intervals is less than a preset proportion of the error threshold, segmentation points are deleted. The process of adding and deleting is repeated to obtain the final set of segmentation points. The starting power point and ending power point corresponding to the power-voltage response curve are combined to form a segmentation boundary sequence and divide multiple power intervals.

[0009] In one alternative implementation, The steady-state weak node with the closest electrical distance to the new energy power generation unit is taken as the fault application point. A three-phase short-circuit disturbance is set at the fault application point, and the transient response of the power control loop and current control loop in the new energy power generation unit is solved to obtain the transient trajectory corresponding to the port of the new energy power generation unit, including: Extract the node number of the steady-state weak node, calculate the electrical distance between the new energy power generation unit access node and each steady-state weak node based on the simplified topology connection relationship, and take the steady-state weak node with the smallest electrical distance as the fault application point; A three-phase short-circuit disturbance is set at the fault application point and the three-phase voltage amplitude is set to zero. Based on the node number, the corresponding row and column elements are located in the pre-obtained modified node admittance matrix and the self-admittance element corresponding to the fault application point is modified to the fault admittance value to obtain the fault state node admittance matrix. The current reference value output by the power control loop of the new energy power generation unit is used as the input data of the current control loop to establish a coupled calculation relationship. Based on the admittance matrix of the fault state node and the initial operating state corresponding to the access node of the new energy power generation unit, the node voltage transient change sequence is obtained by iteratively solving through the prediction and correction algorithm. Based on the node voltage transient change sequence and the coupling calculation relationship, the dual-loop control response time sequence is obtained. The instantaneous power time sequence and port voltage time sequence are extracted from the dual-loop control response time sequence and the trajectory is mapped in the power-voltage plane to obtain the transient trajectory corresponding to the port of the new energy power generation unit.

[0010] In one alternative implementation, Based on the fault state node admittance matrix and the initial operating state corresponding to the access node of the new energy power generation unit, the node voltage transient change sequence is obtained by iteratively solving the prediction and correction algorithm, including: Extract the initial value of the node voltage in the initial operating state corresponding to the access node of the new energy power generation unit, establish the node voltage differential equation based on the admittance matrix of the fault state node, set the integration time step and initialize the current time as the fault occurrence time; Within each integration time step, the forward Euler method is used to calculate the predicted value of the node voltage at the current time based on the node voltage value and the differential value of the node voltage at the previous time step. The backward Euler method is used to recalculate the differential value of the node voltage at the current moment based on the predicted value of the node voltage and the admittance matrix of the fault state node, and the corrected value of the node voltage at the current moment is obtained. It is determined whether the difference between the corrected value of the node voltage and the predicted value of the node voltage is less than a preset convergence threshold. If it is less than the threshold, the corrected value of the node voltage is used as the determined value of the node voltage at the current moment and recorded in the node voltage transient change sequence. The current moment is updated to the moment corresponding to the next integration time step, and the prediction and steps are repeated. If the value is not less than the target value, the corrected value of the node voltage is used as the new predicted value and the correction is repeated until the convergence condition is met or the simulation ends. The transient change sequence of the node voltage is then output.

[0011] In one alternative implementation, The transient trajectory is subjected to Hilbert transform to obtain the instantaneous amplitude envelope and instantaneous frequency curve. The maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve are calculated. The transient stability margin is determined based on the maximum descent depth and the attenuation coefficient, and the evaluation results are generated, including: The coordinate point sequence of the transient trajectory in the power-voltage plane is extracted as a time series signal. The time series signal is subjected to Hilbert transform to obtain an analytical signal. The real and imaginary parts are extracted from the analytical signal and the instantaneous amplitude envelope and instantaneous phase curve are calculated. The instantaneous frequency curve is obtained by time differentiation of the instantaneous phase curve. Identify the minimum amplitude point from the instantaneous amplitude envelope and calculate the amplitude difference between the minimum amplitude point and the pre-acquired initial steady-state operating point as the maximum descent depth; Spectral analysis is performed on the instantaneous frequency curve to extract the frequency component with the largest amplitude as the dominant oscillation frequency component. The oscillation waveform corresponding to the dominant oscillation frequency component is extracted from the instantaneous frequency curve, and the envelope of the oscillation waveform is fitted with an exponential function to obtain the attenuation coefficient. The maximum descent depth is compared with a preset depth threshold to obtain a depth margin index, the attenuation coefficient is compared with a preset attenuation threshold to obtain an attenuation margin index, and the depth margin index and the attenuation margin index are weighted and summed to obtain a transient stability margin. The transient stability margin is associated with the steady-state weak node to obtain the evaluation result and output it.

[0012] A second aspect of this invention provides a system for evaluating the steady-state and transient characteristics of a power grid under large-scale renewable energy grid integration, comprising: The topology simplification module is used to obtain the topology connection relationship of the power grid system, the access node of the new energy power generation unit and the corresponding output time series, identify the non-critical nodes in the topology connection relationship and perform impedance equivalent transformation to obtain the simplified topology connection relationship. The weak point identification module is used to extract the power-voltage response curve corresponding to the output time series and perform piecewise linearization to obtain the piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, the minimum singular value and the corresponding singular vector are calculated by singular value decomposition. For each new energy power generation unit access node, the node corresponding to the element with the largest amplitude in the singular vector is extracted as the steady-state weak node. The transient solution module is used to take the steady-state weak node with the closest electrical distance to the access node of the new energy power generation unit as the fault application point, set a three-phase short-circuit disturbance at the fault application point, and solve the transient response of the power control loop and the current control loop in the new energy power generation unit to obtain the transient trajectory corresponding to the port of the new energy power generation unit. The margin assessment module is used to perform Hilbert transform on the transient trajectory to obtain the instantaneous amplitude envelope and instantaneous frequency curve, calculate the maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve, determine the transient stability margin based on the maximum descent depth and the attenuation coefficient, and generate an assessment result.

[0013] A third aspect of the present invention provides an electronic device, comprising: A processor and a memory for storing processor-executable instructions, wherein the processor is configured to invoke instructions stored in the memory to perform the aforementioned method.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0015] In this invention, a simplified topology is obtained by identifying non-critical nodes and performing impedance equivalent transformation, effectively reducing computational complexity and making the evaluation of large-scale power grid systems more efficient and feasible. Piecewise linearization is used to process the power-voltage response curves corresponding to the output time series, resulting in piecewise admittance functions, which can more accurately characterize the nonlinear characteristics of new energy power generation units and improve the accuracy of steady-state evaluation. Based on the singular value decomposition method, the minimum singular value and corresponding singular vector are calculated to scientifically identify steady-state weak nodes, improving the objectivity and accuracy of the evaluation results. The steady-state weak node with the closest electrical distance to the new energy power generation unit access node is selected as the fault application point, making the evaluation more targeted and able to more effectively expose potential transient stability problems of the system. The transient trajectory is processed by Hilbert transform, and the maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve are calculated, providing a quantitative transient stability margin evaluation index, making the evaluation results more intuitive and accurate. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the method for evaluating the steady-state and transient characteristics of the power grid under large-scale renewable energy grid connection, according to an embodiment of the present invention. Figure 2 This is a flowchart illustrating the voltage transient change solution of the power grid steady-state and transient characteristic evaluation method under large-scale new energy grid connection in an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0019] Figure 1 This is a flowchart illustrating the method for evaluating the steady-state and transient characteristics of a power grid under large-scale renewable energy grid integration, as described in this embodiment of the invention. Figure 1 As shown, the method includes: The topology of the power grid system, the access nodes of new energy power generation units and their corresponding output time series are obtained. Non-critical nodes in the topology are identified and impedance equivalent transformation is performed to obtain a simplified topology. Extract the power-voltage response curve corresponding to the output time series and perform piecewise linearization to obtain the piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, calculate the minimum singular value and the corresponding singular vector through singular value decomposition. For each new energy power generation unit access node, extract the node corresponding to the element with the largest amplitude in the singular vector as the steady-state weak node. The steady-state weak node with the closest electrical distance to the access node of the new energy power generation unit is taken as the fault application point. A three-phase short-circuit disturbance is set at the fault application point, and the transient response of the power control loop and the current control loop in the new energy power generation unit is solved to obtain the transient trajectory corresponding to the port of the new energy power generation unit. The transient trajectory is subjected to Hilbert transform to obtain the instantaneous amplitude envelope and instantaneous frequency curve. The maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve are calculated. The transient stability margin is determined based on the maximum descent depth and the attenuation coefficient, and an evaluation result is generated.

