A collaborative optimization method and system for multiple types of active support equipment at the weak end

Abstract

This invention relates to the field of power system planning and operation optimization technology, specifically disclosing a collaborative optimization method and system for multiple types of active support equipment at the weak-sending end. The method includes establishing a voltage-reactive power droop characteristic equation and constructing a unified power flow-droop static model to describe the system's steady-state operating point; constructing comprehensive performance indicators for multiple scenarios and a unified stiffness constraint model considering the short-circuit capacity of the external grid and the reactive power support capacity of the multiple types of active support equipment; at the capacity configuration layer, minimizing the weighted difference between investment cost and stability benefits is the objective; at the operation layer, minimizing the weighted sum of reactive power output adjustments for multiple types of equipment is the objective. This invention achieves coordinated improvement in configuration and operation to enhance the static margin in the most unfavorable direction and expand the stability domain form by solving capacity configuration using a surrogate model and linear programming at the upper layer, and optimizing reactive power sharing based on Q / V sensitivity and action costs at the lower layer.

Description

Technical Field

[0001] This invention relates to the field of power system planning and operation optimization technology, specifically a collaborative optimization method and system for multiple types of active support equipment at the weak-sending end. Background Technology

[0002] As countries worldwide promote the sustainable transformation of their energy systems, large-scale renewable energy bases and long-distance transmission projects continue to advance, resulting in sending-end power grids exhibiting characteristics of high renewable energy ratios, high power electronics penetration, and weak short-circuit support. Against this backdrop, the transmission of large-scale renewable energy from these bases not only faces voltage and stability constraints caused by insufficient system inertia and short-circuit capacity, but also requires comprehensive consideration of multiple objectives, including renewable energy absorption efficiency, configuration costs, and system resilience, while ensuring safe system operation. This places higher demands on planning and operational coordination. In this context, grid-following (GFL) / grid-forming (GFM) wind turbines, grid-forming static var generators (SVG), synchronous condensers (SC), and energy storage-type voltage support devices are widely used to improve the voltage stiffness and stability margin of weak sending-end systems. However, these active support devices introduce coupling across multiple time scales, control links, and operating modes during voltage control, making the static voltage stability of the sending-end system highly dependent on equipment capacity configuration, control parameter tuning, and operational coordination.

[0003] Traditional reactive power planning methods, which focus on a single operating condition, a single transmission direction, or a single indicator, are insufficient to characterize the most unfavorable static stability boundary under multi-directional power disturbances. On the other hand, there is a significant coupling between capacity configuration and reactive power sharing during operation: capacity configuration alone may result in meeting the minimum stiffness requirement but insufficient robustness in all directions, while operation optimization alone may lead to capacity redundancy but poor marginal benefits or local feasibility but infeasibility across different scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a collaborative optimization method and system for multiple types of active support equipment at the weak end, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A collaborative optimization method for multiple types of active support equipment at the weak feed end, the method comprising:

[0007] Based on the control parameters of various types of active support equipment, voltage-reactive power droop characteristic equations of each type of active support equipment are established, and the droop characteristic equations are coupled with the power flow equations of the power grid to construct a unified power flow-droop static model for describing the steady-state operating point of the system.

[0008] Based on the unified power flow-droop static model, a multi-scenario comprehensive performance index for characterizing the static voltage stability of the system is constructed. The index includes at least the minimum static stability margin and the area of ​​the stability domain.

[0009] A unified stiffness constraint model is constructed that takes into account the short-circuit capacity of the external network and the reactive power support capacity of the various types of active support equipment, and capacity constraints on the equipment to be configured are generated based on the stiffness constraint model.

[0010] At the capacity configuration layer, the rated capacity of the equipment to be configured is used as the decision variable, and the goal is to minimize the weighted difference between investment cost and stability benefit. Under the condition of satisfying the capacity constraints, the optimal capacity configuration scheme is obtained.

[0011] At the operational layer, based on the optimal capacity configuration scheme, with the goal of minimizing the weighted sum of reactive power output adjustments for multiple types of equipment, the reactive power sharing strategy for each active support equipment is solved.

[0012] As a further aspect of the present invention, the construction of the unified power flow-static model includes: characterizing the steady-state behavior of various types of active support equipment as voltage / reactive power droop characteristic equations at a common coupling point, and coupling them with the power flow equations to form a set of nonlinear algebraic equations describing the steady-state operating point of the system.

