A method for rapid identification and configuration selection of transient voltage constraints of synchronous condensers
By combining sensitivity proxy models and integral indices, along with dynamic time windows and piecewise curve parameterization methods, the problems of high computational cost and insufficient accuracy in the selection of synchronous condenser configurations in high-proportion renewable energy grid-connected transmission systems are solved. This enables fast and accurate voltage constraint discrimination and configuration selection, ensuring the safety and stability of the power grid.
Patent Information
- Application Number
- CN202611471653.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies in high-proportion renewable energy grid-connected transmission systems suffer from high computational costs, insufficient evaluation accuracy, and low screening efficiency, making it difficult to quickly identify and screen the transient voltage constraint configuration of synchronous condensers, thus threatening the safety and stability of the power grid.
A sensitivity proxy model is used to replace full-time domain simulation. Pre-screening is performed by combining integral index and threshold criterion. The time window is dynamically determined. The reactive power output capability of the synchronous condenser is described by piecewise curve parameterization. The reactive power sensitivity of the voltage is quickly evaluated by using equivalent linearized voltage. Key parameters are calculated offline and looked up online to ensure the safety and efficiency of the screening results.
It significantly reduces the computational burden in large-scale planning scenarios, accurately reflects the safety margin of transient voltage, ensures the safety and accuracy of screening results, reduces the risk of misjudgment and omission, and improves computational efficiency and engineering feasibility.
Smart Images

Figure CN123121706A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of power system planning technology, and in particular to a method for rapid identification and configuration screening of transient voltage constraints of synchronous condensers. Background Technology
[0002] With the rapid development of new energy power generation, high-proportion new energy grid-connected transmission systems have become an important form of power system. In new energy collection and transmission systems, because new energy power plants are usually located at the end of the grid, the system exhibits significant weak grid characteristics, specifically manifested in low equivalent short-circuit ratio, insufficient inertia support capacity, and limited reactive power and voltage regulation capabilities. In such systems, when typical disturbances such as AC line short-circuit faults and DC commutation failures occur, the transient response of the bus voltage is prone to deterioration, resulting in problems such as large voltage drops, slow recovery speeds, and even voltage instability, seriously threatening the safe and stable operation of the system.
[0003] To improve the transient voltage characteristics of renewable energy transmission systems, configuring synchronous condensers is an effective technical means. Synchronous condensers have advantages such as fast dynamic reactive power response, strong overload capacity, and high operational reliability. They can quickly provide reactive power support during faults, improving the transient recovery capability of the bus voltage. In actual engineering planning, it is necessary to optimize among multiple candidate bus connection locations, different configuration numbers, and capacity levels to form a large-scale candidate configuration set, and then select feasible solutions that meet transient voltage safety constraints.
[0004] However, existing technologies have significant shortcomings in addressing the aforementioned screening requirements. Traditional transient voltage safety assessment methods primarily rely on full-time-domain electromagnetic transient simulation. While this can accurately simulate the voltage response process after a fault disturbance, its computational cost is exorbitant. A single simulation for a single configuration scheme can take tens of minutes or even hours. When facing large-scale planning and screening scenarios involving hundreds or thousands of candidate configurations, performing full-time-domain simulation for each scheme will lead to a dramatic increase in computation time, which is impractical for engineering applications. Furthermore, existing simplified assessment methods often employ fixed time windows and single-point criteria, making it difficult to accurately reflect changes in safety margin throughout the transient voltage recovery process. This is particularly problematic in weak grid scenarios, where it can easily lead to missed or incorrect assessments.
[0005] Therefore, there is an urgent need for a method that can quickly identify and screen large-scale candidate configurations under transient voltage constraints, in order to overcome the technical problems of high computational cost, insufficient evaluation accuracy, and low screening efficiency in the existing technology. Summary of the Invention
[0006] To address this, the present invention provides a method for rapid identification and configuration screening of transient voltage constraints of synchronous condensers, in order to overcome the technical problems of high computational cost, insufficient evaluation accuracy, and low screening efficiency in the prior art.
[0007] To achieve the above objectives, this invention provides a method for rapid identification and configuration selection of transient voltage constraints in synchronous condensers. It includes: Step S1: Obtain the candidate synchronous condenser configuration set, operating condition set, transient voltage safety criteria, synchronous condenser overload reactive power dynamic capability parameters, and new energy transmission system model parameters. Step S2: For each candidate configuration and operating condition, set a judgment time window after the fault disturbance. Based on the parameters of the new energy transmission system model, calculate the equivalent reactive power support required to make the key bus voltage meet the transient voltage safety criterion at each moment within the judgment time window. Based on the dynamic overload reactive power capability parameters of the synchronous condenser, calculate the upper limit of the available overload reactive power dynamic output that the synchronous condenser in the candidate configuration can provide at each moment within the judgment time window. The end point of the judgment time window is dynamically determined according to the actual moment when the key bus voltage under the candidate configuration recovers to meet the transient voltage safety criterion and remains stable. Step S3: At the same time, compare the available overload reactive dynamic output upper limit with the equivalent reactive support amount hourly. Take the positive value of the equivalent reactive support amount exceeding the available overload reactive dynamic output upper limit as the reactive margin deficiency at that time. Integrate the reactive margin deficiency within the discrimination time window to obtain the rapid discrimination index of the candidate configuration under the corresponding operating condition. Step S4: Aggregate the rapid discrimination indicators under each working condition to determine the comprehensive rapid risk indicator of the candidate configuration; Step S5: Compare the comprehensive rapid risk index with a preset threshold. If the comprehensive rapid risk index is less than or equal to the preset threshold, the candidate configuration is determined to have passed the rapid screening under transient voltage safety constraints and enter the fine verification stage. If the comprehensive rapid risk index is greater than the preset threshold, the candidate configuration is determined to not meet the transient voltage safety constraints and is eliminated.
[0008] Furthermore, in step S2, the overload reactive power dynamic capability parameters of the synchronous condenser are generated using a piecewise curve parameterization method, specifically including: The upper limit of available reactive power for a single synchronous condenser under preset operating conditions is determined by a piecewise function, which includes at least three stages: an initial stage, an overload release stage, and a protection action stage. If the candidate configuration includes multiple synchronous condensers, the available reactive power upper limit of each individual synchronous condenser within the same time window is superimposed and combined to obtain the total available reactive power upper limit. If the candidate configuration contains only a single synchronous condenser, the available reactive power limit of that single synchronous condenser is the total available reactive power limit.
