Coordinated control and energy storage capacity joint optimization method for grid-forming converter cluster

Abstract

The application discloses a kind of network construction type converter cluster's cooperative control and energy storage capacity joint optimization method, it is related to network construction type converter technical field, this method includes for different topological structure network construction type converter, establish the cluster state space small signal model of diversified topological and introduce the improved Kuramoto large signal model of rotating operator and synchronous manifold;Based on small signal model and large signal model, establish the normalized cluster model considering small signal stability and large signal stability;Based on the normalized cluster model, with small signal and large signal stability problem as control target, construct network construction type converter cluster cooperative control strategy;Establish the energy storage capacity optimization model with system total cost minimization and stability maximization as target;By solving the optimization model, to realize energy storage capacity and control parameter joint optimization;This method comprehensively improves the stability of network construction type converter cluster under the scene of new energy high proportion access.

Description

Technical Field

[0001] This invention relates to the field of grid-type converter technology, and specifically to a method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster. Background Technology

[0002] As the proportion of new energy power generation continues to increase, the traditional power grid dominated by synchronous machines is gradually transforming into a new type of power system centered on power electronic converters. Grid-connected converters suffer from problems such as lack of inertia, lack of independent grid-connection capability, and susceptibility to instability in weak grids, making it difficult to meet system stability requirements. Grid-forming converters, with their voltage source characteristics, lack of phase-locked loop (PLL) requirements, and inertia and frequency and voltage regulation capabilities, have become key equipment supporting the stable operation of the new power system. Grid-forming converters operate in voltage source mode, overcoming the inherent shortcomings of grid-connected converters.

[0003] With the increasing scale and capacity of photovoltaic and energy storage power plants, power converters often need to be connected in parallel and operate in clusters. However, at present, research on grid-type converters still focuses on the modeling and stability control of individual units, with relatively insufficient research on grid-type converter clusters. Comparing the research trajectory of grid-type converters, it is not difficult to find that the research on power converters is an inevitable process from single units to clusters, and from independent operation to collaborative operation. The reasons for this are determined by the following practical needs: First, in the process of multi-unit collaborative operation, coupling of control loops, operating frequency bands and control objectives between multiple devices is likely to occur. Single-unit optimization control is difficult to model and describe such coupling relationships, and therefore cannot accurately describe the key parameters affecting the stability of the converter cluster, resulting in large model errors. Second, in the actual working conditions with a large number of mutually coupled power converters, the instability of the equipment cluster operation is likely to occur. Such situations are caused by the lack of consideration for the stability of large and small signals of the system and problems such as multi-unit coupling oscillations. Single-unit operation control can only ensure the independent and stable operation of each device, and it is difficult to ensure the stable operation of the parallel converter cluster.

[0004] In summary, existing research mainly focuses on the modeling and control of single grid-type converters, making it difficult to balance the stability of small and large signals and accurately identify key parameters affecting cluster stability. Moreover, multi-machine collaborative control strategies are insufficient, and there is a lack of coordinated optimization between control parameters and energy storage capacity, making it difficult to balance system economy and stability and ensuring the stable operation of parallel converter clusters. Summary of the Invention

[0005] To address the shortcomings of existing technologies, such as the lack of coordinated optimization between control parameters and energy storage capacity, which makes it difficult to balance system economy and stability and ensure the stable operation of parallel converter clusters, this invention proposes a method for coordinated control and joint optimization of energy storage capacity for grid-type converter clusters, thereby solving the problems existing in the prior art.

[0006] A method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster includes the following steps: For network-type converters with different topologies, a small-signal cluster state-space model with diverse topologies and an improved Kuramoto large-signal model incorporating rotation operators and synchronous manifolds are established. Based on the small-signal and large-signal models, a normalized cluster model that balances small-signal and large-signal stability is established, and uncertain time-varying parameters affecting power system stability and their stability intervals are extracted. Based on the normalized cluster model, a collaborative control strategy for a grid-type converter cluster is constructed with the stability issues of small and large signals as the control objective. This collaborative control strategy includes: constructing a group Lyapunov function to suppress multi-machine oscillations through distributed virtual damping and adaptive gain scheduling; performing real-time estimation and adaptive correction of uncertain time-varying parameters affecting power system stability based on a dynamic model and parameter estimator; and constructing a multi-objective collaborative controller to achieve coordinated optimization of frequency regulation, voltage regulation, and power quality within the defined stability range. By combining the extracted uncertain time-varying parameters and their impact on system stability, as well as the energy and power requirements of the coordinated control strategy, an energy storage capacity optimization model is established with the objectives of minimizing the total cost of the power system and maximizing its stability. By solving the optimization model, the joint optimization of energy storage capacity and control parameters can be achieved.