[0013] As a further embodiment of the present invention, the voltage / reactive power droop characteristic equation is expressed as:

[0014] ;

[0015] in, Based on reactive power output; Q m,k For reactive power output of the equipment, k m,k V represents the sag slope of each device. PCC,k V is the voltage amplitude of the k-th PCC. ref,k This is the voltage reference value.

[0016] As a further aspect of the present invention, the construction of the multi-scenario comprehensive performance index includes: determining the static voltage stability limit of each scenario along each direction for a preset set of operating scenarios and a set of disturbance directions;

[0017] A static voltage stability region is constructed based on the static voltage stability limit, and the minimum static stability margin and stability region area are calculated. Then, a comprehensive performance index is obtained by weighted summation based on scenario weights.

[0018] ;

[0019] In the formula, As scene weight, As the minimum margin indicator, A (s) This is an area indicator.

[0020] As a further aspect of the present invention, the construction of the unified stiffness constraint model includes: defining the equivalent short-circuit capacity at the k-th common coupling point as:

[0021] ;

[0022] Among them, S ac,k For external network short-circuit capacity; S GFM,k For the installed capacity of GFM wind turbines; , , These are the short-circuit capacity conversion factors for each device; η GFM ∈(0,1] represents the proportion of support that GFM can utilize in the configuration layer;

[0023] Based on the equivalent short-circuit capacity, a unified generalized short-circuit ratio is constructed, and linear inequality stiffness constraints are constructed according to the set stiffness baseline:

[0024] ;

[0025] in, To unify the generalized short-circuit ratio, This is the set stiffness baseline.

[0026] As a further embodiment of the present invention, the capacity configuration layer uses a proxy model to approximate the mapping relationship between the multi-scenario comprehensive performance index and decision variables. The proxy model is in the form of a quadratic polynomial. The model parameters are fitted by sampling points. An initial linear programming model is constructed using the coefficients of the first-order terms of the proxy model to solve the preliminary configuration. After updating the gradient information of the proxy model at the preliminary configuration points, the final optimal capacity configuration scheme is obtained by solving the solution again.

[0027] As a further aspect of the present invention, the objective function, which aims to minimize the weighted sum of reactive power output adjustments of multiple types of equipment, is constructed based on the reactive power-voltage sensitivity relationship at the common coupling point. The objective function is:

[0028] ;

[0029] in, For action penalty weights, The reactive power output adjustment is the amount at the baseline operating point.

[0030] This invention also provides a collaborative optimization system for multiple types of active support equipment at the weak feed end, the system comprising:

[0031] The model building module is used to establish voltage-reactive droop characteristic equations for various types of active support equipment based on the control parameters of multiple types of active support equipment, and to couple the droop characteristic equations with the power flow equations of the power grid to construct a unified power flow-droop static model for describing the steady-state operating point of the system.

[0032] The stability domain evaluation module is used to construct a comprehensive performance index for characterizing the static voltage stability of the system based on the unified power flow-droop static model. The index includes at least the minimum static stability margin and the stability domain area.

[0033] The stiffness constraint construction module is used to construct a unified stiffness constraint model that takes into account the short-circuit capacity of the external network and the reactive power support capacity of the various types of active support equipment, and to generate capacity constraints for the equipment to be configured based on the stiffness constraint model.

[0034] The capacity collaborative optimization module is used at the capacity configuration layer to solve for the optimal capacity configuration scheme by taking the rated capacity of the equipment to be configured as the decision variable, minimizing the weighted difference between investment cost and stability benefit, and satisfying the capacity constraints.

[0035] The collaborative optimization module is used at the runtime layer to solve for the reactive power sharing strategy of each active support device based on the optimal capacity configuration scheme, with the goal of minimizing the weighted sum of reactive power output adjustments of multiple types of devices.

[0036] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention constructs a unified stiffness baseline constraint for gSCR that considers the short-circuit capacity of the external grid and the contributions of GFM / SVG / SC, explicitly transforming the insufficient external grid strength into the minimum capacity gap of SVG / SC. Finally, under the premise of satisfying the gSCR baseline and the hard voltage constraint, the upper layer solves the capacity configuration using a surrogate model and linear programming, while the lower layer optimizes reactive power sharing based on Q / V sensitivity and action cost, achieving coordinated improvement of static margin in the most unfavorable direction and expanding the stability domain form through configuration and operation. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0038] Figure 1 This is a schematic diagram illustrating the reactive power / voltage droop characteristics of multiple devices.

[0039] Figure 2This is a schematic diagram of the λ-θ static voltage stability domain in a single scenario.