[0009] Furthermore, the available reactive power upper limit is a piecewise function, including: In the initial stage after a fault occurs until the overload release delay time, the available reactive power limit is the rated dynamic reactive power limit. From the overload release delay time to the protection action time, the available reactive power upper limit is the overload reactive power output capacity; After the protection action, the available reactive power limit drops to the value after the limit is applied.
[0010] Further, in step S2, calculating the equivalent reactive power support includes: Construct a reference operating point before the fault disturbance, and obtain the reference bus voltage and the reactive power injection at that operating point; Estimate the voltage deviation of the critical bus at each sampling moment within the discrimination time window; Based on the equivalent voltage reactive power sensitivity coefficient, the voltage deviation at each sampling time is mapped to the corresponding required reactive power support, and the required reactive power support between adjacent sampling times is interpolated to obtain the complete curve of the equivalent reactive power support changing with time.
[0011] Furthermore, the equivalent voltage reactive power sensitivity coefficient is obtained using an offline equivalent network calculation method, including: Based on the network topology and parameters of the new energy transmission system, an equivalent network is constructed using Thevenin's equivalent theory. By applying a unit reactive power injection disturbance at the key bus and solving for the corresponding bus voltage change, the voltage reactive power sensitivity at that bus is obtained. For different operating modes and fault types, a sensitivity curve library is pre-calculated and indexed according to operating condition categories for quick identification stage lookup.
[0012] Furthermore, the starting point of the discrimination time window is the time when the fault occurs, and the ending point is the time when the critical bus voltage recovers to the allowable range specified by the transient voltage safety criterion and does not exceed the range thereafter; when the critical bus voltage fails to recover to the allowable range within a preset time, the preset time is used as the ending point of the discrimination time window; the preset time is taken in the range of 200 milliseconds to 500 milliseconds.
[0013] Furthermore, in step S3: At any given time, if the equivalent reactive power support is less than or equal to the available overload reactive power dynamic output limit, then the reactive power margin shortfall at the corresponding time is zero. If the equivalent reactive power support is greater than the upper limit of the available overload reactive power dynamic output, then the reactive power margin deficiency at the corresponding moment is the difference between the two.
[0014] Furthermore, in step S4, the comprehensive rapid risk index is taken as the maximum value of the rapid discrimination index in all working conditions.
[0015] Furthermore, the preset threshold is determined through an offline calibration process, including: A set of benchmark candidate configurations is selected, and a fine-grained dynamic verification is performed on the set of benchmark candidate configurations to obtain the true feasibility label of each benchmark candidate configuration. Calculate the fast discrimination index for each benchmark candidate configuration, and form a calibration sample set with the corresponding real feasibility label; A threshold is selected with the constraint that the false negative rate does not exceed a preset allowable upper limit. The false negative rate is defined as the proportion of configurations that are truly infeasible but are judged as passable out of all truly infeasible configurations.
[0016] Furthermore, the preset allowable upper limit of the false negative rate is taken in the range of 1% to 5%; under the condition that the false negative rate does not exceed the preset allowable upper limit, the preset threshold is selected in a way that minimizes the false negative rate or minimizes the number of configurations that pass the quick screening. The false negative rate is defined as the proportion of truly feasible configurations that are judged to be rejected to all truly feasible configurations.
[0017] Compared with existing technologies, the advantages of this invention are as follows: This invention uses a sensitivity proxy model to replace full-time-domain simulation, and combines integral indices and threshold criteria for pre-screening, using high-precision simulation only for a small number of configurations, significantly reducing the computational burden in large-scale planning scenarios; by dynamically determining the time window and using an integral index that considers both the magnitude and duration of reactive power deficit, it overcomes the limitations of fixed time windows or single-point criteria, accurately reflecting the true voltage safety margin in transient processes; the aggregation strategy based on the most unfavorable operating conditions ensures the safety of the screening results; simultaneously, the acquisition methods for key equipment parameters are clear, and the thresholds are calibrated offline to balance the risks of missed and false judgments, demonstrating good engineering feasibility.
[0018] Furthermore, by dividing the reactive power upper limit of a single synchronous condenser into an initial stage, an overload release stage, and a protection action post-stage, this invention accurately describes the complete dynamic behavior of the condenser from forced excitation response to protection limiting at different time periods after a fault, avoiding the evaluation bias introduced by simply treating overload capacity as a constant value. By hourly superimposing the reactive power upper limits of multiple condensers within the same time window, it physically and accurately reflects the convergence effect of reactive power support capacity when multiple devices are running in parallel, making this parameterization method uniformly applicable to rapid evaluation of single-machine and multi-machine configurations, and possessing good versatility.
[0019] Furthermore, this invention divides the synchronous condenser into three time periods based on the timing logic of forced excitation start-up, overload release, and protection limiting. The piecewise function form highly matches the actual reactive power output characteristics of the equipment, avoiding the overly optimistic or pessimistic evaluation bias caused by using a single constant value. The parameters corresponding to each stage (rated capacity, overload capacity, delay time, protection action time, and value after limiting) all have clear physical meanings and engineering origins, making it easy to calibrate and correct based on equipment test data or protection settings, thus ensuring the reliability of the model. The piecewise function describes the change of reactive power over time in a step-like form, requiring only simple logical judgments and assignment operations in numerical calculations, without the need for solving complex differential equations, making it suitable for rapid traversal evaluation of a large number of candidate configurations.
[0020] Furthermore, this invention employs an equivalent linearized "voltage reactive power sensitivity" surrogate model, transforming complex nonlinear transient voltage calculations into simple algebraic operations based on the reference operating point and sensitivity coefficients. This avoids iteratively solving the power flow equations across time domains, significantly reducing the computational load for a single evaluation. By mapping and interpolating the voltage deviations at discrete sampling times to obtain a complete continuous curve, it effectively captures the transient voltage change trend while avoiding the high computational burden of dense sampling across the entire time window, achieving a good balance between computational accuracy and efficiency. The required parameters (reference operating point, voltage deviation surrogate, equivalent voltage reactive power sensitivity coefficient) can all be obtained through offline power flow calculations or Thevenin equivalence, eliminating the need for online iterations. This facilitates integration into planning and screening software and is suitable for rapid evaluation scenarios involving large-scale candidate configurations.