[0007] Furthermore, the cluster state space small-signal model of the diversified topology includes at least one of the following components: controlled object dynamics, control delay, voltage and current control, virtual impedance control, active power control, reactive power control, coordinate transformation, coupling sequence impedance, and connection network dynamics.

[0008] Furthermore, the improved Kuramoto large-signal model is expressed as: ; In the formula H gi and D gi These are inertia and droop coefficient, respectively. δ gi ∆ is the phase angle of the equivalent voltage source of the converter. p gi_ref_pu To standardize the input power variation a ij This is the per-unit work angle coefficient; The equivalent relation for phase angle rotation, utilizing the synchronous manifold on the n-dimensional torus, is expressed as: δ g1 δ g2… δ gn T= rot θ ( δ g1 ) rot θ ( δ g2)…rot θ ( δ gn) T, θ ∈[0, 2π); In the formula, rot represents the rotation operator.

[0009] Furthermore, in the distributed virtual damping and adaptive gain scheduling strategy, the adaptive gain is adjusted online according to the real-time operating status of the system to achieve resonance suppression under large signal disturbances.

[0010] Furthermore, the parameter estimator employs a Kalman filter or a particle filter to perform online estimation of time-varying parameters, including line impedance, grid voltage amplitude, and grid voltage phase angle, based on real-time measurement data.

[0011] Furthermore, the multi-objective cooperative controller includes: a single voltage loop control module or a voltage-current dual loop control module, a power controller designed for converters with DC side voltage sources or capacitors respectively, and control logic for achieving coordinated optimization of frequency regulation, voltage regulation and power distribution.

[0012] Furthermore, the objective function of the energy storage capacity optimization model is to minimize the total system cost, and the constraints include: inertia constant constraint, droop coefficient constraint, upper and lower limits of energy storage capacity at each node constraint, power balance constraint, and frequency change rate constraint.

[0013] Furthermore, a hybrid strategy of hierarchical decomposition and heuristics is adopted to solve the optimization model, specifically including: using Benders decomposition to separate capacity decision and runtime scenario sub-problems; introducing pruning strategies and scenario cut sets to accelerate the iterative convergence of the main problem; and combining particle swarm optimization or genetic algorithms to perform approximate optimization solutions for large-scale problems.

[0014] This invention provides a method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster, which has the following advantages: This invention significantly improves the modeling accuracy and key parameter identification capability of multi-machine parallel systems by establishing a normalized cluster model that takes into account both small-signal and large-signal stability, effectively solving the problems of large stability interval errors and difficulty in adapting to diverse topologies in existing models. Based on this, and using a global collaborative control strategy and real-time parameter estimation method, it effectively suppresses multi-machine oscillations and large disturbance instability, enhancing the system's robustness and dynamic response capability under complex operating conditions. Simultaneously, by jointly optimizing control parameter requirements and energy storage capacity configuration, it significantly reduces the construction and operation costs of the energy storage system while ensuring system stability. This method forms a complete technical closed loop from modeling, control to optimization, comprehensively improving the stability, economy, and engineering applicability of grid-type converter clusters in scenarios with high proportions of new energy access, providing key technical support for the safe and reliable operation of new power systems. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the architecture of the grid-type converter cluster modeling and cooperative control method for active grid support in an embodiment of the present invention; Figure 2 This is a diagram illustrating the normalized model and stability analysis architecture of the network-type converter cluster in this embodiment of the invention. Figure 3 This is an architecture diagram of the collaborative control strategy for the operation of a network-type converter cluster in an embodiment of the present invention; Figure 4 This is a diagram illustrating the optimized energy storage capacity architecture of a grid-type converter cluster in an embodiment of the present invention. Figure 5 This is a state-space model structure diagram of a single-unit grid-type converter in an embodiment of the present invention; Figure 5 (a) is a schematic diagram of the impedance model based on the dq rotating coordinates, and (b) is a schematic diagram of the impedance model based on the positive and negative sequence stationary coordinates. Figure 6 This is a schematic diagram of the transient stability analysis of the converter in an embodiment of the present invention; Figure 6 (a) is a schematic diagram of the equal area method analysis, (b) is a schematic diagram of the phase diagram method analysis, and (c) is a schematic diagram comparing the power angle curves before and after optimization. Detailed Implementation