[0020] In one alternative implementation, Identifying non-critical nodes in the topological connections and performing impedance equivalent transformations to obtain simplified topological connections includes: Based on the topological connection relationship, a node association matrix is ​​constructed. The set of shortest electrical paths from the access node of the new energy power generation unit to the system balance node is extracted. The frequency of occurrence of each access node of the new energy power generation unit in the set of shortest electrical paths is counted to obtain the path frequency statistics. After performing the node deletion operation, the algebraic connectivity change rate of the node association matrix is ​​calculated. Non-critical nodes are selected based on the path frequency statistics and the algebraic connectivity change rate. The non-critical nodes are sorted by degree, and the set of adjacent nodes and branch impedance values ​​of each non-critical node are extracted in turn. The equivalent impedance value is obtained by selecting the impedance equivalent calculation method according to the number of nodes in the set of adjacent nodes. When there are two nodes, the series equivalent calculation method is used. When there are three nodes, the star-triangle equivalent transformation calculation method is used. When there are more than three nodes, the admittance matrix block elimination calculation method is used. The equivalent impedance value is updated to the node correlation matrix and the non-critical nodes are deleted. After processing, the simplified topology connection relationship is formed.

[0021] This process acquires the topology of the power grid system, the access nodes of renewable energy generation units, and their corresponding power output time series data. The topology data includes node numbers, branch connections, and branch impedance parameters; the renewable energy access node data includes the access location of each renewable energy generation unit in the grid; and the power output time series data includes the power output values ​​of each renewable energy generation unit at different time points. For example, consider a power grid system with 100 nodes, including 15 renewable energy generation units connected to nodes 23, 27, 35, 42, 49, 56, 61, 68, 72, 77, 83, 87, 91, 95, and 98.

[0022] A node affinity matrix is ​​constructed based on the obtained topological connections. The node affinity matrix is ​​a symmetric matrix, its size being the number of nodes multiplied by the number of nodes. The elements in the matrix represent the connections between nodes. If two nodes are directly connected, the value of the corresponding element is the reciprocal of the impedance of their connecting branch; if two nodes are not directly connected, the value of the corresponding element is 0. In the previous example, if nodes 25 and 36 are connected by a line with an impedance of 0.025 ohms, the value of the corresponding element in the node affinity matrix is ​​40.

[0023] The shortest electrical path set is extracted from the renewable energy generation unit access node to the system balancing node. The system balancing node is usually the reference node of the power grid, and is set as node 1 in the previous example. A breadth-first search algorithm is used to determine the shortest path from each renewable energy access node to the balancing node. For example, the shortest path from node 23 to node 1 is node 23-node 18-node 12-node 7-node 1; the shortest path from node 27 to node 1 is node 27-node 20-node 15-node 10-node 5-node 1. Similarly, the shortest paths from all 15 renewable energy access nodes to the balancing node can be obtained.

[0024] The frequency of each node in the shortest electrical path set is counted to obtain the path frequency statistics. The higher the frequency of a node, the greater its importance in the power transmission path. In the example above, node 1 appears 15 times; node 5 appears 8 times; node 7 appears 6 times; node 10 appears 5 times; nodes 12 and 15 appear 4 and 3 times respectively; while most intermediate nodes appear 1 or 2 times.

[0025] After performing a node deletion operation, the rate of change of algebraic connectivity in the node incidence matrix is ​​calculated. Algebraic connectivity is the second smallest eigenvalue of the node incidence matrix, reflecting the network's connectivity performance. The algebraic connectivity is recalculated after deleting each node in turn, yielding the rate of change before and after deletion. In the previous example, the algebraic connectivity of the original topology was 0.217. After deleting node 30, the algebraic connectivity became 0.213, with a rate of change of 1.84%; after deleting node 45, the algebraic connectivity became 0.206, with a rate of change of 5.07%.

[0026] Non-critical nodes were identified based on path frequency statistics and the rate of change of algebraic connectivity. A path frequency threshold of 2 and a rate of change of algebraic connectivity threshold of 5% were set. A node was classified as non-critical when both its path frequency and rate of change of algebraic connectivity were below the thresholds. In the aforementioned example, nodes 30, 40, 55, 63, 79, and 92 were identified as non-critical nodes.

[0027] The non-critical nodes selected through the screening are sorted by their degree, with nodes having higher degree ranking higher. The degree of a node refers to the number of other nodes directly connected to it. In the example above, node 55 has a degree of 4, node 63 has a degree of 3, nodes 40 and 92 both have a degree of 2, and nodes 30 and 79 both have a degree of 1. The node order after sorting by degree is: node 55, node 63, node 40, node 92, node 30, node 79.

[0028] Each non-critical node is processed sequentially, and its set of adjacent nodes and the impedance values ​​of the connected branches are extracted. For node 55, its adjacent nodes are nodes 32, 46, 67 and 81, and the impedances of the connected branches are 0.030 ohms, 0.025 ohms, 0.035 ohms and 0.020 ohms, respectively.

[0029] The appropriate impedance equivalence calculation method is selected based on the number of nodes in the adjacent node set. For the case with two adjacent nodes, a series equivalence calculation method is used, where the equivalent impedance is the sum of the impedances of the two branches. For example, when processing node 40, its adjacent nodes are nodes 31 and 52, with connection impedances of 0.015 ohms and 0.020 ohms respectively. After equivalence, the impedance between nodes 31 and 52 is 0.035 ohms.

[0030] For the case with 3 adjacent nodes, the star-delta equivalent transformation calculation method is used. When processing node 63, its adjacent nodes are nodes 47, 58, and 74, with connection impedances of 0.018 ohms, 0.022 ohms, and 0.015 ohms, respectively. After equivalence, the equivalent impedance between nodes 47 and 58 is 0.040 ohms, the equivalent impedance between nodes 47 and 74 is 0.033 ohms, and the equivalent impedance between nodes 58 and 74 is 0.037 ohms.

[0031] For cases with more than three adjacent nodes, a block-based elimination method for the admittance matrix is ​​used. When processing node 55, the branch impedance is converted into admittance values. The admittances of node 55 and its four adjacent nodes are 33.333, 40.000, 28.571, and 50.000, respectively. The admittance matrix is ​​constructed and divided into blocks. Node 55 is eliminated through matrix operations to obtain an equivalent admittance matrix. Then, the admittance values ​​are converted back to impedance values ​​to update the node correlation matrix.

[0032] The calculated equivalent impedance value is used to update the node incidence matrix, and the corresponding non-critical node rows and columns are deleted from the matrix. For example, after processing node 40, the element value between nodes 31 and 52 is updated to 28.571 in the node incidence matrix, and the row and column corresponding to node 40 are deleted.

[0033] After processing all non-critical nodes, a simplified topology is formed. In the aforementioned example, the original 100 nodes were simplified to 94 critical nodes, a simplification rate of 6%. Testing verified that the voltage amplitude error at critical nodes before and after simplification was less than 0.3%, the phase angle error was less than 0.2 degrees, and the power flow deviation was less than 1%, meeting the accuracy requirements for power grid analysis.

[0034] In this embodiment, by comprehensively considering the frequency of the shortest electrical path from the access node of the new energy power generation unit to the system balance node and the rate of change of the proximal connectivity before and after node deletion, the importance of nodes can be objectively measured from two dimensions: global electrical connectivity and the criticality of new energy power transmission. This allows for the accurate differentiation of non-critical nodes that have a relatively small impact on system operation and power transmission, improving the reliability and safety margin of the topology simplification results. By implementing orderly deletion of non-critical nodes and adaptively selecting series equivalent, star-delta transformation, or admittance matrix block elimination impedance equivalent methods based on the number of their adjacent nodes, node compression is achieved while ensuring the continuity of equivalent electrical characteristics. This effectively maintains the equivalent impedance distribution and power flow transmission characteristics of the original network, significantly reducing the risk of electrical parameter distortion and improving the approximation accuracy of the simplified network to the behavior of the original system.

[0035] In one alternative implementation, The power-voltage response curve corresponding to the output time series is extracted and piecewise linearized to obtain a piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, the minimum singular value and the corresponding singular vector are calculated through singular value decomposition. For each new energy power generation unit access node, the node corresponding to the element with the largest amplitude in the singular vector is extracted as a steady-state weak node, including: The power output value and the corresponding voltage measurement value are extracted from the output time series and curve fitting is performed to obtain the power-voltage response curve. Based on the slope change characteristics of the power-voltage response curve in different power ranges, multiple power range segments are obtained by identifying segmentation points through an adaptive curvature detection algorithm. Linear regression fitting is performed on the power range segments to obtain linearized slope coefficients, and the piecewise admittance value corresponding to the power range segment is determined according to the linearized slope coefficients to obtain the piecewise admittance function. Based on the simplified topology connection relationship, a node admittance matrix is ​​constructed, and the self-admittance element corresponding to the access node of the new energy power generation unit is replaced with the piecewise admittance value corresponding to the current operating power in the piecewise admittance function to obtain the modified node admittance matrix. Perform singular value decomposition on the modified node admittance matrix to obtain a singular value sequence and a right singular vector matrix. Extract the singular value with the smallest value in the singular value sequence as the minimum singular value. Extract the singular vector in the right singular vector matrix that corresponds to the minimum singular value. Calculate the magnitude of each element in the singular vector and establish a mapping relationship with the node number in the simplified topology connection relationship. For each new energy power generation unit access node, after excluding itself from the mapping relationship, the node corresponding to the element with the largest amplitude among the remaining elements is extracted as the steady-state weak node.