[0021] Furthermore, this invention completes the calculation of voltage reactive power sensitivity entirely offline. During online rapid identification, it only needs to directly look up the table according to the operating condition category, without performing time-consuming operations such as equivalent network construction, power flow calculation, or matrix inversion in real time, which significantly reduces the calculation delay in the rapid identification stage. By pre-calculating the sensitivity for different operating modes (load level, renewable energy output) and fault types (three-phase short circuit, two-phase short circuit, etc.) and establishing an index library, the rapid identification stage can match the current operating condition and call the corresponding sensitivity value, ensuring the accuracy of the equivalent reactive power support calculation under different scenarios. The computationally intensive but time-insensitive sensitivity solution work is moved forward, while the real-time-critical table lookup and comparison work is left in the online stage, forming a clear division of labor of "offline database construction - online table lookup", which facilitates modular implementation and subsequent expansion and maintenance.
[0022] Furthermore, this invention uses the actual recovery of the key bus voltage to the allowable range and its stability as the endpoint criterion, enabling the time window to adapt to the differences in voltage recovery speed under different operating conditions, fully covering the effective range of transient response, and avoiding evaluation bias caused by a fixed time window; a preset time limit of 200-500ms is set, which is forcibly truncated when the voltage recovery is slow or the oscillation does not converge, to prevent the time window from being extended indefinitely and causing calculation loss of control, thus ensuring the calculation convergence and timeliness of the screening method.
[0023] Furthermore, this invention only includes the reactive power margin deficiency when system demand exceeds equipment supply, and sets it to zero when the synchronous condenser capacity is sufficient to cover the demand. This avoids algebraically canceling out the redundant margin and the deficiency, thus accurately locating the periods when voltage instability risk actually exists during transient processes. The calculation logic directly reflects the engineering essence of how much is lacking; the larger the difference, the more insufficient the voltage safety margin at that moment, providing a basic quantity with clear physical meaning for subsequent integration and comprehensive risk quantification. By adopting a point-by-point judgment and positive truncation method, there is no need to introduce additional weights or empirical coefficients, which not only ensures the efficiency of the calculation but also allows the basic quantity to reflect both the severity and duration of the risk after integration over the time window, enhancing the objectivity and interpretability of the final evaluation index.
[0024] Furthermore, this invention selects a threshold based on the primary constraint that the false negative rate (actually infeasible but mistakenly judged as pass) does not exceed a preset allowable upper limit (1% to 5%). This mechanism ensures that the screening process will not allow configurations with transient voltage instability risks to proceed to subsequent stages, thus safeguarding the first line of defense for the safe operation of the power grid. The calibration sample is composed of fine verification results based on the benchmark sample set and measured data of rapid discrimination indicators. The threshold selection has a clear data-driven basis, avoiding the arbitrariness and uncertainty of human experience in determining values. Different implementers can obtain consistent threshold results by using the same set of sample sets.
[0025] Furthermore, this invention strictly controls the false negative rate within the safety limit of 1% to 5%, prioritizing ensuring that no risky configurations are overlooked. On this basis, it further minimizes the false positive rate, that is, minimizing the false rejection of feasible solutions while ensuring the safe operation of the power grid, thus balancing safety and the sufficiency of solution selection. When improving computational efficiency is the main objective, the threshold can be determined by minimizing the number of configurations that pass the rapid screening, thereby minimizing the number of candidate configurations that enter the high-cost fine-grained verification stage and concentrating limited computational resources on the few most valuable solutions. Attached Figure Description
[0026] Figure 1 This is a flowchart of the method for rapid identification and configuration screening of transient voltage constraints of synchronous condensers provided in the embodiments of this application. Detailed Implementation
[0027] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.
[0028] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.
[0029] Figure 1 This is a flowchart of the method for rapid identification and configuration screening of transient voltage constraints of synchronous condensers provided in the embodiments of this application.
[0030] like Figure 1 As shown, the technical solution provided in this application includes the following steps: Step S1: Obtain the candidate synchronous condenser configuration set, operating condition set, transient voltage safety criteria, synchronous condenser overload reactive power dynamic capability parameters, and new energy transmission system model parameters. Specifically, the candidate synchronous condenser configuration set includes multiple distributed synchronous condenser deployment schemes to be evaluated, each defining the access bus location, number of synchronous condensers, and capacity level. The operating condition set includes combinations of various operating modes and fault types, such as typical disturbance scenarios like AC line short-circuit faults and DC commutation failures under different load levels. The transient voltage safety criteria define the constraints that the critical bus voltage must meet during fault disturbances and recovery processes, which may be expressed as voltage allowable range, peak limit, recovery time requirements, or over-limit integral indicators.
[0031] The overload reactive power dynamic capability parameter of a synchronous condenser is used to characterize its ability to provide additional reactive power support during faults. This parameter is determined based on the electromagnetic transient characteristics of the synchronous condenser itself, and is obtained as follows: Rated dynamic non-functional upper limit and overload reactive power output The excitation system mathematical model and strong excitation capability curve provided by the synchronous condenser manufacturer are obtained, or the short-time overcurrent capability index specified in the national standard GB / T7064 "Technical Requirements for Salient Pole Synchronous Generators" is converted. Overload release delay time , refers to the delay time required from the moment the fault occurs until the synchronous condenser excitation system completes the strong excitation response and the reactive power output is increased to the overload capacity level. This parameter is calculated based on the excitation system voltage response time constant Td0′ and the peak voltage multiple, or determined by actual measurement through the equipment factory type test. Protective action moment This refers to the moment when the overexcitation limiter or rotor overload protection of the synchronous condenser activates and begins to limit reactive power output. This parameter is determined based on the setting value and action curve of the protection device. Value after limiting This refers to the upper limit of the non-functional capacity that the synchronous condenser can output for a long period of time after the protection action. This parameter is determined based on the rated capacity and cooling conditions of the equipment.
[0032] The above parameters can be given individually for each device, or they can be categorized by device model and stored in a table in the parameter library for later use.