[0040] Figure 3 The flowchart of the optimized framework provided for the embodiments of the present invention.

[0041] Figure 4 This is a schematic diagram of the weak-sending-end power grid access to the main grid provided in an embodiment of the present invention.

[0042] Figure 5 The diagrams show the λ-θ static voltage stability domain in representative scenarios (before and after configuration optimization), where (a) represents scenario A, (b) represents scenario A, and (c) represents scenario A.

[0043] Figure 6 Figure 1 shows the capacity configuration results for various types of active support equipment and a comparison chart of different strategies. Figure 2 shows the capacity configuration results, and Figure 3 shows a comparison chart of different strategies.

[0044] Figure 7 A comparison diagram of reactive actions in different scenarios.

[0045] Figure 8 A comparison chart of reactive power support ratios in different scenarios.

[0046] Figure 9 For different scenarios, λ min Comparison chart. Detailed Implementation

[0047] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0048] In this embodiment of the invention, a collaborative optimization method for multiple types of active support equipment at a weak feed end is provided, the method comprising:

[0049] Based on the control parameters of various types of active support equipment, voltage-reactive power droop characteristic equations of each type of active support equipment are established, and the droop characteristic equations are coupled with the power flow equations of the power grid to construct a unified power flow-droop static model for describing the steady-state operating point of the system.

[0050] Based on the unified power flow-droop static model, a multi-scenario comprehensive performance index for characterizing the static voltage stability of the system is constructed. The index includes at least the minimum static stability margin and the area of ​​the stability domain.

[0051] A unified stiffness constraint model is constructed that takes into account the short-circuit capacity of the external network and the reactive power support capacity of the various types of active support equipment, and capacity constraints on the equipment to be configured are generated based on the stiffness constraint model.

[0052] At the capacity configuration layer, the rated capacity of the equipment to be configured is used as the decision variable, and the goal is to minimize the weighted difference between investment cost and stability benefit. Under the condition of satisfying the capacity constraints, the optimal capacity configuration scheme is obtained.

[0053] At the operational layer, based on the optimal capacity configuration scheme, with the goal of minimizing the weighted sum of reactive power output adjustments for multiple types of equipment, the reactive power sharing strategy for each active support equipment is solved.

[0054] As a preferred embodiment of the present invention, the construction of the unified power flow-static model includes: characterizing the steady-state behavior of various types of active support equipment as voltage / reactive power droop characteristic equations at a common coupling point, and coupling them with the power flow equations to form a set of nonlinear algebraic equations describing the steady-state operating point of the system.

[0055] The voltage / reactive power droop characteristic equation is expressed as:

[0056] ;

[0057] in, Based on reactive power output; Q m,k For reactive power output of the equipment, k m,k V represents the sag slope of each device. PCC,k V is the voltage amplitude of the k-th PCC. ref,k This is the voltage reference value.

[0058] In this embodiment, a unified power flow-droop static model is established, which includes grid-following (GFL) / grid-forming (GFM) wind turbines, grid-forming static var generators (SVG), and synchronous condensers (SC). The steady-state behavior of each type of active support equipment is characterized as the voltage / reactive power droop characteristic equation at the common coupling point.

[0059] See Figure 1 As shown, under a unified benchmark (k) GFM,k =220, k svg,k =800, k SC,k =300) Comparison of voltage droop characteristics of three types of equipment: SVG has the largest slope and a higher saturation upper limit, reflecting its strong reactive power support capability when the system experiences large disturbances and undervoltage; SC has a moderate slope and is suitable as a continuously adjustable reactive power compensation resource; GFM has a gentler slope and a smaller capacity, making it more suitable for fine-tuning reactive power.

[0060] To assess the impact of the aforementioned equipment on the overall voltage distribution of the sending-end power grid, it is necessary to couple the characteristics of these equipment with the power flow equations. The problem of finding the steady-state operating point of the system is transformed into solving a system of nonlinear algebraic equations composed of network equations and equipment constraints:

[0061] ;

[0062] In the formula, the state variable Includes voltage magnitude and phase angle at each node; control variables The reactive power vector of each active support device; function This corresponds to the droop control characteristics and capacity limit constraints. By solving this system of equations, the steady-state operating point of the system, which takes into account the control characteristics of various types of equipment, can be obtained.

[0063] As a preferred embodiment of the present invention, the construction of the multi-scenario comprehensive performance index includes: determining the static voltage stability limit of each scenario along each direction for a preset set of operating scenarios and a set of disturbance directions;

[0064] A static voltage stability region is constructed based on the static voltage stability limit, and the minimum static stability margin and stability region area are calculated. Then, a comprehensive performance index is obtained by weighted summation based on scenario weights.