[0036] Based on the acquired power output time series data, the power output values ​​and corresponding voltage measurement values ​​of the new energy power generation units were extracted. Taking a certain new energy power generation unit as an example, samples were taken every 15 minutes from the 24-hour power output time series, resulting in 96 power-voltage data pairs. The power values ​​ranged from 0 to 60 MW, and the corresponding voltage values ​​ranged from 0.97 to 1.03 per unit. These data points were plotted on a coordinate graph, with the horizontal axis representing the power output value and the vertical axis representing the voltage measurement value. Curve fitting was then performed to obtain the power-voltage response curve.

[0037] Based on the slope variation characteristics of the power-voltage response curve across different power ranges, an adaptive curvature detection algorithm is used to identify segmentation points of the curve. This is achieved by calculating the curvature value at each point on the curve and setting a curvature threshold of 0.05. When the curvature value exceeds the threshold, the point is marked as a potential segmentation point. To avoid excessive segmentation, a minimum interval of 5 MW is set between adjacent segmentation points. In the aforementioned example, the curvature detection algorithm identified three segmentation points located at 12 MW, 27 MW, and 45 MW, respectively, thus dividing the power range into four segments: 0 to 12 MW, 12 to 27 MW, 27 to 45 MW, and 45 to 60 MW.

[0038] Linear regression was performed on each power range to obtain the linearized slope coefficients. In the 0-12 MW range, the slope was 0.0010, indicating that for every 1 MW increase in power, the voltage increased by 0.0010 per unit. In the 12-27 MW range, the slope was 0.0003; in the 27-45 MW range, the slope was -0.0002; and in the 45-60 MW range, the slope was -0.0008. The piecewise admittance value for each range was calculated based on the linearized slope coefficients by multiplying the inverse of the slope by a baseline coefficient. With the baseline coefficient set to 100, the piecewise admittance values ​​for the four ranges were 1000, 3333, -5000, and -1250, respectively, in megawatts per unit, forming the piecewise admittance function.

[0039] Based on the established simplified topology, a node admittance matrix is ​​constructed. The node admittance matrix is ​​a complex matrix where diagonal elements represent the self-admittance of each node, and off-diagonal elements represent the mutual admittance between nodes. In the previous example, the simplified power grid contains 94 nodes, and the node admittance matrix is ​​94 x 94. The self-admittance element value is determined by the sum of the admittances of all branches connected to the current node, while the mutual admittance element value is the negative of the admittance of the branch connecting two nodes. For example, if the branch impedance between node 5 and node 8 is 0.03 ohms, the corresponding admittance value is 33.33. The elements at positions 5,5 and 8,8 in the node admittance matrix need to be increased by 33.33, while the elements at positions 5,8 and 8,5 need to be decreased by 33.33.

[0040] Replace the self-admittance element corresponding to the new energy power generation unit access node with the piecewise admittance value corresponding to the current operating power in the piecewise admittance function. Assume the current power output of new energy power generation unit access node 23 is 35 MW. According to the piecewise admittance function, the power value is in the range of 27 to 45 MW, and the corresponding piecewise admittance value is -5000. Increase the self-admittance element value at position 23 in the node admittance matrix by -5000 to obtain the corrected node admittance matrix. Perform the same operation on all 15 new energy power generation unit access nodes, finding the corresponding piecewise admittance value based on their current power output and updating the self-admittance element accordingly.

[0041] Singular value decomposition (SVD) is performed on the modified nodal admittance matrix to obtain a sequence of singular values ​​and a right singular vector matrix. SVD is an operation that decomposes a matrix into the product of three matrices: the left singular vector matrix, the singular value diagonal matrix, and the transpose of the right singular vector matrix. The singular value sequence, arranged in descending order, represents the "strength" of the matrix in different directions. In the previous example, the singular value sequence of the modified nodal admittance matrix is ​​820.5, 745.2, 698.3, ..., 0.156. The smallest singular value in the sequence is extracted as the minimum singular value, which is 0.156 in this example.

[0042] Extract the singular vector corresponding to the minimum singular value from the right singular vector matrix. The singular vector represents the weakest eigendirection of the admittance matrix. Calculate the magnitude of each element in this singular vector and establish a mapping relationship with the node numbers in the simplified topological connectivity. In the example, the singular vector has 94 element magnitudes: 0.012, 0.025, 0.037, ..., corresponding to node numbers 1 to 94.

[0043] For each new energy power generation unit access node, after excluding itself in the mapping relationship, the node corresponding to the element with the largest amplitude among the remaining elements is extracted as the steady-state weak node of that new energy power generation unit. For example, for access node 23, after excluding itself, the element with the largest amplitude in the singular vector corresponds to node 41, so node 41 is the steady-state weak node of node 23. Similarly, the steady-state weak nodes of other new energy power generation unit access nodes are determined. The steady-state weak node of node 27 is node 52, the steady-state weak node of node 35 is node 64, and so on.

[0044] In this embodiment, by fitting power and voltage data in the output time series and constructing a power-voltage response curve, and combining adaptive curvature detection to achieve automatic segmentation of the power range, the nonlinear electrical characteristics of the new energy power generation unit under different operating ranges can be accurately characterized. This effectively avoids model distortion caused by changes in operating points, making the admittance modeling more consistent with the actual operating state and improving the accuracy of steady-state analysis. The piecewise admittance function is dynamically introduced into the node admittance matrix to correct the self-admittance of the new energy access node, which significantly enhances the ability to characterize the volatility of new energy and improves the robustness of the steady-state analysis results. By performing singular value decomposition on the corrected node admittance matrix and performing node correlation strength analysis based on the minimum singular value and its corresponding singular vector, the system can identify the associated nodes that are most coupled with the new energy access node and have the greatest impact on steady-state security from the perspective of the overall system stability margin. This effectively avoids misjudgment and improves the global consistency and physical rationality of weak node identification.

[0045] In one alternative implementation, Based on the slope variation characteristics of the power-voltage response curve in different power ranges, multiple power range segments are obtained by identifying segmentation points using an adaptive curvature detection algorithm, including: The power-voltage response curve is subjected to adaptive density sampling along the power axis to obtain multiple sampling points. For each sampling point, adjacent sampling points are selected and the first derivative value is calculated. The first derivative value is subjected to second difference and a filter is applied to remove high-frequency noise to obtain the filtered second derivative value. The absolute value of the second derivative of the filter is extracted to form a curvature amplitude sequence. Statistical operations are performed on the curvature amplitude sequence to obtain the mean and standard deviation, and a weighted sum is obtained to obtain the curvature threshold. The difference between adjacent elements of the curvature amplitude sequence is calculated and the abrupt change position is identified as the curvature amplitude abrupt change position. Traverse the sampling points and remove sampling points whose absolute value of the second derivative of the filter is less than the curvature threshold and are not located at the position of curvature amplitude change. Add the retained sampling points as candidate segmentation points to the candidate segmentation point list. Temporary power intervals are divided based on candidate segmentation points, and linear regression fitting is performed to obtain linearized fitting error. If the linearized fitting error exceeds a preset error threshold, segmentation points are added. If the linearized fitting error of adjacent temporary power intervals is less than a preset proportion of the error threshold, segmentation points are deleted. The process of adding and deleting is repeated to obtain the final set of segmentation points. The starting power point and ending power point corresponding to the power-voltage response curve are combined to form a segmentation boundary sequence and divide multiple power intervals.

[0046] Adaptive density sampling is performed on the power-voltage response curve along the power axis, with the sampling density dynamically adjusted according to the curve's rate of change. In regions of significant curve variation, the sampling interval is set to 0.5 MW; in regions of gradual change, the sampling interval can be increased to 2 MW. Taking a new energy power generation unit with a power range of 0 to 60 MW as an example, a total of 85 sampling points were obtained through adaptive density sampling. The sampling point density is higher in the power output ranges of 10 to 15 MW and 40 to 50 MW.