[0033] The parameters of the renewable energy transmission system model include the power grid topology, line impedance, transformer parameters, load characteristics, and renewable energy unit models, which are used to calculate voltage deviation and reactive power demand.
[0034] Step S2: For each candidate configuration and operating condition, set a judgment time window after the fault disturbance. Based on the parameters of the new energy transmission system model, calculate the equivalent reactive power support required to make the key bus voltage meet the transient voltage safety criterion at each moment within the judgment time window. Based on the dynamic overload reactive power capability parameters of the synchronous condenser, calculate the upper limit of the available overload reactive power dynamic output that the synchronous condenser in the candidate configuration can provide at each moment within the judgment time window. The end point of the judgment time window is dynamically determined according to the actual moment when the key bus voltage under the candidate configuration recovers to meet the transient voltage safety criterion and remains stable. In this step, a discrimination time window is first set for each operating condition. The starting point of the discrimination time window is set to the moment the fault occurs. The ending point of the discrimination time window is not a fixed value, but is dynamically determined based on the actual recovery of the critical bus voltage under the candidate configuration. Specifically, the ending point is the moment when the critical bus voltage recovers to the allowable range specified by the transient voltage safety criteria, and does not subsequently exceed that range. This dynamic determination method can more accurately capture the entire process of voltage transient response and avoid evaluation bias that may be caused by a fixed time window.
[0035] After determining the discrimination time window, the equivalent reactive power support required to ensure the critical bus voltage meets the transient voltage safety criterion at each moment within the discrimination time window is calculated based on the parameters of the new energy transmission system model. This calculation process employs an equivalent linearized "voltage reactive power sensitivity" proxy calculation path. First, a reference operating point before the fault disturbance is constructed, and the reference bus voltage and the reactive power injection at that operating point are obtained. This reference operating point reflects the steady-state or quasi-steady-state operating state of the system before the fault occurs. Next, the voltage deviation proxy amount of the critical bus at each sampling moment within the discrimination time window is estimated. Specifically, this application provides a preferred method for estimating the voltage deviation proxy amount, including the following sub-steps: Sub-step S2a involves performing Thevenin equivalents on the new energy transmission system based on the system power flow profile before the fault occurred, with the critical bus as the boundary node, to obtain the equivalent potential at the critical bus. and equivalent impedance ; Sub-step S2b involves determining each sampling time within the discrimination time window. Determine the equivalent short-circuit impedance at the fault point based on the fault type (e.g., three-phase short circuit, single-phase short circuit, etc.). And calculate the impact coefficient of the fault point on the voltage drop of the critical bus. This coefficient can be determined by the electrical distance between the fault point and the critical busbar, as well as... and The pressure distribution relationship is obtained; Sub-step S2c calculates the voltage deviation proxy at each sampling time. : ,in This is the reference voltage amplitude of the critical busbar before the fault; As an alternative, the voltage deviation proxy can also be obtained by interpolation using a table lookup of the voltage-time response curve based on time-domain simulation pre-calculation. Regardless of the specific method used, the obtained voltage deviation proxy should reflect the deviation trend of the critical bus voltage relative to the reference value under fault disturbance.
[0036] Subsequently, based on the equivalent voltage reactive power sensitivity coefficient, the voltage deviation at each sampling time is mapped to the corresponding required reactive power support. The equivalent voltage reactive power sensitivity coefficient characterizes the approximately linear relationship between voltage deviation and reactive power injection, satisfying that the voltage deviation is approximately equal to the product of the sensitivity coefficient and the reactive power injection. By interpolating the required reactive power support between adjacent sampling times, a complete curve of the equivalent reactive power support changing with time is obtained.
[0037] Simultaneously, based on the overload reactive power dynamic capability parameters of the synchronous condenser, the upper limit of the available overload reactive power dynamic output that the synchronous condenser can provide in the candidate configurations at each time point within the discrimination time window is calculated. This upper limit reflects the maximum reactive power support capability that the synchronous condenser can actually provide during a fault, taking into account the overload characteristics of the equipment itself and the impact of protection actions.
[0038] Step S3: At the same time, compare the available overload reactive dynamic output upper limit with the equivalent reactive support amount hourly. Take the positive value of the equivalent reactive support amount exceeding the available overload reactive dynamic output upper limit as the reactive margin deficiency at that time. Integrate the reactive margin deficiency within the discrimination time window to obtain the rapid discrimination index of the candidate configuration under the corresponding operating condition. Specifically, at each moment within the discrimination time window Compare the equivalent reactive power support at that moment. With available overload reactive power dynamic output limit .like The reactive power margin at the corresponding moment is insufficient. It is zero; if The reactive power margin at the corresponding moment is insufficient. The difference between the two is, Subsequently, the insufficient reactive power margin was addressed. In judging the time window Integrating within the range yields a rapid discrimination index for the candidate configuration under this operating condition. This indicator quantifies the degree and duration of the inadequacy of the reactive power support capacity provided by the synchronous condenser relative to system demand under this operating condition.
[0039] Step S4: Aggregate the rapid discrimination indicators under each working condition to determine the comprehensive rapid risk indicator of the candidate configuration; In this step, the rapid discrimination indicators calculated for the same candidate configuration under all operating conditions in the operating condition set are aggregated to determine the comprehensive rapid risk indicator for that candidate configuration. The aggregation method is designed to reflect the risk level of candidate configurations under the most unfavorable operating conditions, thereby ensuring the safety of the screening results.
[0040] Step S5: Compare the comprehensive rapid risk index with a preset threshold. If the comprehensive rapid risk index is less than or equal to the preset threshold, the candidate configuration is determined to have passed the rapid screening under transient voltage safety constraints and enter the fine verification stage. If the comprehensive rapid risk index is greater than the preset threshold, the candidate configuration is determined to not meet the transient voltage safety constraints and is eliminated.
[0041] Specifically, the comprehensive rapid risk index obtained in step S4 With preset threshold Compare. If If so, the candidate configuration is determined to have passed the rapid screening under transient voltage safety constraints, and is retained for subsequent fine verification; if If a candidate configuration fails to meet the transient voltage safety constraints, it is directly eliminated, and no further high-cost simulation calculations are performed. This screening mechanism significantly reduces the number of candidate configurations that need to enter the fine verification stage, thereby improving overall computational efficiency. The fine verification stage can use electromagnetic transient simulation (PSCAD / EMTDC) or electromechanical-electromagnetic hybrid simulation to accurately verify the candidate configurations that have passed the rapid screening, in order to confirm whether their transient voltage response strictly meets the safety criteria.