[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0017] This invention proposes a collaborative control and energy storage capacity optimization method for grid-connected converter clusters used for active grid support. Taking grid-connected converter clusters as the research object and improving equipment operation stability and economy as the core objectives, the research focuses on three aspects: mechanism analysis and modeling, cluster control, and equipment capacity optimization. Figure 1 As shown, it includes the following stages: 1. Precise model of network converter cluster and analysis of its coupling characteristics.

[0018] Converter modeling and stability mechanism analysis form the foundation for subsequent collaborative control strategies and provide model support for energy storage capacity optimization. However, three main challenges exist in converter modeling and stability mechanism analysis: First, grid-type converter topologies are diverse, including traditional two-level topologies and modular multi-level topologies for high-voltage, high-current scenarios. The operating modes and control models of different topologies vary significantly, making it difficult to achieve unified modeling using a single analytical approach and modeling architecture. Second, grid-type converter cluster systems exhibit strong nonlinearity, with coupling relationships between devices, control loops, and operating frequency bands, making it difficult to establish precise, accurate, and concise models. Finally, in practical engineering, the instability causes of grid-type converter clusters are diverse, including system stability issues caused by small disturbances and stability problems caused by large disturbances such as low-voltage ride-through. Existing analysis methods do not consider both large-signal and small-signal stability, making it difficult to accurately analyze the causes of system instability, quickly determine control targets, and effectively identify stable intervals.

[0019] This invention aims to combine different converter topologies, using a modular analysis approach as its core concept. It analyzes stability from two dimensions: small-signal and large-signal stability. At the cluster level, it implements a normalized model for diverse topologies and a small-signal model for cluster control that considers multi-factor coupling. It also proposes an improved Kuramoto large-signal model for clusters, incorporating complex multi-machine interconnections. This approach balances large-signal and small-signal stability analysis, summarizes system normalization modeling methods for diverse topology scenarios, identifies key parameters affecting stability and their stability ranges when system parameters fluctuate, and ultimately forms an accurate, concise, and practical cluster model and large / small-signal stability analysis method. Specifically, it includes the following: (1) Cluster normalization model of diversified network converter topologies.

[0020] This study investigates different topologies such as traditional two-level converters and MMC converters, analyzes their operating modes, control loops, and operating parameter characteristics, and proposes a normalized modeling approach and method for diverse models. This allows for the modeling of the cluster state space and the establishment of a model that includes controlled objects, time delays, voltage and current control, virtual impedance, active power (or DC voltage) control, reactive power control, coordinate system transformation, coupling sequence impedance, and connection networks.

[0021] By establishing a state-space model of a single network converter (such as...) Figure 5 (a) shows the impedance model based on the dq rotating coordinate system, and (b) shows the impedance model based on the positive and negative sequence stationary coordinate system. This solves the problems of incomplete topology coverage and low parameter accuracy in existing models. The converter is divided into four basic types: DC-AC grid-connected, DC-grid AC-grid-connected, DC-grid AC-grid-connected, and DC-AC-grid-connected. The impedance model based on the dq rotating coordinate system is improved to an impedance model based on the positive and negative sequence stationary coordinate system. State-space models for the four types are established respectively. Analysis of their zero-pole distribution diagrams shows that the model analysis results are consistent with the experimental results, indicating that the model is accurate and effective. Furthermore, a large-signal model of the converter is established, and the nonlinearity introduced by the duty cycle and the product of state variables is analyzed. The transient stability of the converter is analyzed using the equal area method and phase diagram method, such as... Figure 6 As shown, (a) is the equal area method, (b) is the phase diagram method, and (c) is a comparison of the power angle curves before and after optimization.