[0047] For each sampling point, the first derivative is calculated using adjacent sampling points, employing the central difference method. This involves dividing the voltage difference between two adjacent sampling points by the corresponding power difference. For example, a sampling point with a power value of 25 MW has a voltage value of 1.012 per unit. The previous sampling point had a power value of 23.5 MW and a voltage value of 1.010 per unit, while the next sampling point had a power value of 26.5 MW and a voltage value of 1.013 per unit. The calculated first derivative is 0.00067 per unit per MW, meaning that near this point, for every 1 MW increase in power, the voltage increases by 0.00067 per unit.

[0048] The calculated first derivative value is subjected to a second difference to obtain the second derivative value. The second difference calculation method is to divide the difference between adjacent first derivatives by the corresponding power difference. For a 25 MW sampling point, the first derivatives of two adjacent points are 0.00080 and 0.00045, respectively. The calculated second derivative is -0.00012 per unit squared per MW, representing the rate of change of the first derivative. Since the sampling process may introduce noise, a filter is applied to the second derivative value to remove high-frequency noise. A moving average filter with a window width of 5 is used to weight the second derivative values ​​of the current point and the two points before and after it, with weights set to 0.1, 0.2, 0.4, 0.2, and 0.1, respectively. After filtering, the second derivative of the 25 MW sampling point is corrected to -0.00009 per unit squared per MW.

[0049] The absolute values ​​of the filtered second derivatives are extracted to form a curvature amplitude sequence. For 85 sampling points, 85 curvature amplitude values ​​are obtained. Statistical operations are performed to calculate the mean and standard deviation of the curvature amplitude sequence. In the previous example, the mean of the curvature amplitude sequence is 0.00015, and the standard deviation is 0.00010. A weighted summation method is used to calculate the curvature threshold, with a weight of 3.0, meaning the curvature threshold equals the mean plus 3.0 times the standard deviation, resulting in 0.00045. Simultaneously, the difference between adjacent elements in the curvature amplitude sequence is calculated. When the difference exceeds 2.5 times the standard deviation, the current position is marked as a curvature amplitude abrupt change location. In the previous example, power values ​​of 12.5 MW, 27.5 MW, and 46.0 MW were identified as curvature amplitude abrupt change locations.

[0050] Iterate through all sampling points and remove points that meet the following conditions: the absolute value of the second derivative of the filter is less than the curvature threshold and the point is not located at a sudden change in curvature amplitude. In the example, 64 out of 85 sampling points were removed, and the remaining 21 points were added to the candidate segmentation point list as candidate segmentation points. The candidate segmentation points include power values ​​of 10.5 MW, 12.0 MW, 12.5 MW, 13.0 MW, 13.5 MW, 26.0 MW, 27.0 MW, 27.5 MW, 28.0 MW, 28.5 MW, 43.5 MW, 45.0 MW, 46.0 MW, 47.0 MW, and 48.5 MW.

[0051] Temporary power intervals are divided based on candidate segmentation points. Initially, adjacent candidate segmentation points serve as boundaries. Linear regression fitting is performed on each temporary power interval, and the root mean square error (RMSE) is calculated as the linearization fitting error. An error threshold of 0.002 per unit is set. For the power interval from 0 MW to 10.5 MW, the fitted linearization slope coefficient is 0.00094, and the RMSE is 0.00112, which is less than the error threshold. For the power interval from 10.5 MW to 12.0 MW, the RMSE is 0.00315, exceeding the error threshold, requiring additional segmentation points. After adding a segmentation point at a power value of 11.5 MW, the RMSEs of the two new intervals decrease to 0.00095 and 0.00102, respectively.

[0052] The linearization fitting error of adjacent temporary power intervals is checked. If both are less than a preset percentage of the error threshold (set to 50%, i.e., 0.001), the segmentation points between adjacent temporary power intervals are considered for deletion. For example, the fitting errors for the power intervals 27.5 MW to 28.0 MW and 28.0 MW to 28.5 MW are 0.00062 and 0.00075, respectively, both less than 0.001. The segmentation point at 28.0 MW is deleted, and the intervals are merged into the new interval of 27.5 MW to 28.5 MW. The fitting error of the new interval is 0.00080, which still meets the requirements. Through iterative processes of adding and deleting segmentation points, the final segmentation points are determined to be 11.5 MW, 12.5 MW, 27.5 MW, and 46.0 MW.

[0053] By combining the starting power point of 0 MW and the ending power point of 60 MW corresponding to the power-voltage response curves, a complete segmented boundary sequence is formed: 0 MW, 11.5 MW, 12.5 MW, 27.5 MW, 46.0 MW, and 60 MW. Based on this, five power ranges are obtained: 0 MW to 11.5 MW, 11.5 MW to 12.5 MW, 12.5 MW to 27.5 MW, 27.5 MW to 46.0 MW, and 46.0 MW to 60 MW. Linear regression fitting is performed on these five power ranges, and the obtained linearized slope coefficients are 0.00094, 0.00026, 0.00031, -0.00017, and -0.00076 per MW, respectively, with root mean square errors all less than the error threshold.

[0054] In this embodiment, by implementing adaptive density sampling along the power axis and combining first-order derivative and filtered second-order derivative analysis, the curvature characteristics that truly reflect changes in operating characteristics in the power-voltage response curve can be highlighted while suppressing the influence of measurement noise. This effectively reduces the risk of noise amplification and the introduction of false inflection points, and improves the stability and reliability of segmentation determination. By statistically modeling the curvature amplitude sequence and adaptively generating curvature thresholds, and combining the identification rules of curvature abrupt change locations, candidate segmentation points with physical significance can be automatically screened according to the differences in curve shape under different new energy units and different operating conditions. This avoids the problem of too many segments or missing segments caused by fixed thresholds or empirical settings, significantly enhancing the versatility and adaptability of the method. By introducing an iterative mechanism for supplementing and deleting segmentation points based on linearization fitting error, the power interval segments can accurately approximate the original power-voltage response curve while avoiding unnecessary redundant segments, significantly improving the overall consistency and linearization effect of the segmentation results.

[0055] In one alternative implementation, The steady-state weak node with the closest electrical distance to the new energy power generation unit is taken as the fault application point. A three-phase short-circuit disturbance is set at the fault application point, and the transient response of the power control loop and current control loop in the new energy power generation unit is solved to obtain the transient trajectory corresponding to the port of the new energy power generation unit, including: Extract the node number of the steady-state weak node, calculate the electrical distance between the new energy power generation unit access node and each steady-state weak node based on the simplified topology connection relationship, and take the steady-state weak node with the smallest electrical distance as the fault application point; A three-phase short-circuit disturbance is set at the fault application point and the three-phase voltage amplitude is set to zero. Based on the node number, the corresponding row and column elements are located in the pre-obtained modified node admittance matrix and the self-admittance element corresponding to the fault application point is modified to the fault admittance value to obtain the fault state node admittance matrix. The current reference value output by the power control loop of the new energy power generation unit is used as the input data of the current control loop to establish a coupled calculation relationship. Based on the admittance matrix of the fault state node and the initial operating state corresponding to the access node of the new energy power generation unit, the node voltage transient change sequence is obtained by iteratively solving through the prediction and correction algorithm. Based on the node voltage transient change sequence and the coupling calculation relationship, the dual-loop control response time sequence is obtained. The instantaneous power time sequence and port voltage time sequence are extracted from the dual-loop control response time sequence and the trajectory is mapped in the power-voltage plane to obtain the transient trajectory corresponding to the port of the new energy power generation unit.

[0056] The node numbers of the identified steady-state weak nodes are extracted. Taking the new energy power generation unit in the aforementioned embodiment as an example, the steady-state weak node connected to node 23 is node 41, the steady-state weak node connected to node 27 is node 52, and the steady-state weak node connected to node 35 is node 64. Based on the established simplified topology connection relationship, the electrical distance between the new energy power generation unit access node and each steady-state weak node is calculated. The electrical distance is calculated using the branch impedance accumulation method, that is, the sum of the impedances of all branches traversed from the starting point to the ending point. For node 23 and its steady-state weak node 41, there are multiple possible paths. The shortest electrical distance is calculated to be 0.065 ohms using the Dijkstra shortest path algorithm; the electrical distance between node 27 and node 52 is 0.082 ohms; and the electrical distance between node 35 and node 64 is 0.047 ohms. The steady-state weak node with the smallest electrical distance is selected as the fault application point, which is node 64 in this embodiment.

[0057] A three-phase short-circuit disturbance is set at the fault application point, and the three-phase voltage amplitude is set to zero. A three-phase short circuit is a typical severe fault and can effectively test the transient response characteristics of new energy power generation units. Based on the node number, the corresponding row and column elements are located in the pre-obtained modified node admittance matrix, which has already been constructed in the aforementioned implementation steps. The self-admittance element corresponding to the fault application point is modified to the fault admittance value, set to 100,000 in Siemens, representing a very small impedance value. In the aforementioned example, the self-admittance element corresponding to node 64 is modified, i.e., the element at position (64, 64) in the modified node admittance matrix is ​​added to the original value by 100,000. This is equivalent to connecting a branch with extremely low impedance to ground at node 64, forming a three-phase short circuit state. After modification, the fault state node admittance matrix is ​​obtained, which reflects the admittance characteristics of the power grid under fault conditions.