[0042] This invention employs a sensitivity proxy model to replace full-time-domain simulation and combines integral indices and threshold criteria for pre-screening. High-precision simulation is used only for a small number of configurations, significantly reducing the computational burden in large-scale planning scenarios. By dynamically determining the time window and using an integral index that considers both the magnitude and duration of reactive power deficit, it overcomes the limitations of fixed time windows or single-point criteria, accurately reflecting the true voltage safety margin during transient processes. The aggregation strategy based on the most unfavorable operating conditions ensures the safety of the screening results. At the same time, the acquisition methods for key equipment parameters are clear, and the thresholds are calibrated offline to account for both missed and false judgments, demonstrating good engineering feasibility.
[0043] Specifically, in step S2, the overload reactive power dynamic capability parameters of the synchronous condenser are generated using a piecewise curve parameterization method, specifically including: The upper limit of available reactive power for a single synchronous condenser under preset operating conditions is determined by a piecewise function, which includes at least three stages: an initial stage, an overload release stage, and a protection action stage. If the candidate configuration includes multiple synchronous condensers, the available reactive power upper limit of each individual synchronous condenser within the same time window is superimposed and combined to obtain the total available reactive power upper limit. If the candidate configuration contains only a single synchronous condenser, the available reactive power limit of that single synchronous condenser is the total available reactive power limit.
[0044] Specifically, for a single synchronous condenser, its available reactive power limit is... Over time The changing relationship is expressed as a piecewise function. This piecewise function is divided into different time phases based on the overload characteristics and protection logic of the synchronous condenser. If multiple synchronous condensers are deployed on multiple access buses in the candidate configuration, the available reactive power upper limit of each synchronous condenser is superimposed and synthesized within the same time window to obtain the total available reactive power upper limit of the candidate configuration. This superposition and synthesis is based on the conventional assumption that the reactive power of each synchronous condenser is aggregated equally through the bus and there is no shared limit. If the candidate configuration contains only a single synchronous condenser, then superposition is not required, and the upper limit of the available reactive power of that single synchronous condenser is the upper limit of the total available reactive power.
[0045] This invention accurately describes the complete dynamic behavior of a single synchronous condenser from forced excitation response to protection limiting during different time periods after a fault by dividing the reactive power upper limit of a single synchronous condenser into an initial stage, an overload release stage, and a protection action post-stage. This avoids the evaluation bias introduced by simply treating overload capacity as a constant value. By hourly superimposing the reactive power upper limits of multiple synchronous condensers within the same time window, it physically and accurately reflects the convergence effect of reactive power support capacity when multiple devices are running in parallel. This parameterization method can be uniformly applied to the rapid evaluation of single-machine and multi-machine configurations, and has good versatility.
[0046] Specifically, the available reactive power upper limit is a piecewise function, including: In the initial stage after a fault occurs until the overload release delay time, the available reactive power limit is the rated dynamic reactive power limit. From the overload release delay time to the protection action time, the available reactive power upper limit is the overload reactive power output capacity; After the protection action, the available reactive power limit drops to the value after the limit is applied.
[0047] Specifically, let the time of the fault be... The first stage is the initial stage, and the time interval is... ,in The overload release delay period is the upper limit of available reactive power of the synchronous condenser during this phase. Maintain at the rated dynamic non-functional upper limit This reflects the state immediately after a fault occurs and before the excitation system has fully responded to the overload release; the second stage is the overload release stage, with a time interval of [missing information]. ,in To protect the timing of actions, the upper limit of available reactive power for the synchronous condenser is set during this phase. Upgraded to overload reactive power output capability This reflects the state of the synchronous condenser providing additional reactive power support using its overload capacity after the excitation system responds; the third stage is the stage after the protection action, with a time interval of During this phase, the upper limit of available reactive power for synchronous condensers. The value after dropping to the limit This indicates that the overload protection or control system of the synchronous condenser has started to take effect, limiting its continued output of excessive reactive power to protect the equipment.
[0048] This invention divides the synchronous condenser into three time periods based on the timing logic of forced excitation start-up, overload release, and protection limiting. The piecewise function form closely matches the actual reactive power output characteristics of the equipment, avoiding overly optimistic or pessimistic evaluation biases caused by using a single constant value. The parameters corresponding to each stage (rated capacity, overload capacity, delay time, protection action time, and value after limiting) all have clear physical meanings and engineering origins, facilitating calibration and correction based on equipment test data or protection settings, thus ensuring the reliability of the model. The piecewise function describes the change of reactive power over time in a step-like form, requiring only simple logical judgments and assignment operations in numerical calculations, without the need for solving complex differential equations, making it suitable for rapid traversal evaluation of a large number of candidate configurations.
[0049] Specifically, in step S2, calculating the equivalent reactive power support includes: Construct a reference operating point before the fault disturbance, and obtain the reference bus voltage and the reactive power injection at that operating point; Estimate the voltage deviation of the critical bus at each sampling moment within the discrimination time window; Based on the equivalent voltage reactive power sensitivity coefficient, the voltage deviation at each sampling time is mapped to the corresponding required reactive power support, and the required reactive power support between adjacent sampling times is interpolated to obtain the complete curve of the equivalent reactive power support changing with time.
[0050] Specifically, a baseline operating point is first constructed before the fault disturbance. This operating point is based on the system operating state before the fault occurs, and the reference voltage of the critical bus is obtained. and the reactive power injection at that operating point Next, the sampling times of the critical bus within the discrimination time window are estimated. Voltage deviation agent quantity The voltage deviation proxy reflects the trend of bus voltage change relative to the reference voltage under fault disturbance. Subsequently, based on the equivalent voltage reactive power sensitivity coefficient... The voltage deviation at each sampling time Mapped to the corresponding required reactive power support. The mapping relationship satisfies Thus we can obtain Finally, the required reactive power support between adjacent sampling times is interpolated, for example using piecewise linear interpolation, to obtain a complete curve of the equivalent reactive power support changing with time. .