[0065] ;

[0066] In the formula, As scene weight, As the minimum margin indicator, A (s) This is an area indicator.

[0067] In this embodiment, the set of operating scenarios S and the set of disturbance directions θ are considered. h A directional load growth search is performed on the current-sag static model.

[0068] That is, the baseline state in a given scenario s∈S Below, we define the disturbance direction angle θ∈[0,2π) on the P–Q plane, and denote the active and reactive power increment vectors along this direction as... and Its corresponding power increment direction vector is:

[0069] ;

[0070] The corresponding parameterized load trajectory is defined as follows:

[0071] ;

[0072] In the formula, , Let λ be the increased load vector. As λ gradually increases from 0, the system experiences a gradually increasing power disturbance along direction θ until the power flow equation has no solution or the voltage reaches the safety constraint boundary. At this point, λ is the static voltage stability limit in that direction. The limit points in all directions within the interval [0, 2π) are then considered. Connecting in polar coordinates, the λ-θ static voltage stability region constituting scenario s is shown below. Figure 2 As shown.

[0073] Considering the diversity of system operation, a scene weight ω is introduced. s Construct comprehensive performance indicators for multiple scenarios:

[0074] ;

[0075] In the formula, As scene weight, As the minimum margin indicator, A (s) For area indicators, , This will serve as a key performance indicator for subsequent capacity configuration models.

[0076] As a preferred embodiment of the present invention, the construction of the unified stiffness constraint model includes: defining the equivalent short-circuit capacity S at the k-th common coupling point. eq,k for:

[0077] ;

[0078] Among them, S ac,k For external network short-circuit capacity; S GFM,k For the installed capacity of GFM wind turbines; , , These are the short-circuit capacity conversion factors for each device; η GFM ∈(0,1] represents the proportion of support that GFM can utilize in the configuration layer;

[0079] Based on the equivalent short-circuit capacity, a unified generalized short-circuit ratio is constructed, and linear inequality stiffness constraints are constructed according to the set stiffness baseline:

[0080] ;

[0081] in, To unify the generalized short-circuit ratio, This is the set stiffness baseline.

[0082] In this embodiment, to quantify the comprehensive contribution of multiple devices to the system strength, for the k-th PCC node, the equivalent short-circuit capacity S, taking into account the active support capabilities of multiple devices, is calculated. eq,k :

[0083] To ensure the system has sufficient disturbance rejection stiffness, the equivalent gSCR at each PCC must satisfy a unified lower limit. :

[0084] ;

[0085] Among them, S wind,k This represents the equivalent short-circuit capacity of the wind turbine at the kth common coupling point.

[0086] We obtain linear inequality constraints regarding the capacity to be matched for SVG and SC:

[0087] ;

[0088] In the formula, The minimum external reactive power support capacity gap required to meet stiffness requirements.

[0089] In a preferred embodiment of the present invention, the capacity configuration layer uses a proxy model to approximate the mapping relationship between the multi-scenario comprehensive performance index and decision variables. The proxy model is in the form of a quadratic polynomial. The model parameters are fitted by sampling points. An initial linear programming model is constructed using the coefficients of the first-order terms of the proxy model to solve the preliminary configuration. After updating the gradient information of the proxy model at the preliminary configuration points, the final optimal capacity configuration scheme is obtained by solving the solution again.

[0090] In this embodiment, see Figure 3 The upper-level model of the active support equipment configuration-operation optimization framework: Using the SVG and SC rated capacity at each PCC as decision variables x, Latin hypercube sampling is used to generate samples and a quadratic multinomial surrogate model is trained to predict domain indicators. A linear programming model is constructed with the objective of minimizing the weighted difference between investment cost and stability benefit. Under the constraints of unified gSCR, minimum reactive power demand of PCC, and minimum stability domain area, the optimal capacity configuration x is solved. * The objective function for capacity configuration layer optimization is:

[0091] ;

[0092] In the formula, the first term is the equipment investment cost; and These represent the iteration points of the proxy model in the r-th round. The gradient with respect to the capacity vector; and These represent the baseline minimum static stability margin and the baseline stability domain area when the system has no active support in scenario s, respectively. This is a vector of unit capacity cost coefficients for equipment, weighted by the total life-cycle investment cost and operation and maintenance cost; W λ With W AThe payoff weighting coefficients (W) for minimum static stability margin and stability region area, respectively. λ W A >0), used to balance economic efficiency and stability objectives.