[0058] A coupled calculation relationship is established using the current reference value output from the power control loop of the new energy power generation unit as the input data for the current control loop. The new energy power generation unit typically adopts a dual-loop control structure, with an outer loop for power control and an inner loop for current control. The power control loop generates a current reference value through a proportional-integral controller based on the deviation between the power command value and the actual measured power value. The current control loop generates a modulation wave signal through a proportional-integral controller based on the deviation between the current reference value and the actual measured current value, controlling the inverter output current. A dual-loop control model is established, with the proportional gain of the power control loop set to 0.2 and the integral time constant set to 0.05 seconds; the proportional gain of the current control loop is set to 2.0 and the integral time constant set to 0.002 seconds. Initially, it is assumed that the active power output of the new energy power generation unit at node 35 is 30 MW, the reactive power output is 0 Mvar, the port voltage is 1.02 per unit, and the phase angle is 0 degrees.

[0059] Based on the admittance matrix of the fault-state node and the initial operating state corresponding to the access node of the new energy power generation unit, the transient change sequence of the node voltage is obtained by iteratively solving the prediction and correction algorithm. The prediction and correction algorithm is a numerical integration method suitable for solving rigid differential equation systems. The time step is set to 0.0001 seconds, the total simulation time is 0.5 seconds, the fault occurrence time is 0.1 seconds, and the fault duration is 0.1 seconds.

[0060] Based on the transient change sequence of node voltage and the coupling calculation relationship of the dual-loop control, the dual-loop control response time sequence is obtained. The port voltage is calculated based on the node voltage; the power deviation is calculated based on the port voltage and the current power output; the current reference value is calculated using the power controller; the current deviation is calculated based on the current reference value and the actual current; the modulation wave signal is calculated using the current controller; and the output voltage is calculated based on the modulation wave signal and the DC-side voltage. These calculations are performed once at each time step, forming the dual-loop control response time sequence. The instantaneous power time sequence and the port voltage time sequence are extracted from the dual-loop control response time sequence. In the aforementioned example, before the fault occurred, the port voltage of the new energy power generation unit was stable at 1.02 per unit, and the active power was stable at 30 MW; 0.1 seconds after the fault occurred, the port voltage rapidly dropped to 0.25 per unit, and the active power plummeted to 7.5 MW; after the fault was cleared, the port voltage recovered to 1.00 per unit, and the active power gradually recovered to 28.5 MW, stabilizing after 0.3 seconds.

[0061] Trajectory mapping is performed in the power-voltage plane to obtain the transient trajectory corresponding to the port of the new energy power generation unit. Using instantaneous power as the abscissa and port voltage as the ordinate, the power-voltage data pairs at each time point are plotted on the plane to form a continuous trajectory. The transient trajectory can be divided into four stages: the pre-fault steady-state stage, where trajectory points cluster around a power of 30 MW and a voltage of 1.02 per unit; the fault occurrence stage, where the trajectory rapidly moves from the initial point to around a power of 7.5 MW and a voltage of 0.25 per unit; the fault clearance stage, where the trajectory rapidly recovers from a low power and low voltage point to a higher power and voltage point; and the stable recovery stage, where the trajectory gradually converges to around a power of 28.5 MW and a voltage of 1.00 per unit.

[0062] In this embodiment, by calculating the electrical distance between the access node of the new energy power generation unit and each steady-state weak node based on the simplified topology connection relationship, and selecting the steady-state weak node with the smallest electrical distance as the fault application point, it is possible to ensure that the applied disturbance has the most significant impact on the new energy unit in terms of electrical coupling, which significantly enhances the representativeness and sensitivity of the fault scenario and improves the coverage of potential weak operating conditions by transient analysis. By directly introducing three-phase short-circuit fault admittance modeling at the node admittance matrix level and dynamically updating it in combination with the corrected node admittance matrix, it is possible to accurately reflect the changes in electrical constraints under fault conditions while maintaining the consistency of the network structure, which significantly improves the accuracy and numerical stability of node voltage calculation during faults. By establishing the coupling calculation relationship between the power control loop and the current control loop of the new energy power generation unit and combining the prediction and correction algorithm to iteratively solve the node voltage under fault conditions, it is possible to finely characterize the dynamic response process of the new energy power generation unit under strong disturbance conditions, effectively reduce the model fragmentation error, and improve the authenticity and credibility of transient response analysis.

[0063] In one alternative implementation, Based on the fault state node admittance matrix and the initial operating state corresponding to the access node of the new energy power generation unit, the node voltage transient change sequence is obtained by iteratively solving the prediction and correction algorithm, including: Extract the initial value of the node voltage in the initial operating state corresponding to the access node of the new energy power generation unit, establish the node voltage differential equation based on the admittance matrix of the fault state node, set the integration time step and initialize the current time as the fault occurrence time; Within each integration time step, the forward Euler method is used to calculate the predicted value of the node voltage at the current time based on the node voltage value and the differential value of the node voltage at the previous time step. The backward Euler method is used to recalculate the differential value of the node voltage at the current moment based on the predicted value of the node voltage and the admittance matrix of the fault state node, and the corrected value of the node voltage at the current moment is obtained. It is determined whether the difference between the corrected value of the node voltage and the predicted value of the node voltage is less than a preset convergence threshold. If it is less than the threshold, the corrected value of the node voltage is used as the determined value of the node voltage at the current moment and recorded in the node voltage transient change sequence. The current moment is updated to the moment corresponding to the next integration time step, and the prediction and steps are repeated. If the value is not less than the target value, the corrected value of the node voltage is used as the new predicted value and the correction is repeated until the convergence condition is met or the simulation ends. The transient change sequence of the node voltage is then output.

[0064] The initial node voltage values ​​are extracted from the initial operating state of the new energy power generation unit's access node. Taking a regional power grid as an example, this grid contains 120 nodes, with node 35 being the access node for the new energy power generation unit. In the initial operating state, the voltage amplitude of node 35 is 1.02 per unit, and the phase angle is 0.05 radians. This initial value will be used as the starting value for subsequent transient simulations. Based on the fault state node admittance matrix obtained above, a node voltage differential equation is established. The establishment of the node voltage differential equation is based on the dynamic characteristics of the power grid during transient processes, considering network impedance, load characteristics, and the dynamic response characteristics of the new energy power generation unit. This differential equation can be expressed as: the time derivative of the node voltage is equal to the product of the inverse of the node admittance matrix and the difference between the node injected current. In this regional power grid, the dynamic response characteristics of the new energy power generation unit are reflected through the corresponding control system model, including power control and current control components.

[0065] The integration time step is set to 0.0001 seconds, which meets the accuracy requirements for calculating high-frequency electromagnetic transients and the dynamic response of the control system during power system transients. The fault occurrence time is set to 0.1 seconds, and the current time is initialized as the fault occurrence time, i.e., 0.1 seconds. At this time, the fault has been applied to the steady-state weak node 64, and the admittance matrix of the fault state node has been updated, reflecting the impact of the fault on the power grid topology. Based on the initial operating state, the initial value of the node voltage is set to the steady-state value before the fault occurs, with the voltage of node 35 at 1.02 per unit and the phase angle at 0.05 radians. The differential value of the node voltage is initialized by substituting the initial node voltage into the differential equation and solving it. In this embodiment, the differential amplitude of the voltage of node 35 at the initial time is -2.5 per unit per second, and the differential phase angle is -15 radians per second, indicating that the node voltage will decrease rapidly and the phase angle will change rapidly after the fault occurs.

[0066] Within each integration time step, the forward Euler method is used to calculate the predicted node voltage at the current time based on the node voltage value and its derivative from the previous time step. The forward Euler method is a first-order explicit integration method; the predicted node voltage at the current time step equals the node voltage value at the previous time step plus its derivative multiplied by the time step. Taking the first integration step as an example, the current time step is 0.1001 seconds, the previous time step is 0.1000 seconds, the voltage of node 35 at the previous time step is 1.02 per unit, and the voltage derivative is -2.5 per unit per second. The calculated predicted node voltage at the current time step is 0.9975 per unit, and the predicted phase angle is 0.0485 radians. While the forward Euler method is simple and fast, it may face numerical stability issues when dealing with rigid differential equations; therefore, it needs to be combined with the backward Euler method for correction.