[0051] This invention employs an equivalent linearized "voltage reactive power sensitivity" surrogate model, transforming complex nonlinear transient voltage calculations into simple algebraic operations based on a reference operating point and sensitivity coefficients. This avoids iteratively solving power flow equations across time domains, significantly reducing the computational load for a single evaluation. By mapping and interpolating the voltage deviations at discrete sampling times to obtain a complete continuous curve, it effectively captures the trend of transient voltage changes while avoiding the high computational burden of dense sampling across the entire time window, achieving a good balance between computational accuracy and efficiency. The required parameters (reference operating point, voltage deviation surrogate, equivalent voltage reactive power sensitivity coefficient) can all be obtained through offline power flow calculations or Thevenin equivalence, eliminating the need for online iterations. This facilitates integration into planning and screening software and is suitable for rapid evaluation scenarios involving large-scale candidate configurations.
[0052] Specifically, the equivalent voltage reactive power sensitivity coefficient is obtained using an offline equivalent network calculation method, including: Based on the network topology and parameters of the new energy transmission system, an equivalent network is constructed using Thevenin's equivalent theory. By applying a unit reactive power injection disturbance at the key bus and solving for the corresponding bus voltage change, the voltage reactive power sensitivity at that bus is obtained. For different operating modes and fault types, a sensitivity curve library is pre-calculated and indexed according to operating condition categories for quick identification stage lookup.
[0053] Specifically, the equivalent voltage reactive power sensitivity coefficient The acquisition process is completed offline. Based on the network topology and parameters of the new energy transmission system, an equivalent network is constructed using Thevenin's equivalent theory. A unit reactive power injection disturbance is applied at the critical bus. Solve for the corresponding bus voltage change. Voltage reactive power sensitivity That is For different operating modes (such as different load levels and different renewable energy outputs) and fault types (such as three-phase short circuits and two-phase short circuits), the corresponding voltage and reactive power sensitivities are pre-calculated, and a sensitivity curve library indexed by operating condition category is established. In the rapid identification stage, the corresponding sensitivity curve is directly looked up in the table according to the current operating condition category, without the need for complex network calculations in real time, thereby improving the calculation speed.
[0054] This invention completes the calculation of voltage reactive power sensitivity entirely offline. During online rapid identification, it only needs to directly look up the table according to the operating condition category, without performing time-consuming operations such as equivalent network construction, power flow calculation, or matrix inversion in real time, which significantly reduces the calculation delay in the rapid identification stage. By pre-calculating the sensitivity for different operating modes (load level, renewable energy output) and fault types (three-phase short circuit, two-phase short circuit, etc.) and establishing an index library, the rapid identification stage can match the current operating condition and call the corresponding sensitivity value, ensuring the accuracy of the equivalent reactive power support calculation under different scenarios. The computationally intensive but time-insensitive sensitivity solution work is moved forward, while the real-time-critical table lookup and comparison work is left in the online stage, forming a clear division of labor of "offline database construction - online table lookup", which facilitates modular implementation and subsequent expansion and maintenance.
[0055] Specifically, the starting point of the discrimination time window is the time when the fault occurs, and the ending point is the time when the critical bus voltage recovers to the allowable range specified by the transient voltage safety criterion and does not exceed the range thereafter; when the critical bus voltage fails to recover to the allowable range within a preset time, the preset time is used as the ending point of the discrimination time window; the preset time is taken in the range of 200 milliseconds to 500 milliseconds.
[0056] Specifically, the starting point of the judgment time window is fixed at the moment the fault occurs. Determine the end of the time window. Determined dynamically according to the following logic: First, the critical bus voltage V(t) is restored to the allowable range specified by the transient voltage safety criteria. The criteria are as follows: After V(t) enters this interval, it is further determined whether it remains within this interval for multiple consecutive sampling points (e.g., 5 consecutive sampling points with a total duration of not less than 20 milliseconds). If so, this moment is determined as the recovery and stabilization moment and is used as the end point of the discrimination time window. .
[0057] If, during the aforementioned determination process, the critical bus voltage is within a preset time period... If the above conditions for restoring stability are consistently not met, then a forced measure will be taken. This serves as the endpoint of the time window for judgment, to prevent the time window from extending indefinitely. The preset duration... The value ranges from 200 milliseconds to 500 milliseconds. Preferably, 300 milliseconds is used for AC short circuit faults and 400 milliseconds is used for DC commutation failure faults.
[0058] This invention uses the actual recovery of the critical bus voltage to the allowable range and its stability as the endpoint criterion, enabling the time window to adapt to the differences in voltage recovery speed under different operating conditions, fully covering the effective range of transient response, and avoiding evaluation bias caused by a fixed time window; a preset time limit of 200-500ms is set, which is forcibly truncated when the voltage recovery is slow or the oscillation does not converge, to prevent the time window from being extended indefinitely and causing calculation loss of control, thus ensuring the calculation convergence and timeliness of the screening method.
[0059] Specifically, in step S3: At any given time, if the equivalent reactive power support is less than or equal to the available overload reactive power dynamic output limit, then the reactive power margin shortfall at the corresponding time is zero. If the equivalent reactive power support is greater than the upper limit of the available overload reactive power dynamic output, then the reactive power margin deficiency at the corresponding moment is the difference between the two.
[0060] In this embodiment, the calculation process for insufficient reactive power margin is as follows: the difference between the reactive power required by the system and the reactive power provided by the synchronous condenser directly represents the degree of lack of transient voltage safety margin at that moment.
[0061] when When D(t) is zero; when When D(t) takes the difference between the two, compares them hourly and only includes the positive difference, which can avoid the positive and negative deviations canceling each other out within the whole time window and accurately reflect the instantaneous severity of insufficient reactive power support during the transient process.
[0062] In discretized numerical computation, if the sampling step size is Δt, then the index can be quickly determined. The integral form can be discretized as follows: ; in Let Δt be the k-th sampling time, and N be the total number of sampling points. Preferably, Δt is between 1 and 5 milliseconds to effectively capture the abrupt changes in reactive power before and after the overload release phase and the protection action. This indicator takes into account both the amplitude and duration of reactive power deficit, avoiding the one-sidedness of single-point criteria.