[0093] The following engineering constraints must be met:

[0094] ;

[0095] ;

[0096] ;

[0097] ;

[0098] In the formula, Let α be the active power of the k-th PCC under the baseline operating condition; Q This is the empirical coefficient for the minimum reactive power ratio. This constraint ensures that the system still has basic local reactive power balance capability when the output of the GFM is limited or shut down. The baseline area without SVG / SC configuration. This is the minimum area protection coefficient.

[0099] The expression for the quadratic polynomial proxy model is:

[0100] ;

[0101] In the formula, , The perturbation directions θ at the lower edge of scene s are respectively h The predicted values ​​of the static voltage stability limit and the predicted area of ​​the stability region; x⊙x is the eigenvector containing only the squared terms of a single variable. The coefficient of the linear term, The coefficients of the quadratic term have the same dimension as x.

[0102] Minimum static voltage stability margin in scenario s It is defined as the minimum value of the predicted value for each discrete disturbance direction. This is a preset set of discrete perturbation directions.

[0103] The model employs offline sampling and fitting, online LP solution, iteratively updates the objective function using the first-order gradient information of the surrogate model, and verifies the key directions through continuous power flow, as detailed below:

[0104] Benchmark evaluation: For each scenario s, in the absence of configuration ( ) Calculation benchmark and , serving as normalized weights and constraint benchmarks.

[0105] Sampling and Fitting: Within the feasible capacity region, N sets of samples are generated using Latin hypercube sampling (LHS) combined with a boundary enhancement strategy. Calling a unified static model to calculate the real and The quadratic surrogate model described in Section 2.4 is fitted, and the intercept term is corrected by lower quantiles to construct a conservative lower bound. .

[0106] Round 1 LP: Ignore quadratic terms and use the coefficients of the linear terms from the surrogate model. Establish the initial linear programming model described in Section 3.2 and solve it to obtain the initial configuration. .

[0107] Round 2 LP: In Calculate gradient Update the coefficients of the linear terms in the objective function and constraints, and resolve the LP to obtain the final configuration. (If the second round of LP is not feasible, then return to the first round of LP).

[0108] Back-substitution verification: Substitute x * Substitute the complete λ-θ scan and power flow-droop iteration procedures to recalculate the real-world scenarios for each condition. and It can also optionally call CPF to verify the results in key directions to ensure that the final solution meets gSCR and voltage hard constraints.

[0109] As a preferred embodiment of the present invention, see [link to previous document]. Figure 3 Actively support equipment configuration – operation optimization framework lower-level model: in capacity configuration x * Once determined, an operation layer optimization model is constructed based on the Q / V sensitivity matrix of the PCC point, with the goal of minimizing the "weighted sum of reactive power action costs of multiple devices", and the reactive power sharing strategies of GFM, SVG and SC are solved in each scenario.

[0110] The objective function, which aims to minimize the weighted sum of reactive power output adjustments for multiple types of equipment, is constructed based on the reactive power-voltage sensitivity relationship at the common coupling point. The objective function is:

[0111] ;

[0112] in, For action penalty weights, The reactive power output adjustment is the amount at the baseline operating point.

[0113] In this embodiment, the runtime optimization must meet the following engineering constraints:

[0114] ;

[0115] ;

[0116] In the formula, This represents the optimal capacity for transmission to the upper layer. It provides power for the baseline operating conditions. S is the reference voltage for scenario s. VQ This is the local Q / V sensitivity matrix at PCC.

[0117] This invention employs a λ-θ directional limit search to obtain critical points under different disturbance directions, constructs the boundary of the static voltage stability domain, and quantifies all-directional robustness using minimum margin and domain area. Furthermore, it constructs a unified gSCR stiffness baseline constraint considering the external grid short-circuit capacity and the contributions of GFM / SVG / SC, explicitly transforming insufficient external grid strength into a minimum capacity gap of SVG / SC. Finally, under the premise of satisfying the gSCR baseline and voltage hard constraints, the upper layer solves the capacity configuration using a surrogate model and linear programming, while the lower layer optimizes reactive power sharing based on Q / V sensitivity and action cost, achieving coordinated improvement in static margin in the most unfavorable direction and expanding the stability domain form through configuration and operation.

[0118] This invention also provides a collaborative optimization system for multiple types of active support equipment at the weak-feeding end, the system comprising:

[0119] The model building module is used to establish voltage-reactive droop characteristic equations for various types of active support equipment based on the control parameters of multiple types of active support equipment, and to couple the droop characteristic equations with the power flow equations of the power grid to construct a unified power flow-droop static model for describing the steady-state operating point of the system.