[0067] The backward Euler method is used to recalculate the differential value of the node voltage at the current moment based on the predicted node voltage and the node admittance matrix under fault conditions. The backward Euler method is an implicit integration method, requiring the solution of a system of nonlinear equations. Substituting the predicted node voltage at the current moment into the differential equations yields the new differential value. For node 35, when the predicted voltage at the current moment is 0.9975 per unit, the calculated differential voltage value is -3.2 per unit per second. Based on the updated differential value, the backward Euler method is used to calculate the correction value of the node voltage at the current moment. The calculation process of the backward Euler method is: the corrected node voltage value at the current moment equals the node voltage value at the previous moment plus the differential value of the node voltage at the current moment multiplied by the time step. The calculated corrected voltage value for node 35 at the current moment is 0.9968 per unit, and the phase angle correction value is 0.0482 radians.

[0068] The algorithm determines whether the difference between the corrected node voltage and the predicted node voltage is less than a preset convergence threshold. The preset convergence threshold is set to 0.0001 per unit, and it needs to be sufficiently small to ensure calculation accuracy. For node 35, the absolute value of the difference between the corrected voltage value (0.9968 per unit) and the predicted value (0.9975 per unit) is 0.0007 per unit, which is greater than the convergence threshold of 0.0001 per unit. Therefore, further iterative calculation is required. The corrected voltage value of 0.9968 per unit is used as the new predicted value, and the Euler method calculation step is re-executed. After iterative calculation, a new voltage differential value of -3.1 per unit per second and a new voltage corrected value of 0.9969 per unit are obtained. The absolute value of the difference between the corrected value and the predicted value is 0.0001 per unit, which is equal to the convergence threshold, satisfying the convergence condition.

[0069] When the difference between the corrected value and the predicted value of the node voltage is less than the preset convergence threshold, the corrected value of the node voltage is used as the determined value of the node voltage at the current moment and recorded in the node voltage transient change sequence. At time 0.1001, the determined value of the voltage of node 35 is 0.9969 per unit, and the determined value of the phase angle is 0.0482 radians. The current time is updated to the time corresponding to the next integration time step, i.e., 0.1002 seconds, and the prediction and correction steps are repeated. The same calculation process is performed at each time step until the simulation end time of 0.5 seconds is reached. For large-scale power grids, there may be computational convergence issues. To ensure the stability of the simulation process, the maximum number of iterations is set to 10. If convergence is not achieved after reaching the upper limit of the number of iterations, the relaxation factor method is used to assist convergence, and the relaxation factor is set to 0.8.

[0070] During the fault duration (0.1 to 0.2 seconds), the voltage at the fault point approached zero, and the voltages at other nodes also decreased significantly. The voltage at node 35 dropped to a minimum of 0.25 per unit during the fault, with a phase shift of 0.32 radians. After the fault was cleared (0.2 seconds later), the system began to recover. During the recovery process, the voltage at node 35 oscillated, with a maximum overshoot of 0.08 per unit and an oscillation frequency of approximately 5 Hz. After approximately 0.15 seconds of oscillation, the system stabilized, and the voltage at node 35 eventually stabilized at 1.00 per unit, with a phase angle of 0.07 radians. The entire simulation recorded node voltage data at 4000 time points, forming a complete sequence of transient node voltage changes.

[0071] In this embodiment, by starting with the initial node voltage value at the time of fault occurrence and combining it with the node admittance matrix of the fault state to establish the node voltage differential equation, the dynamic evolution process of the node voltage can be continuously described under the condition of abrupt changes in electrical parameters. This significantly improves the physical continuity and time resolution of the transient voltage change characterization. By first using the forward Euler method for voltage prediction and then using the backward Euler method for voltage correction within each integration time step, the numerical divergence problem that is prone to occur in explicit methods in strongly nonlinear and rigid systems can be effectively suppressed while ensuring computational step efficiency. This significantly enhances numerical stability and reduces computational complexity and iteration burden. By introducing an adaptive convergence determination mechanism based on the difference between the predicted and corrected values, unnecessary repeated iterations can be automatically terminated while ensuring solution accuracy. This allows the node voltage to converge quickly to a reasonable solution within each time step, effectively improving simulation efficiency and avoiding error accumulation.

[0072] Figure 2 This is a flowchart illustrating the voltage transient change solution of the power grid steady-state and transient characteristic evaluation method under large-scale new energy grid connection in an embodiment of the present invention.

[0073] In one alternative implementation, The transient trajectory is subjected to Hilbert transform to obtain the instantaneous amplitude envelope and instantaneous frequency curve. The maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve are calculated. The transient stability margin is determined based on the maximum descent depth and the attenuation coefficient, and the evaluation results are generated, including: The coordinate point sequence of the transient trajectory in the power-voltage plane is extracted as a time series signal. The time series signal is subjected to Hilbert transform to obtain an analytical signal. The real and imaginary parts are extracted from the analytical signal and the instantaneous amplitude envelope and instantaneous phase curve are calculated. The instantaneous frequency curve is obtained by time differentiation of the instantaneous phase curve. Identify the minimum amplitude point from the instantaneous amplitude envelope and calculate the amplitude difference between the minimum amplitude point and the pre-acquired initial steady-state operating point as the maximum descent depth; Spectral analysis is performed on the instantaneous frequency curve to extract the frequency component with the largest amplitude as the dominant oscillation frequency component. The oscillation waveform corresponding to the dominant oscillation frequency component is extracted from the instantaneous frequency curve, and the envelope of the oscillation waveform is fitted with an exponential function to obtain the attenuation coefficient. The maximum descent depth is compared with a preset depth threshold to obtain a depth margin index, the attenuation coefficient is compared with a preset attenuation threshold to obtain an attenuation margin index, and the depth margin index and the attenuation margin index are weighted and summed to obtain a transient stability margin. The transient stability margin is associated with the steady-state weak node to obtain the evaluation result and output it.

[0074] The coordinate sequence of the transient trajectory in the power-voltage plane is extracted as a time series signal. Taking a new energy power plant with an installed capacity of 50 MW connected to node 35 as an example, the transient trajectory under fault conditions was obtained. This trajectory contains data at 5000 time points, with a time span of 0 to 0.5 seconds and a sampling interval of 0.0001 seconds. Power time series and voltage time series are extracted from the power-voltage plane. The initial value of the power time series is 30 MW, the minimum value is 7.5 MW, and the final stable value is 28.5 MW; the initial value of the voltage time series is 1.02 per unit, the minimum value is 0.25 per unit, and the final stable value is 1.00 per unit. To analyze the frequency characteristics and stability of the system, the power time series is used as the main analysis object, as the signal can better reflect the response characteristics of the new energy power generation unit to system disturbances.

[0075] An analytic signal is obtained by applying a Hilbert transform to the power time series signal. The Hilbert transform is a signal processing technique that converts real signals into analytic signals, facilitating the extraction of instantaneous characteristics. The Hilbert transform is implemented using the Fast Fourier Transform (FFT) method. The original signal undergoes a Fourier transform; negative frequency components are set to zero, and positive frequency components are multiplied by 2; an inverse Fourier transform is then performed to obtain the analytic signal. In this embodiment, the Hilbert transform is applied to the power time series to obtain a complex analytic signal. The real and imaginary parts are extracted from the analytic signal; the real part is the original power signal, and the imaginary part is the result of the Hilbert transform of the original signal. The instantaneous amplitude envelope of the analytic signal is calculated, which is the square root of the sum of the squares of the real and imaginary parts. In this embodiment, the initial value of the instantaneous amplitude envelope is 30 MW, drops sharply at 0.1 seconds after the fault occurs, reaches its lowest point at 0.15 seconds (6.2 MW), gradually recovers after the fault is cleared with oscillations, and finally stabilizes around 28.5 MW.

[0076] The instantaneous phase curve of the analytic signal is calculated, which is the arctangent of the ratio of the imaginary part to the real part. During the calculation, a four-quadrant arctangent function is used to ensure phase continuity. The initial value of the instantaneous phase curve is close to 0 radians, monotonically increasing with time, but the rate of increase varies with the system state. The instantaneous frequency curve is obtained by differentiating the instantaneous phase curve over time, representing the change in the local oscillation frequency of the signal. The differentiation uses the center difference method, where the current frequency equals the phase difference between two adjacent points divided by the time interval. To reduce noise introduced by the numerical differentiation, the obtained instantaneous frequency curve is smoothed using a Butterworth low-pass filter with a cutoff frequency set to 50 Hz. The processed instantaneous frequency curve is close to 0 Hz during steady-state operation, but exhibits significant fluctuations after a fault occurs, with a maximum frequency shift reaching 12 Hz. After the fault is cleared, it gradually stabilizes at an oscillation frequency of 2 to 3 Hz.