[0022] (2) Improved Kuramoto large-signal model for grid-type converter clusters.

[0023] This study investigates the Kuramoto large-signal cluster model, combining diverse topology characteristics, active and reactive power control decoupling, line dynamics, and other factors to realize a Kuramoto large-signal model suitable for grid converter clusters.

[0024] (3) Take into account the stability of large / small signals of the grid-type converter cluster.

[0025] By integrating small-signal and large-signal stability analysis, the key parameters affecting system stability and their safe operating range are accurately extracted, and finally a refined modeling and large / small-signal stability analysis method suitable for grid-type converter clusters is constructed.

[0026] 2. A method for controlling the operation of a network-type converter cluster based on global collaborative control.

[0027] The operation and control strategy of grid-connected converter clusters is not only a core technology for ensuring the steady-state and transient stability of the system, but also provides a scientific basis for subsequent capacity configuration and design. However, in practical applications, three severe technical challenges are faced: First, when multiple converters are connected to the grid, due to differences in loop parameters, communication and measurement delays, and line non-ideals, mutually coupled oscillations are easily generated and amplified to instability; second, the dynamic fluctuations of grid impedance, converter component characteristics, and ambient temperature with operating conditions weaken the robustness of fixed-parameter controllers; and finally, when pursuing multiple objectives such as stability, dynamic response, and harmonic factors, constraints and conflicts often arise between the indicators.

[0028] Therefore, this invention takes stability as its core objective while considering other performance indicators, and achieves cluster control of networked converters from a global perspective. It researches and develops multi-converter cooperative control strategies, and achieves highly robust control and multi-objective cooperative control through real-time estimation of control parameters. Specifically, it includes the following: (1) Cooperative control strategy for network-type converter clusters.

[0029] With global stability as the goal, a group of Lyapunov functions is constructed and distributed virtual damping and adaptive gain scheduling are adopted to achieve fast resonance suppression under large / small signals. At the same time, phase sequence impedance and parameter coupling are preserved through equivalent reduction modeling, so as to achieve high-precision cooperative control with a low-dimensional model.

[0030] (2) Real-time estimation method for control system parameters.

[0031] A model containing parameters to be estimated is established based on a dynamic model, and a Kalman or particle filter estimator is trained using simulation / experimental data. During runtime, real-time measurements are combined, and the estimator is used to adaptively correct the model parameters, thereby achieving high-precision estimation of the time-varying characteristics of the system.

[0032] (3) Multi-objective collaborative control method.

[0033] By combining single voltage loop and voltage-current dual loop control, the grid connection function and power quality of the grid-connected converter are guaranteed. On this basis, a dedicated power controller is designed for two types of converters with voltage sources and capacitors connected to the DC side respectively, so as to realize online voltage regulation and frequency regulation, and ensure the system's fast response and steady-state accuracy under multiple operating conditions.

[0034] 3. Energy storage capacity optimization method for grid-type converter clusters.

[0035] Energy storage capacity determines control parameters of a cluster system, such as inertia and droop coefficient, and is closely related to stability. However, the construction cost of energy storage equipment used in grid-type converter clusters is relatively high, making it essential to reduce energy storage capacity while ensuring stability.

[0036] Therefore, based on the analysis of the control system parameters of the grid-type converter, this invention derives the necessary energy storage capacity required for effective control. Simultaneously, by combining factors such as energy storage prices and control strategies, it achieves synergistic optimization of energy storage capacity and control strategy, thereby ensuring the stable operation of the converter cluster and the economical design of equipment capacity. Specifically, it includes the following: (1) Analysis of the allowable range of energy storage based on the requirements of converter control parameters.

[0037] To address the energy storage requirements of grid-type converter clusters, this paper combines typical operating conditions and stability analysis to evaluate the energy and power demands of control parameters such as inertia constant and power droop coefficient during short-term frequency support and power regulation. Based on this, a virtual inertia simulation based on DC link capacitors is proposed. By responding to frequency / voltage fluctuations in real time through charging and discharging, it provides equivalent inertia support and reactive power assistance for the cluster, meeting the inertia and droop performance requirements and eliminating the need for additional energy storage deployment.

[0038] (2) Capacity optimization problem model considering control parameter requirements.