[0120] The stability domain evaluation module is used to construct a comprehensive performance index for characterizing the static voltage stability of the system based on the unified power flow-droop static model. The index includes at least the minimum static stability margin and the stability domain area.

[0121] The stiffness constraint construction module is used to construct a unified stiffness constraint model that takes into account the short-circuit capacity of the external network and the reactive power support capacity of the various types of active support equipment, and to generate capacity constraints for the equipment to be configured based on the stiffness constraint model.

[0122] The capacity collaborative optimization module is used at the capacity configuration layer to solve for the optimal capacity configuration scheme by taking the rated capacity of the equipment to be configured as the decision variable, minimizing the weighted difference between investment cost and stability benefit, and satisfying the capacity constraints.

[0123] The collaborative optimization module is used at the runtime layer to solve for the reactive power sharing strategy of each active support device based on the optimal capacity configuration scheme, with the goal of minimizing the weighted sum of reactive power output adjustments of multiple types of devices.

[0124] See Figure 4 A weak-sender-end simulation was constructed in the IEEE-39 node system. Buses 15, 20, 29, and 31 were selected as common coupling points for renewable energy power plants. Each PCC was connected to the aggregated wind farm and could be configured with grid-type SVG and SC. The wind turbine grid connection mode adopted a hybrid configuration of GFL / GFM, with a single unit rated capacity of 100MVA, and the proportion of grid-type wind turbines (GFM) in the power plant was 30%. The upper limit of reactive power capacity per station for both SVG and SC was set to 300MVar. By adjusting the equivalent short-circuit capacity of the external grid and related channel parameters, the initial gSCR of each PCC was made to fall within the range of 2.00–2.67, thus reflecting the typical characteristics of "weak source and grid, and insufficient reactive power support" in a high-proportion renewable energy sending-end system. The wind power installed capacity and external grid short-circuit capacity parameters of each PCC are shown in Table 1, where S wind For the short-circuit capacity of the wind farm, S ac For the short-circuit capacity of the external grid, three typical operating scenarios are set based on the power balance characteristics of the sending end: Scenario A (baseline): medium load, typical wind power output; Scenario B (high load): high load, low wind power (load-dominated unfavorable scenario); Scenario C (high wind power): relatively high wind power, medium load (wind power-dominated unfavorable scenario). This paper adopts the envelope design principle at the configuration layer, uniformly setting the lower limit of gSCR to the maximum value of the demand in each scenario (i.e., 2.45). In the operation layer optimization, the reactive power penalty weight is set to c. SVG =0.2、c GFM =0.5、c SC =3.0 The system prioritizes the rapid adjustment potential of power electronic devices to reduce mechanical losses and engineering costs caused by frequent rotation of equipment. The discrete number of λ-θ scanning direction angles is set to H=72 (interval 5°). The limiting factor λ along each direction is searched using a bisection method, with a termination threshold of 1×10⁻⁶. −4 pu.

[0125] Table 1. Rated parameters (MW / MVAR) of equipment at each point of common coupling (PCC)

[0126]

[0127] Operating Condition 1: Analysis of the Static Voltage Stability Region Distribution Characteristics (λ-θ)

[0128] See Figure 5 Using the method described in this invention, a λ-θ scan is performed on the static voltage stability domain around each PCC in three representative scenarios. Figure 5 The λ-θ statically stable domain boundaries for scenarios A-Case C are given with and without active support, and with SVG and SC configurations. Solid lines represent the boundaries after configuration, and dashed lines represent the boundaries before configuration. In Case A, the optimization scheme affects λ in each direction.min The improvement is relatively balanced, and the overall λ min The static stable area is increased by approximately 2.8%, with an improvement of approximately 0.35%. In Case B, the optimized configuration significantly pushes the λ-θ boundary further away in the direction of radial load increase, resulting in an improvement of approximately 8.8% in the stable domain area, indicating that the proposed configuration method can identify and focus on amplifying the most dangerous perturbation directions. In Case C, the baseline scheme itself has a high λ-θ value. min With a larger stability region area, the optimized λ min The increase is about 0.39%, and the area increase is about 0.6%. The configuration optimization is mainly reflected in the adjustment and balancing of some directional boundaries, rather than a large-scale overall expansion. This is consistent with the engineering intuition that the overall margin is relatively sufficient in wind power-dominated scenarios.