[0077] The minimum amplitude point is identified from the instantaneous amplitude envelope, and the amplitude difference between this point and the initial steady-state operating point is calculated as the maximum descent depth. By searching for the global minimum on the instantaneous amplitude envelope, the minimum amplitude point is identified at 0.15 seconds, with a value of 6.2 MW. The power value at the initial steady-state operating point is 30 MW, therefore the maximum descent depth is 23.8 MW, accounting for 79.3% of the initial value. The maximum descent depth reflects the vulnerability of the system after disturbance; the greater the descent depth, the weaker the system's disturbance resistance. Spectral analysis is performed on the instantaneous frequency curve, and the frequency component with the largest amplitude is extracted as the dominant oscillation frequency component. The spectral analysis uses the Fast Fourier Transform method to transform the instantaneous frequency curve (0.2 seconds to 0.5 seconds) after fault clearing. In the obtained spectrum, the amplitude is largest at 2.5 Hz, therefore 2.5 Hz is identified as the dominant oscillation frequency component.

[0078] The oscillation waveform corresponding to the dominant oscillation frequency component is extracted from the instantaneous frequency curve. This is achieved by processing the original instantaneous frequency curve through a bandpass filter with a center frequency of 2.5 Hz and a bandwidth of 0.5 Hz. The filtered signal mainly contains frequency components around 2.5 Hz, more clearly reflecting the characteristics of the dominant oscillation mode. The amplitude envelope of the filtered oscillation waveform is calculated using the method of connecting the signal's local extrema. An exponential function is then fitted to the envelope, with the fitting function being that the amplitude equals the initial amplitude multiplied by a negative power of the exponential decay factor, where the exponential decay factor is the product of time and the decay coefficient. The fitting parameters are determined using the least squares method, yielding a decay coefficient of 5.2, representing the oscillation decay rate. A larger decay coefficient indicates better system damping characteristics and a faster recovery to a steady state.

[0079] The depth margin index is obtained by comparing the maximum descent depth with a preset depth threshold. The preset depth threshold is set to 25 MW, representing the maximum acceptable power descent of the system. In this embodiment, the maximum descent depth is 23.8 MW, which is less than the preset threshold of 25 MW. The depth margin index is calculated as 1 minus the ratio of the maximum descent depth to the preset depth threshold, resulting in 0.048. A depth margin index greater than 0 indicates a certain margin in terms of power descent depth; a larger value indicates a more sufficient margin. The attenuation margin index is obtained by comparing the attenuation coefficient with a preset attenuation threshold. The preset attenuation threshold is set to 3.0, representing the minimum attenuation rate required by the system. In this embodiment, the attenuation coefficient is 5.2, which is greater than the preset threshold of 3.0. The attenuation margin index is calculated as the ratio of the attenuation coefficient to the preset attenuation threshold minus 1, resulting in 0.733. An attenuation margin index greater than 0 indicates that the system has sufficient damping characteristics to effectively suppress oscillations; a larger value indicates better damping characteristics.

[0080] The transient stability margin is obtained by weighted summation of the depth margin and attenuation margin indices. The weight of the depth margin index is set to 0.4, and the weight of the attenuation margin index is set to 0.6. This weighting considers that in large-scale renewable energy grid-connected systems, damping characteristics are generally more critical than the power drop depth. The calculated transient stability margin is 0.458. A transient stability margin greater than 0 indicates that the system has sufficient stability under current operating conditions and disturbances; a larger value indicates better system stability and greater resilience to similar disturbances. The transient stability margin is correlated with the steady-state weak node to obtain the evaluation results. In this example, the steady-state weak node is node 64, with a transient stability margin of 0.458. The combined results show that the steady-state weak node 64 exhibits some stability under three-phase short-circuit fault conditions, but the margin is not large. It is recommended to enhance the reactive power compensation capability near this node or adjust the control parameters of the renewable energy generation unit to improve the system's transient stability.

[0081] In this embodiment, by performing a Hilbert transform on the power-voltage plane transient trajectory and constructing an analytical signal, the instantaneous amplitude, phase, and frequency evolution information of the trajectory can be extracted simultaneously under non-stationary disturbance conditions, improving the ability to characterize complex transient behaviors. By extracting the maximum descent depth from the instantaneous amplitude envelope and comparing it with the initial steady-state operating point, it is beneficial to evaluate the system's ability to withstand strong disturbances and its safety margin. By performing spectral analysis on the instantaneous frequency curve and identifying the dominant oscillation frequency component, and by obtaining the attenuation coefficient by combining the exponential fitting of the oscillation envelope, the damping characteristics of transient oscillations can be accurately characterized, directly reflecting the decay rate of the system's oscillation energy, and improving the physical consistency and comparability of transient stability determination.

[0082] A second aspect of this invention provides a system for evaluating the steady-state and transient characteristics of a power grid under large-scale renewable energy grid integration, comprising: The topology simplification module is used to obtain the topology connection relationship of the power grid system, the access node of the new energy power generation unit and the corresponding output time series, identify the non-critical nodes in the topology connection relationship and perform impedance equivalent transformation to obtain the simplified topology connection relationship. The weak point identification module is used to extract the power-voltage response curve corresponding to the output time series and perform piecewise linearization to obtain the piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, the minimum singular value and the corresponding singular vector are calculated by singular value decomposition. For each new energy power generation unit access node, the node corresponding to the element with the largest amplitude in the singular vector is extracted as the steady-state weak node. The transient solution module is used to take the steady-state weak node with the closest electrical distance to the access node of the new energy power generation unit as the fault application point, set a three-phase short-circuit disturbance at the fault application point, and solve the transient response of the power control loop and the current control loop in the new energy power generation unit to obtain the transient trajectory corresponding to the port of the new energy power generation unit. The margin assessment module is used to perform Hilbert transform on the transient trajectory to obtain the instantaneous amplitude envelope and instantaneous frequency curve, calculate the maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve, determine the transient stability margin based on the maximum descent depth and the attenuation coefficient, and generate an assessment result.

[0083] A third aspect of the present invention provides an electronic device, comprising: A processor and a memory for storing processor-executable instructions, wherein the processor is configured to invoke instructions stored in the memory to perform the aforementioned method.

[0084] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0085] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating the steady-state and transient characteristics of a power grid under large-scale renewable energy grid integration, characterized in that, include: The topology of the power grid system, the access nodes of new energy power generation units and their corresponding output time series are obtained. Non-critical nodes in the topology are identified and impedance equivalent transformation is performed to obtain a simplified topology. Extract the power-voltage response curve corresponding to the output time series and perform piecewise linearization to obtain the piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, calculate the minimum singular value and the corresponding singular vector through singular value decomposition. For each new energy power generation unit access node, extract the node corresponding to the element with the largest amplitude in the singular vector as the steady-state weak node. The steady-state weak node with the closest electrical distance to the access node of the new energy power generation unit is taken as the fault application point. A three-phase short-circuit disturbance is set at the fault application point, and the transient response of the power control loop and the current control loop in the new energy power generation unit is solved to obtain the transient trajectory corresponding to the port of the new energy power generation unit. The transient trajectory is subjected to Hilbert transform to obtain the instantaneous amplitude envelope and instantaneous frequency curve. The maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve are calculated. The transient stability margin is determined based on the maximum descent depth and the attenuation coefficient, and an evaluation result is generated.

2. The method according to claim 1, characterized in that, Identifying non-critical nodes in the topological connections and performing impedance equivalent transformations to obtain simplified topological connections includes: Based on the topological connection relationship, a node association matrix is ​​constructed. The set of shortest electrical paths from the access node of the new energy power generation unit to the system balance node is extracted. The frequency of occurrence of each access node of the new energy power generation unit in the set of shortest electrical paths is counted to obtain the path frequency statistics. After performing the node deletion operation, the algebraic connectivity change rate of the node association matrix is ​​calculated. Non-critical nodes are selected based on the path frequency statistics and the algebraic connectivity change rate. The non-critical nodes are sorted by degree, and the set of adjacent nodes and branch impedance values ​​of each non-critical node are extracted in turn. The equivalent impedance value is obtained by selecting the impedance equivalent calculation method according to the number of nodes in the set of adjacent nodes. When there are two nodes, the series equivalent calculation method is used. When there are three nodes, the star-triangle equivalent transformation calculation method is used. When there are more than three nodes, the admittance matrix block elimination calculation method is used. The equivalent impedance value is updated to the node correlation matrix and the non-critical nodes are deleted. After processing, the simplified topology connection relationship is formed.