[0063] This invention only accounts for insufficient reactive power margin when system demand exceeds equipment supply. When the synchronous condenser's capacity is sufficient to cover the demand, it directly sets the reactive power margin to zero, avoiding algebraic cancellation between redundant margin and insufficient margin. This allows for precise location of the actual voltage instability risk during transient processes. The calculation logic directly reflects the engineering essence of the deficit; a larger difference indicates a more insufficient voltage safety margin at that moment, providing a fundamental quantity with clear physical meaning for subsequent integration and comprehensive risk quantification. By adopting point-by-point judgment and positive truncation, no additional weights or empirical coefficients are needed, ensuring both computational efficiency and enabling the fundamental quantity to reflect both the severity and duration of the risk after integration over the time window, thus enhancing the objectivity and interpretability of the final evaluation indicators.
[0064] Specifically, in step S4, the comprehensive rapid risk index is the maximum value of the rapid discrimination index in all working conditions.
[0065] Specifically, let the set of operating conditions be... Candidate configurations under operating conditions The following quick discrimination indicators are Then the comprehensive rapid risk index of candidate configurations. Defined as: The use of maximum value aggregation reflects the worst-case safety-oriented principle, meaning that candidate configurations must meet safety constraints under all operating conditions, especially the highest-risk ones. If the rapid discrimination index is still less than or equal to the preset threshold under the worst-case condition, it is more likely to meet the constraints under other operating conditions, thus ensuring the safety of the screening results.
[0066] Specifically, the preset threshold is determined through an offline calibration process, including: A set of benchmark candidate configurations is selected, and a fine-grained dynamic verification is performed on the set of benchmark candidate configurations to obtain the true feasibility label of each benchmark candidate configuration. Calculate the fast discrimination index for each benchmark candidate configuration, and form a calibration sample set with the corresponding real feasibility label; A threshold is selected with the constraint that the false negative rate does not exceed a preset allowable upper limit. The false negative rate is defined as the proportion of configurations that are truly infeasible but are judged as passable out of all truly infeasible configurations.
[0067] Specifically, preset threshold The determination process is as follows: First, select a baseline candidate configuration set. This set contains a number of typical candidate configurations, covering different layout schemes, capacity levels, and access locations. Next, fine-grained dynamic verification is performed on each configuration in the baseline candidate configuration set. For example, electromagnetic transient simulation or high-precision equivalent dynamic models are used to calculate its transient voltage index, and its feasibility is determined according to transient voltage safety criteria. If the voltage constraint is met, it is marked as "feasible"; otherwise, it is marked as "infeasible". Then, a rapid discrimination index for each baseline candidate configuration is calculated. This data, along with corresponding real feasibility labels, forms a calibration sample set. Finally, a threshold is selected with the constraint that the false negative rate does not exceed a preset allowable upper limit. The false negative rate is defined as a data point that is truly infeasible but is judged as passable. This ensures that the false negative rate meets the preset allowable upper limit, while optimizing other performance indicators.
[0068] This invention selects a threshold based on the primary constraint that the false negative rate (actually infeasible but mistakenly judged as pass) does not exceed a preset allowable upper limit (1% to 5%). This mechanism ensures that the screening process will not allow configurations with transient voltage instability risks to proceed to subsequent stages, thus safeguarding the first line of defense for the safe operation of the power grid. The calibration sample is composed of fine verification results based on the benchmark sample set and measured data of rapid discrimination indicators. The threshold selection has a clear data-driven basis, avoiding the arbitrariness and uncertainty of human experience in determining values. Different implementers can obtain consistent threshold results by using the same set of sample sets.
[0069] Specifically, the preset allowable upper limit of the false negative rate is between 1% and 5%; under the condition that the false negative rate does not exceed the preset allowable upper limit, the preset threshold is selected in a way that minimizes the false negative rate or minimizes the number of configurations that pass the quick screening. The false negative rate is defined as the proportion of truly feasible configurations that are judged to be rejected to all truly feasible configurations.
[0070] Specifically, the preset upper limit for the false negative rate. The value ranges from 1% to 5%. This meets the requirement of a low false negative rate. Under the condition of selecting a preset threshold The goal is to minimize the false positive rate or the number of configurations that pass the fast screening. The false positive rate is defined as the proportion of truly feasible configurations that are rejected out of all truly feasible configurations. In this way, while ensuring safety (controlling the false negative rate), we can maximize the accuracy of the screening (reducing the false positive rate) or minimize the number of configurations that enter the fine-grained verification stage (improving computational efficiency).
[0071] This invention strictly controls the false negative rate within a safe range of 1% to 5%, prioritizing ensuring that no risky configurations are overlooked. On this basis, it further minimizes the false positive rate, that is, minimizing the false rejection of feasible solutions while ensuring the safe operation of the power grid, thus balancing security and the sufficiency of solution selection. When improving computational efficiency is the primary objective, the threshold can be determined by minimizing the number of configurations that pass the rapid screening, thereby minimizing the number of candidate configurations that enter the high-cost fine-grained verification stage and concentrating limited computational resources on the few most valuable solutions.
[0072] Specifically, the fine-grained verification phase includes: For candidate configurations that pass the rapid screening, the transient indicators of key bus voltage are calculated using an equivalent dynamic model or electromagnetic transient simulation. The final feasibility is determined according to the transient voltage safety criteria, and the final feasible configuration set or the ranking result used for subsequent optimization is output.
[0073] In this embodiment, 36 candidate configurations were quickly screened and then entered the fine-grained verification stage. The PSD-BPA electromechanical transient simulation program was used to perform full-time-domain simulations of all 36 operating conditions for each of the 36 configurations. Three transient indicators—voltage drop depth, peak transient overvoltage, and voltage recovery time—were extracted for each operating condition and compared with transient voltage safety criteria. Configurations were deemed feasible if all three indicators were met simultaneously. After evaluation, 32 configurations were deemed feasible, while 4 were eliminated because the peak transient overvoltage exceeded 1.15 pu. Finally, a set of 32 feasible configurations was output, sorted from smallest to largest by the comprehensive rapid risk index. Among the 18 configurations with a comprehensive rapid risk index of 0, the scheme with the lowest total capacity (600 Mvar) and the fewest access points (1) was recommended as the optimal configuration.