[0039] An energy storage capacity optimization model based on control parameter requirements is constructed, incorporating energy / power requirements such as cluster inertia (inertia constant) and droop coefficient into the capacity constraints, with the energy storage capacity of each node as the decision variable. With the objectives of minimizing total cost and system stability, a dual constraint is formed by combining the demand curve and peak power demand, and power flow and dynamic response constraints are introduced to ensure that the control parameter support requirements are met under disturbance scenarios. Finally, a cost-effective energy storage configuration scheme is solved by mixed integer programming.

[0040] (3) Effective solution to the energy storage capacity optimization problem.

[0041] For the aforementioned energy storage capacity optimization model, a hierarchical decomposition and heuristic hybrid strategy can be adopted to improve solution efficiency. First, Benders decomposition is used to separate capacity decisions from sub-problems of each operating scenario, and initial solutions are obtained quickly through polynomial relaxation. Then, pruning and scenario cut sets are introduced into the main problem to accelerate iterative convergence. For large-scale or real-time requirements, a metaheuristic method combining particle swarm optimization / genetic algorithms and local search can be used to significantly shorten the computation time while ensuring accuracy, thereby achieving efficient and convergent energy storage capacity optimization solutions.

[0042] 1. Normalized model and stability analysis of network-type converter clusters, such as... Figure 2 As shown, it includes the following steps: (1) Small-signal model establishment and analysis: Fully analyze the stability of the network-type converter cluster under small disturbances, establish an improved state-space expression for modeling, and analyze its stability, including the following steps: Step 1: State-space representation modeling. Based on existing research and previous work, and considering the operational characteristics of two-level and modular multilevel topologies of network converters, the state-space representations of individual units are established and improved by category.

[0043] Step 2: Coupling Analysis. Analyze the coupling relationships between different control loops and different levels within the single machine to achieve the analysis and effective modeling of the coupling mechanism.

[0044] Step 3: Cluster Modeling. Analyze the system structure, operating conditions, and coupling relationships between devices of the network-type converter cluster, and establish a normalized cluster state-space model.

[0045] (2) System large-signal modeling: The improved Kuramoto large-signal model of the grid-type converter cluster will be gradually removed from the idealized assumptions of the existing large-signal model, including the following steps: Step 1: Study the existing large-signal model of grid-type converter clusters. The focus is on the Kuramoto large-signal model of the cluster (see the following formula):

[0046] ; In the formula H gi and D gi These are inertia and droop coefficient, respectively. δ gi ∆ is the phase angle of the equivalent voltage source of the converter. p gi_ref_pu To standardize the input power variation a ij This is the per-unit work angle coefficient.

[0047] Step 2: Improve the Kuramoto large-signal model; using the synchronous manifold on the n-dimensional torus, consider the following equivalent relationship of phase angle rotation: δ g1 δ g2… δ gn T= rot θ ( δ g1 ) rot θ ( δ g2)…rot θ ( δ gn) T, θ ∈[0, 2π); In the formula, rot represents the rotation operator; a large-signal model that meets the actual operation requirements of grid-type converter clusters is proposed, and the model is further simplified.

[0048] Step 3: Instability Source Analysis. Based on typical operating conditions, analyze the impact of large-signal instability sources (such as power angle exceeding limits, voltage collapse, and DC link voltage exceeding limits) on the system's large-signal stability.

[0049] (3) Establish a normalized model that takes into account the stability of both large and small signals and conduct instability source analysis: Based on the established model and combined with the typical operating conditions of the cluster system, summarize the instability sources and classify them according to the stability of large / small signals to lay the foundation for subsequent stability analysis.

[0050] 2. Cooperative control strategy for network-type converter cluster operation, such as... Figure 3 As shown, the specific steps include: (1) Realize basic voltage and current control and frequency and voltage regulation control Step 1: Compare single-voltage-loop and dual-voltage-current control strategies. Based on the established state-space model, design a voltage controller through closed-loop pole placement to suppress LCL filter resonance while avoiding coupling between the power loop and voltage loop. To ensure system stability both in-grid and off-grid, the filter resonant frequency should satisfy the following relationship:

[0051] ; In the formula f r1 and f r2 respectively grid connection LCL Hefei-Limited Internet LC The resonant frequency of the filter.