[0129] Operating Condition 2: Capacity Configuration Solution

[0130] Under the constraints of unified gSCR and static stability domain indices, the optimized SVG and SC capacity configurations are shown below. Figure 6 As shown in (a), PCC15 and PCC20 have relatively high initial gSCRs but are still close to or slightly below the unified lower limit, and their configurations are mainly SVG, with PCC20 having an SVG of approximately 94 Mvar. PCC15 has only a small amount of SC (approximately 8 MVar) configured to reinforce the weakest constraint. PCC29 and PCC31 have lower initial gSCRs and are mainly configured with SC, with PCC29 having an SC of approximately 234 MVar and PCC31 having an SC of approximately 300 MVar. PCC31 also has an SVG of approximately 274 MVar configured to enhance the flexibility and redundancy of voltage regulation. Overall, the total capacity of SVG is approximately 368 MVar, the total capacity of SC is approximately 542 MVar, and the total capacity is approximately 910 MVar. This result indicates that in weak channel nodes dominated by gSCR hard constraints, SC must be relied upon to provide equivalent stiffness; in relatively strong nodes, SVG is more economical and has the advantage of rapid regulation.

[0131] To evaluate the advantages of the proposed method over traditional strategies in terms of capacity, this invention constructs four comparative configuration strategies: a minimum cost scheme, a pure SVG scheme, a pure SC scheme, and an SVG / SC evenly distributed scheme. Under the premise of a consistent total capacity (909.25 MVar), the λ of each strategy is compared in three scenarios. min Indicators and results can be found in [link / reference]. Figure 6 (b) The normalized areas of the minimum cost scheme and the pure SC scheme are 1.05 and 1.04, respectively, while the normalized area of ​​the synergistic strategy in this paper reaches 1.12, which is about 7% and 8% higher than the minimum cost scheme and the pure SC scheme, respectively.

[0132] Operating Condition 3: Comparison of Reactive Power Actions Before and After Operational Layer Optimization

[0133] This invention performs coordinated optimization of voltage-reactive power for SVG, SC, and GFM wind turbines at the operational layer. Specifically, by setting action penalty weights, SVG is defined as a low-cost priority adjustment resource, and SC is defined as a high-cost suppressed resource, guiding the system to prioritize the rapid adjustment potential of SVG / GFM while minimizing the frequency of SC actions. The reactive power actions, reactive power support ratio, and λ are presented in three representative scenarios. min See comparison results Figures 7 to 9 As shown. Figure 7 As shown, the total reactive power for different cases is approximately 260, 244, and 264 MVar, respectively. Although c SC At its highest, due to the hard constraints of weak node voltage and the limited configuration space of SVG, SC still plays a leading supporting role in some scenarios, reflecting the structural passive dominance under weak channel conditions; in the weaker high load Case B, the action amount of GFM and SVG increases relatively, indicating that the optimization tends to prioritize the marginal adjustment capability of low-cost power electronic resources. Figure 8 In the three scenarios, GFM accounts for approximately 38%–39%, SVG accounts for approximately 11%–15%, and SC accounts for approximately 46%–50%, indicating that under the combined effect of capacity and operating parameters, power electronic resources and spinning resources form a synergistic support, with SC providing strength as a safety net and GFM / SVG providing rapid and flexible compensation. Figure 9 The three representative scenarios of λ are given min Comparison results under three modes: baseline scheme, capacity configuration only, and capacity and operation co-optimization.

[0134] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A collaborative optimization method for multiple types of active support equipment at the weak feed end, characterized in that, The method includes: Based on the control parameters of various types of active support equipment, voltage-reactive power droop characteristic equations of each type of active support equipment are established, and the droop characteristic equations are coupled with the power flow equations of the power grid to construct a unified power flow-droop static model for describing the steady-state operating point of the system. Based on the unified power flow-droop static model, a multi-scenario comprehensive performance index for characterizing the static voltage stability of the system is constructed. The index includes at least the minimum static stability margin and the area of ​​the stability domain. A unified stiffness constraint model is constructed that takes into account the short-circuit capacity of the external network and the reactive power support capacity of the various types of active support equipment, and capacity constraints on the equipment to be configured are generated based on the stiffness constraint model. At the capacity configuration layer, the rated capacity of the equipment to be configured is used as the decision variable, and the goal is to minimize the weighted difference between investment cost and stability benefit. Under the condition of satisfying the capacity constraints, the optimal capacity configuration scheme is obtained. At the operational layer, based on the optimal capacity configuration scheme, the reactive power sharing strategy of each active support device is obtained by minimizing the weighted sum of reactive power output adjustments of multiple types of devices.