3. The method according to claim 1, characterized in that, The power-voltage response curve corresponding to the output time series is extracted and piecewise linearized to obtain a piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, the minimum singular value and the corresponding singular vector are calculated through singular value decomposition. For each new energy power generation unit access node, the node corresponding to the element with the largest amplitude in the singular vector is extracted as a steady-state weak node, including: The power output value and the corresponding voltage measurement value are extracted from the output time series and curve fitting is performed to obtain the power-voltage response curve. Based on the slope change characteristics of the power-voltage response curve in different power ranges, multiple power range segments are obtained by identifying segmentation points through an adaptive curvature detection algorithm. Linear regression fitting is performed on the power range segments to obtain linearized slope coefficients, and the piecewise admittance value corresponding to the power range segment is determined according to the linearized slope coefficients to obtain the piecewise admittance function. Based on the simplified topology connection relationship, a node admittance matrix is ​​constructed, and the self-admittance element corresponding to the access node of the new energy power generation unit is replaced with the piecewise admittance value corresponding to the current operating power in the piecewise admittance function to obtain the modified node admittance matrix. Perform singular value decomposition on the modified node admittance matrix to obtain a singular value sequence and a right singular vector matrix. Extract the singular value with the smallest value in the singular value sequence as the minimum singular value. Extract the singular vector in the right singular vector matrix that corresponds to the minimum singular value. Calculate the magnitude of each element in the singular vector and establish a mapping relationship with the node number in the simplified topology connection relationship. For each new energy power generation unit access node, after excluding itself from the mapping relationship, the node corresponding to the element with the largest amplitude among the remaining elements is extracted as the steady-state weak node.

4. The method according to claim 3, characterized in that, Based on the slope variation characteristics of the power-voltage response curve in different power ranges, multiple power range segments are obtained by identifying segmentation points using an adaptive curvature detection algorithm, including: The power-voltage response curve is subjected to adaptive density sampling along the power axis to obtain multiple sampling points. For each sampling point, adjacent sampling points are selected and the first derivative value is calculated. The first derivative value is subjected to second difference and a filter is applied to remove high-frequency noise to obtain the filtered second derivative value. The absolute value of the second derivative of the filter is extracted to form a curvature amplitude sequence. Statistical operations are performed on the curvature amplitude sequence to obtain the mean and standard deviation, and a weighted sum is obtained to obtain the curvature threshold. The difference between adjacent elements of the curvature amplitude sequence is calculated and the abrupt change position is identified as the curvature amplitude abrupt change position. Traverse the sampling points and remove sampling points whose absolute value of the second derivative of the filter is less than the curvature threshold and are not located at the position of curvature amplitude change. Add the retained sampling points as candidate segmentation points to the candidate segmentation point list. Temporary power intervals are divided based on candidate segmentation points, and linear regression fitting is performed to obtain linearized fitting error. If the linearized fitting error exceeds a preset error threshold, segmentation points are added. If the linearized fitting error of adjacent temporary power intervals is less than a preset proportion of the error threshold, segmentation points are deleted. The process of adding and deleting is repeated to obtain the final set of segmentation points. The starting power point and ending power point corresponding to the power-voltage response curve are combined to form a segmentation boundary sequence and divide multiple power intervals.

5. The method according to claim 1, characterized in that, The steady-state weak node with the closest electrical distance to the new energy power generation unit is taken as the fault application point. A three-phase short-circuit disturbance is set at the fault application point, and the transient response of the power control loop and current control loop in the new energy power generation unit is solved to obtain the transient trajectory corresponding to the port of the new energy power generation unit, including: Extract the node number of the steady-state weak node, calculate the electrical distance between the new energy power generation unit access node and each steady-state weak node based on the simplified topology connection relationship, and take the steady-state weak node with the smallest electrical distance as the fault application point; A three-phase short-circuit disturbance is set at the fault application point and the three-phase voltage amplitude is set to zero. Based on the node number, the corresponding row and column elements are located in the pre-obtained modified node admittance matrix and the self-admittance element corresponding to the fault application point is modified to the fault admittance value to obtain the fault state node admittance matrix. The current reference value output by the power control loop of the new energy power generation unit is used as the input data of the current control loop to establish a coupled calculation relationship. Based on the admittance matrix of the fault state node and the initial operating state corresponding to the access node of the new energy power generation unit, the node voltage transient change sequence is obtained by iteratively solving through the prediction and correction algorithm. Based on the node voltage transient change sequence and the coupling calculation relationship, the dual-loop control response time sequence is obtained. The instantaneous power time sequence and port voltage time sequence are extracted from the dual-loop control response time sequence and the trajectory is mapped in the power-voltage plane to obtain the transient trajectory corresponding to the port of the new energy power generation unit.

6. The method according to claim 5, characterized in that, Based on the fault state node admittance matrix and the initial operating state corresponding to the access node of the new energy power generation unit, the node voltage transient change sequence is obtained by iteratively solving the prediction and correction algorithm, including: Extract the initial value of the node voltage in the initial operating state corresponding to the access node of the new energy power generation unit, establish the node voltage differential equation based on the admittance matrix of the fault state node, set the integration time step and initialize the current time as the fault occurrence time; Within each integration time step, the forward Euler method is used to calculate the predicted value of the node voltage at the current time based on the node voltage value and the differential value of the node voltage at the previous time step. The backward Euler method is used to recalculate the differential value of the node voltage at the current moment based on the predicted value of the node voltage and the admittance matrix of the fault state node, and the corrected value of the node voltage at the current moment is obtained. It is determined whether the difference between the corrected value of the node voltage and the predicted value of the node voltage is less than a preset convergence threshold. If it is less than the threshold, the corrected value of the node voltage is used as the determined value of the node voltage at the current moment and recorded in the node voltage transient change sequence. The current moment is updated to the moment corresponding to the next integration time step, and the prediction and steps are repeated. If the value is not less than the target value, the corrected value of the node voltage is used as the new predicted value and the correction is repeated until the convergence condition is met or the simulation ends. The transient change sequence of the node voltage is then output.

7. The method according to claim 1, characterized in that, The transient trajectory is subjected to Hilbert transform to obtain the instantaneous amplitude envelope and instantaneous frequency curve. The maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve are calculated. The transient stability margin is determined based on the maximum descent depth and the attenuation coefficient, and the evaluation results are generated, including: The coordinate point sequence of the transient trajectory in the power-voltage plane is extracted as a time series signal. The time series signal is subjected to Hilbert transform to obtain an analytical signal. The real and imaginary parts are extracted from the analytical signal and the instantaneous amplitude envelope and instantaneous phase curve are calculated. The instantaneous frequency curve is obtained by time differentiation of the instantaneous phase curve. Identify the minimum amplitude point from the instantaneous amplitude envelope and calculate the amplitude difference between the minimum amplitude point and the pre-acquired initial steady-state operating point as the maximum descent depth; Spectral analysis is performed on the instantaneous frequency curve to extract the frequency component with the largest amplitude as the dominant oscillation frequency component. The oscillation waveform corresponding to the dominant oscillation frequency component is extracted from the instantaneous frequency curve, and the envelope of the oscillation waveform is fitted with an exponential function to obtain the attenuation coefficient. The maximum descent depth is compared with a preset depth threshold to obtain a depth margin index, the attenuation coefficient is compared with a preset attenuation threshold to obtain an attenuation margin index, and the depth margin index and the attenuation margin index are weighted and summed to obtain a transient stability margin. The transient stability margin is associated with the steady-state weak node to obtain the evaluation result and output it.

8. A system for evaluating the steady-state and transient characteristics of a power grid under large-scale renewable energy grid integration, used to implement the method described in any one of claims 1-7, characterized in that, include: The topology simplification module is used to obtain the topology connection relationship of the power grid system, the access node of the new energy power generation unit and the corresponding output time series, identify the non-critical nodes in the topology connection relationship and perform impedance equivalent transformation to obtain the simplified topology connection relationship. The weak point identification module is used to extract the power-voltage response curve corresponding to the output time series and perform piecewise linearization to obtain the piecewise admittance function. Based on the node admittance matrix of the piecewise admittance function and the simplified topology connection relationship, the minimum singular value and the corresponding singular vector are calculated by singular value decomposition. For each new energy power generation unit access node, the node corresponding to the element with the largest amplitude in the singular vector is extracted as the steady-state weak node. The transient solution module is used to take the steady-state weak node with the closest electrical distance to the access node of the new energy power generation unit as the fault application point, set a three-phase short-circuit disturbance at the fault application point, and solve the transient response of the power control loop and the current control loop in the new energy power generation unit to obtain the transient trajectory corresponding to the port of the new energy power generation unit. The margin assessment module is used to perform Hilbert transform on the transient trajectory to obtain the instantaneous amplitude envelope and instantaneous frequency curve, calculate the maximum descent depth of the instantaneous amplitude envelope and the attenuation coefficient of the dominant oscillation frequency component of the instantaneous frequency curve, determine the transient stability margin based on the maximum descent depth and the attenuation coefficient, and generate an assessment result.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images