[0052] Step 2: Design power controllers for two types of grid-connected converters with DC-connected voltage sources and capacitors to achieve voltage / frequency regulation functions; based on previous work, design an active power controller to achieve customized inertia, droop, damping, and secondary frequency regulation functions. The transfer function form of some components is as follows: Design a reactive power controller to achieve reactive power compensation or voltage regulation functions as needed, and verify the rationality of the voltage and power controller design through experiments.

[0053] (2) Realize system information prediction Step 1: Based on the analysis results of the key parameters that dominate the stability of the cluster system in Research Content 1, and combined with typical system operating conditions, identify and summarize the key unknown parameters that dominate the stability of the system (such as line impedance), and quantitatively analyze their variation range and impact on system stability.

[0054] Step 2: Based on the information obtained during the operation of the cluster system, a method for real-time prediction of system parameters is proposed.

[0055] Step 3: Compare the proposed parameter prediction methods from multiple dimensions, including system robustness, dynamic performance, and accuracy. Verify the effectiveness of the proposed parameter prediction methods through experiments.

[0056] (3) Cluster Cooperative Controller Design Based on system parameters, an adaptive cooperative controller is designed to achieve cooperative control of the grid converter cluster system, solving the problems of multi-machine oscillation and system instability. The steps include: Step 1: Analyze the impact of time-varying parameters. Collect time-varying parameters in the system and analyze their impact on the control system design, converting them into quantitative indicators to participate in the controller design.

[0057] Step 2: Adaptive Controller Implementation. A prediction-based adaptive controller is proposed. Based on the deterministic equivalence principle, control parameters such as inertia, droop, active power command, reactive power command, and virtual impedance are adjusted in real time to reduce oscillations and solve various stability problems faced by the cluster system.

[0058] 3. Optimization of energy storage capacity for grid-type converter clusters, such as... Figure 4 As shown, the specific steps include: (1) Control requirements analysis A thorough analysis of the relationship between the control loop parameter range of the grid-type converter cluster and the capacity of the energy storage device is conducted, and the characteristics of existing energy storage devices are analyzed, including the following steps: Step 1: Energy Storage Component Characteristic Analysis. Collect data on the power characteristics, energy characteristics, and price of existing energy storage devices, and analyze to determine the characteristics and advantages of each component.

[0059] Step 2: Energy Storage Capacity Tolerance Analysis. Based on different topologies and engineering scenarios, analyze and determine the permissible range of decision variables for energy storage capacity optimization.

[0060] Step 3: Controller Parameter Analysis. Analyze the relationship between the control loop parameters of the grid-type converter and the energy storage capacity, thereby quantifying the correspondence between control parameters and energy storage capacity.

[0061] Step 4: Operational Condition Analysis. Analyze the system structure and operational conditions to determine the engineering conditions that the energy storage configuration should address.

[0062] (2) Optimization model establishment Analyze the decision variables, constraints, and optimization objectives of the optimization problem, and establish a collaborative optimization model. This includes the following steps:

[0063] Step 1: Energy Storage Component Modeling. An optimization model for the energy storage components is established by combining voltage, current, and power characteristics. The H∞ norm of the matrix is ​​used to analyze the energy storage configuration requirements of the converter cluster, laying a theoretical foundation for cluster energy storage capacity optimization. The relevant formula for the H∞ norm of matrix A being less than γ is as follows:

[0064] ; Step 2: Establish a collaborative optimization model. Based on the control loop parameter requirements, the inertia coefficient is analyzed using the following formula to determine the upper and lower limits of the decision variables. With cost and stability as optimization objectives, a real-time collaborative optimization model for the control loop and energy storage capacity is established.

[0065] The inertia coefficients are as follows: ; In the formula: H cap The equivalent inertia coefficient of the DC link capacitor. K fv These are inertia control parameters. Based on the cluster system parameters, the range of inertia coefficients that the converter cluster can simulate is quantitatively analyzed.

[0066] (3) Model Solving: Analyze the above optimization model, simplify and decouple the complex optimization problem, and propose a corresponding solution algorithm, including the following steps: Step 1: Scenario Generation; Combining historical data and based on the scenario generation method, considering the probability of occurrence of each working condition, typical operating scenarios are generated.