2. The collaborative optimization method for multiple types of active support equipment at the weak feed end according to claim 1, characterized in that, The construction of the unified power flow-static model includes: characterizing the steady-state behavior of various types of active support equipment as voltage / reactive power droop characteristic equations at a common coupling point, and coupling them with the power flow equations to form a set of nonlinear algebraic equations describing the steady-state operating point of the system.

3. The collaborative optimization method for multiple types of active support equipment at the weak feed end according to claim 2, characterized in that, The voltage / reactive power droop characteristic equation is expressed as: ； in, Based on reactive power output; Q m,k For reactive power output of the equipment, k m,k V represents the sag slope of each device. PCC,k Let V be the voltage amplitude of the k-th PCC. ref,k This is the voltage reference value.

4. The collaborative optimization method for multiple types of active support equipment at the weak feed end according to claim 1, characterized in that, The construction of the multi-scenario comprehensive performance index includes: determining the static voltage stability limit of each scenario along each direction for a preset set of operating scenarios and a set of disturbance directions; A static voltage stability region is constructed based on the static voltage stability limit, and the minimum static stability margin and stability region area are calculated. Then, a comprehensive performance index is obtained by weighted summation based on scenario weights. ； In the formula, As scene weight, As the minimum margin indicator, A (s) This is an area indicator.

5. The collaborative optimization method for multiple types of active support equipment at the weak feed end according to claim 1, characterized in that, The construction of the unified stiffness constraint model includes: defining the equivalent short-circuit capacity at the k-th common coupling point as: ； Among them, S ac,k For external network short-circuit capacity; S GFM,k For the installed capacity of GFM wind turbines; , , These are the short-circuit capacity conversion factors for each device; η GFM ∈(0,1] represents the proportion of support that GFM can utilize in the configuration layer; Based on the equivalent short-circuit capacity, a unified generalized short-circuit ratio is constructed, and linear inequality stiffness constraints are constructed according to the set stiffness baseline: ； in, To unify the generalized short-circuit ratio, This is the set stiffness baseline.

6. The collaborative optimization method for multiple types of active support equipment at the weak feed end according to claim 1, characterized in that, The capacity configuration layer uses a surrogate model to approximate the mapping relationship between the multi-scenario comprehensive performance index and decision variables. The surrogate model is in the form of a quadratic polynomial. The model parameters are fitted by sampling points. An initial linear programming model is constructed using the coefficients of the first-order terms of the surrogate model to solve the preliminary configuration. After updating the gradient information of the surrogate model at the preliminary configuration points, the final optimal capacity configuration scheme is obtained by solving the solution again.

7. The collaborative optimization method for multiple types of active support equipment at the weak feed end according to claim 1, characterized in that, The objective function, which aims to minimize the weighted sum of reactive power output adjustments for multiple types of equipment, is constructed based on the reactive power-voltage sensitivity relationship at the common coupling point. The objective function is: ； in, For action penalty weights, The reactive power output adjustment is the amount at the baseline operating point.

8. A collaborative optimization system for multiple types of active support equipment at a weak feed end, used to implement the collaborative optimization method for multiple types of active support equipment at a weak feed end as described in any one of claims 1-7, characterized in that, The system includes: The model building module is used to establish voltage-reactive droop characteristic equations for various types of active support equipment based on the control parameters of multiple types of active support equipment, and to couple the droop characteristic equations with the power flow equations of the power grid to construct a unified power flow-droop static model for describing the steady-state operating point of the system. The stability domain evaluation module is used to construct a comprehensive performance index for characterizing the static voltage stability of the system based on the unified power flow-droop static model. The index includes at least the minimum static stability margin and the stability domain area. The stiffness constraint construction module is used to construct a unified stiffness constraint model that takes into account the short-circuit capacity of the external network and the reactive power support capacity of the various types of active support equipment, and to generate capacity constraints for the equipment to be configured based on the stiffness constraint model. The capacity collaborative optimization module is used at the capacity configuration layer to solve for the optimal capacity configuration scheme by taking the rated capacity of the equipment to be configured as the decision variable, minimizing the weighted difference between investment cost and stability benefit, and satisfying the capacity constraints. The collaborative optimization module is used at the runtime layer to solve for the reactive power sharing strategy of each active support device based on the optimal capacity configuration scheme, with the goal of minimizing the weighted sum of reactive power output adjustments of multiple types of devices.

Images