[0074] Specifically, after the candidate configuration passes the rapid screening and enters the fine-grained verification stage, the method further includes: The transient indices of the key bus voltage under each candidate configuration are calculated using an equivalent dynamic model or electromagnetic transient simulation. The transient indicators are compared with the transient voltage safety criteria to determine the final feasibility of each candidate configuration. A final set of feasible configurations is generated based on the judgment results; If multiple feasible configurations exist, sort them in ascending order of the comprehensive rapid risk index values, and output the sorting results for subsequent optimization.
[0075] The above embodiments are merely illustrative examples and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this application.
Claims
1. A method for rapid identification and configuration selection of transient voltage constraints in synchronous condensers, characterized in that, include: Step S1: Obtain the candidate synchronous condenser configuration set, operating condition set, transient voltage safety criteria, synchronous condenser overload reactive power dynamic capability parameters, and new energy transmission system model parameters. Step S2: For each candidate configuration and operating condition, set a judgment time window after the fault disturbance. Based on the parameters of the new energy transmission system model, calculate the equivalent reactive power support required to make the key bus voltage meet the transient voltage safety criterion at each moment within the judgment time window. Based on the dynamic overload reactive power capability parameters of the synchronous condenser, calculate the upper limit of the available overload reactive power dynamic output that the synchronous condenser in the candidate configuration can provide at each moment within the judgment time window. The end point of the judgment time window is dynamically determined according to the actual moment when the key bus voltage under the candidate configuration recovers to meet the transient voltage safety criterion and remains stable. Step S3: At the same time, compare the available overload reactive dynamic output upper limit with the equivalent reactive support amount hourly. Take the positive value of the equivalent reactive support amount exceeding the available overload reactive dynamic output upper limit as the reactive margin deficiency at that time. Integrate the reactive margin deficiency within the discrimination time window to obtain the rapid discrimination index of the candidate configuration under the corresponding operating condition. Step S4: Aggregate the rapid discrimination indicators under each working condition to determine the comprehensive rapid risk indicator of the candidate configuration; Step S5: Compare the comprehensive rapid risk index with a preset threshold. If the comprehensive rapid risk index is less than or equal to the preset threshold, the candidate configuration is determined to have passed the rapid screening under transient voltage safety constraints and enter the fine verification stage. If the comprehensive rapid risk index is greater than the preset threshold, the candidate configuration is determined to not meet the transient voltage safety constraints and is eliminated.
2. The method according to claim 1, characterized in that, In step S2, the overload reactive power dynamic capability parameters of the synchronous condenser are generated using a piecewise curve parameterization method, specifically including: The upper limit of available reactive power for a single synchronous condenser under preset operating conditions is determined by a piecewise function, which includes at least three stages: an initial stage, an overload release stage, and a protection action stage. If the candidate configuration includes multiple synchronous condensers, the available reactive power upper limit of each individual synchronous condenser within the same time window is superimposed and combined to obtain the total available reactive power upper limit. If the candidate configuration contains only a single synchronous condenser, the available reactive power limit of that single synchronous condenser is the total available reactive power limit.
3. The method according to claim 2, characterized in that, The available reactive power upper limit is a piecewise function, including: In the initial stage after a fault occurs until the overload release delay time, the available reactive power limit is the rated dynamic reactive power limit. From the overload release delay time to the protection action time, the available reactive power upper limit is the overload reactive power output capacity; After the protection action, the available reactive power limit drops to the value after the limit is applied.
4. The method according to claim 1, characterized in that, In step S2, calculating the equivalent reactive power support includes: Construct a reference operating point before the fault disturbance, and obtain the reference bus voltage and the reactive power injection at that operating point; Estimate the voltage deviation of the critical bus at each sampling moment within the discrimination time window; Based on the equivalent voltage reactive power sensitivity coefficient, the voltage deviation at each sampling time is mapped to the corresponding required reactive power support, and the required reactive power support between adjacent sampling times is interpolated to obtain the complete curve of the equivalent reactive power support changing with time.
5. The method according to claim 4, characterized in that, The equivalent voltage reactive power sensitivity coefficient is obtained using an offline equivalent network calculation method, including: Based on the network topology and parameters of the new energy transmission system, an equivalent network is constructed using Thevenin's equivalent theory. By applying a unit reactive power injection disturbance at the key bus and solving for the corresponding bus voltage change, the voltage reactive power sensitivity at that bus is obtained. For different operating modes and fault types, a sensitivity curve library is pre-calculated and indexed according to operating condition categories for quick identification stage lookup.
6. The method according to claim 1, characterized in that, The starting point of the discrimination time window is the time when the fault occurs, and the ending point is the time when the critical bus voltage recovers to the allowable range specified by the transient voltage safety criterion and does not exceed the range thereafter. When the critical bus voltage fails to recover to the allowable range within a preset time, the preset time is used as the ending point of the discrimination time window. The preset time is taken in the range of 200 milliseconds to 500 milliseconds.
7. The method according to claim 1, characterized in that, In step S3: At any given time, if the equivalent reactive power support is less than or equal to the available overload reactive power dynamic output limit, then the reactive power margin shortfall at the corresponding time is zero. If the equivalent reactive power support is greater than the upper limit of the available overload reactive power dynamic output, then the reactive power margin deficiency at the corresponding moment is the difference between the two.
8. The method according to claim 1, characterized in that, In step S4, the comprehensive rapid risk index is taken as the maximum value of the rapid discrimination index in all working conditions.
9. The method according to claim 1, characterized in that, The preset threshold is determined through an offline calibration process, including: A set of benchmark candidate configurations is selected, and a fine-grained dynamic verification is performed on the set of benchmark candidate configurations to obtain the true feasibility label of each benchmark candidate configuration. Calculate the fast discrimination index for each benchmark candidate configuration, and form a calibration sample set with the corresponding real feasibility label; A threshold is selected with the constraint that the false negative rate does not exceed a preset allowable upper limit. The false negative rate is defined as the proportion of configurations that are truly infeasible but are judged as passable out of all truly infeasible configurations.
10. The method according to claim 9, characterized in that, The preset allowable upper limit of the false negative rate is between 1% and 5%. Under the condition that the false negative rate does not exceed the preset allowable upper limit, the preset threshold is selected in a way that minimizes the false negative rate or minimizes the number of configurations that pass the quick screening. The false negative rate is defined as the proportion of truly feasible configurations that are judged to be rejected to all truly feasible configurations.