[0067] Step 2: Problem reduction and decoupling; analyze the nonlinear characteristics and coupling relationships in the optimization problem, use methods such as piecewise linearization to reduce the problem order, and use methods such as dual problem solving to simplify the calculation.

[0068] Step 3: Algorithm Improvement and Problem Solving; Combining existing artificial intelligence methods and intelligent problem-solving algorithms, improve the problem-solving algorithm and prove its convergence for use in solving capacity optimization problems.

[0069] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster, characterized in that, Includes the following steps: For network-type converters with different topologies, a small-signal cluster state-space model with diverse topologies and an improved Kuramoto large-signal model incorporating rotation operators and synchronous manifolds are established. Based on the small-signal and large-signal models, a normalized cluster model that balances small-signal and large-signal stability is established, and uncertain time-varying parameters affecting power system stability and their stability intervals are extracted. Based on the normalized cluster model, a collaborative control strategy for the network-type converter cluster is constructed with the stability issues of small and large signals as the control objective. The cooperative control strategy includes: constructing a group Lyapunov function and suppressing multi-machine oscillations through distributed virtual damping and adaptive gain scheduling strategies; Based on the dynamic model and parameter estimator, uncertain time-varying parameters affecting the stability of the power system are estimated and adaptively corrected in real time; a multi-objective cooperative controller is constructed to achieve coordinated optimization of frequency regulation, voltage regulation and power quality within the determined stability range; By combining the extracted uncertain time-varying parameters and their impact on system stability, as well as the energy and power requirements of the coordinated control strategy, an energy storage capacity optimization model is established with the objectives of minimizing the total cost of the power system and maximizing its stability. By solving the optimization model, the joint optimization of energy storage capacity and control parameters can be achieved.

2. The method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster according to claim 1, characterized in that, The cluster state-space small-signal model with diverse topologies includes at least one of the following components: controlled object dynamics, control delay, voltage and current control, virtual impedance control, active power control, reactive power control, coordinate transformation, coupling sequence impedance, and connection network dynamics.

3. The method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster according to claim 1, characterized in that, The improved Kuramoto large-signal model is expressed as: ； In the formula H gi and D gi These are inertia and droop coefficient, respectively. δ gi ∆ is the phase angle of the equivalent voltage source of the converter. p gi_ref_pu To standardize the input power variation a ij This is the per-unit work angle coefficient; The equivalent relation for phase angle rotation, utilizing the synchronous manifold on the n-dimensional torus, is expressed as: δ g1 δ g2… δ gn T= rot θ ( δ g1 ) rot θ ( δ g2)…rot θ ( δ gn) T, θ ∈[0,2π); In the formula, rot represents the rotation operator.

4. The method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster according to claim 1, characterized in that, In the distributed virtual damping and adaptive gain scheduling strategy, the adaptive gain is adjusted online according to the real-time operating status of the power system to achieve resonance suppression under large signal disturbances.

5. The method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster according to claim 1, characterized in that, The parameter estimator employs a Kalman filter or a particle filter to perform online estimation of time-varying parameters, including line impedance, grid voltage amplitude, and grid voltage phase angle, based on real-time measurement data.

6. The method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster according to claim 1, characterized in that, The multi-objective cooperative controller includes: a single voltage loop control module or a voltage-current dual loop control module, a power controller designed for converters with DC side voltage sources or capacitors respectively, and control logic for achieving coordinated optimization of frequency regulation, voltage regulation and power distribution.

7. The method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster according to claim 1, characterized in that, The objective function of the energy storage capacity optimization model is to minimize the total cost of the power system. The constraints include: inertia constant constraint, droop coefficient constraint, upper and lower limits of energy storage capacity at each node constraint, power balance constraint, and frequency change rate constraint.

8. The method for joint optimization of collaborative control and energy storage capacity of a grid-type converter cluster according to claim 1, characterized in that, The optimization model is solved using a hybrid strategy of hierarchical decomposition and heuristics, specifically including: using Benders decomposition to separate capacity decision and runtime scenario subproblems; introducing pruning strategies and scenario cut sets to accelerate the iterative convergence of the main problem; and combining particle swarm optimization or genetic algorithms to approximate optimization solutions for large-scale problems.

Images