A simulation method suitable for multi-type adjustable resource cooperative control comparison and verification

Abstract

The present application relates to power grid simulation data processing technical field, especially to a kind of simulation method suitable for the collaborative control contrast verification of multiple types of adjustable resources, comprising: the full-order state space model of multiple types of adjustable resources is built, and different high-order reduced-order model and low-order reduced-order model of fidelity and complexity are generated;Adaptive variable step size strategy is designed, the local truncation error is estimated in real time and combined with system dynamic characteristics trigger rule, and the simulation step is dynamically adjusted;The multiple control strategies to be verified are run in a typical daily fluctuation scenario on high-order reduced-order model, obtain the preliminary performance indicators of each strategy, and select the preferred strategy;Preferred strategy is run in different scene library on low-order reduced-order model, and the long-term statistical performance, robustness and economy of preferred strategy are obtained;For the finally selected collaborative control strategy, switch back to full-order state space model for rechecking simulation.The present application can effectively improve the verification efficiency of collaborative control strategy.

Description

Technical Field

[0001] This invention relates to the field of power grid simulation data processing technology, and in particular to a simulation method applicable to comparative verification of collaborative control of multiple types of adjustable resources. Background Technology

[0002] With the increasing penetration of new energy sources such as wind power and photovoltaics, as well as power electronic adjustable resources such as energy storage and flexible DC transmission, into power systems, the form and operating characteristics of power grids are becoming increasingly complex. Against this backdrop, achieving coordinated control of multiple types of adjustable resources is a key technical challenge for maintaining voltage stability. Numerical simulation is typically required to verify, compare, and select different coordinated control strategies. However, due to the diverse dynamic response characteristics of various types of adjustable resources, the time scales of dynamic responses differ significantly when performing voltage coordination, leading to a trade-off between model accuracy and simulation efficiency in the simulation verification of coordinated control strategies for multiple types of adjustable resources.

[0003] Currently, commonly used simulation order reduction and acceleration methods mainly include model order reduction techniques such as singular perturbation-based methods and balanced truncation methods, as well as numerical integration methods with fixed or simple variable step sizes. However, these methods are mostly designed for single-type equipment or single analysis targets, and the resulting reduced-order models are often of fixed order and single fidelity, which cannot adapt to the different requirements of model accuracy and speed in multi-stage verification such as "rapid screening" and "fine evaluation". At the same time, their step size control strategies are not deeply integrated with system dynamics and model fidelity, making it difficult to significantly improve the verification efficiency of multiple strategies and scenarios while ensuring the overall simulation accuracy. Therefore, existing technologies lack a systematic simulation acceleration solution that is phased, adaptive, and balances efficiency and accuracy for the verification of collaborative control strategies for multiple types of resources. Summary of the Invention

[0004] This invention provides a simulation method applicable to comparative verification of collaborative control of multiple types of adjustable resources, which can effectively solve the problems in the background technology.

[0005] This invention provides a simulation method for comparative verification of collaborative control of multiple types of adjustable resources, comprising the following steps: Construct a full-order state-space model for multiple types of adjustable resources, and generate high-order and low-order reduced-order models with different fidelity and complexity based on this full-order state-space model; An adaptive variable step size strategy is designed to dynamically adjust the simulation step size by estimating the local truncation error in real time and combining it with the system's dynamic characteristics triggering rules. Multiple control strategies to be verified are run on a typical daily fluctuation scenario on a high-order reduced-order model. Using adaptive variable step size, preliminary performance indicators of each strategy are obtained, and the preferred strategy is selected. The preferred strategy is then run on a low-order reduced-order model with different scenario libraries to obtain the long-term statistical performance, robustness, and economy of the preferred strategy. For the finally selected cooperative control strategy, switch back to the full-order state-space model for verification simulation to confirm the reliability of the reduced-order model verification conclusion.

[0006] Furthermore, the construction of the full-order state-space model is specifically as follows: Establish a small-signal full-order state-space model that includes at least photovoltaic, wind power, and energy storage resources. The full-order state-space model is in the following form: ; ; In the formula, The rate of change of the state; These are measured variables; For state variables; A is the input variable; B is the system matrix; C is the input matrix; D is the output matrix; and D is the feedforward matrix.

[0007] Furthermore, generating higher-order and lower-order reduced-order models with different fidelity and complexity is specifically as follows: A high-order reduced-order model is generated by using the balance truncation method based on Hankel singular values. The high-order reduced-order model achieves order reduction by retaining the dominant dynamic mode that has the greatest impact on input and output behavior in the full-order state-space model and discarding the secondary modes. The singular perturbation theory is used to eliminate or simplify the ultrafast dynamic variables as fast variables in the high-order reduced model, and a low-order reduced model is constructed that retains only the slow dynamics related to minute-level voltage regulation and multi-resource power allocation.

[0008] Furthermore, the specific steps for generating a higher-order reduced-order model are as follows: Define the controllable Gramian matrix of the system and observable Gramian matrix They are respectively: ; ; Establish the Lyapunov equation: ; ; Controllable Gramian matrix and observable Gramian matrix To find solutions to the Lyapunov equations; For controllable Gramian matrix and observable Gramian matrix Perform Cholesky decomposition and calculate and ; ; ; Singular value decomposition of the Cholesky product yields matrices U, Λ, and V satisfying the following relationship: ; The non-singular transformation matrix T is calculated as follows: ; Based on the nonsingular transformation matrix T, the full-order state-space model is transformed into: ; ; In the formula, , , , ; For the transformed full-order state-space model, the corresponding Gramian controllable Gramian matrix and observable Gramian matrix satisfy: ; In the formula, These are the Hankle singular values ​​of the transformed full-order state-space model; State variables are categorized into important state variables based on the magnitude of their Hankle singular values. and unimportant state variables The two parts, and the corresponding full-order state-space model, can be rewritten as: ; ; Will Setting it to zero, the resulting higher-order reduced-order model is: ; ; In the formula, ; ; ; .

[0009] Furthermore, the specific steps for generating a low-order reduced-order model are as follows: For variables in a higher-order reduced-order model whose time scale has fast variable characteristics When performing fast dynamic reduction, the remaining variables are combined into normal rate variables. Then, the multi-timescale high-order reduction model is equivalent to: ; ; In the formula, These are singular perturbation parameters; Assuming the function on the right-hand side of the above equation does not contain ε, the higher-order reduced-order model can be linearized at the equilibrium point as follows: ; In the formula, , , , f and g respectively and The derivative matrix; The lower-order reduction model is then: ; And boundary layer systems: .

[0010] Furthermore, the adaptive variable step size strategy is designed as follows: The local truncation error LTE at each step of the simulation was estimated using the embedded Runge-Kutta method; Let the current time be The current step size is The system state variables are After one-step integration using the embedded Runge-Kutta method, approximate solutions for the two next-step states are obtained: , ; Then the estimated value of the local truncation error in this step ; Pick The norm is used as the global error scalar estimate for this step. : ; Preset error tolerance upper limit and lower limit ; After each integration step using the embedded Runge-Kutta method, determine: like If the simulation step size is reduced, the integration process should be restarted. like If the result is not found, accept the result of this step and maintain the current simulation step size to proceed to the next calculation. like If the result is positive, accept the result of this step and increase the simulation step size in the next step.

[0011] Furthermore, the specific calculation of the simulation step size is as follows: ; In the formula, For the new simulation step size; This is the current simulation step size or the step size of the last failed simulation. ρ is the expected tolerance; p is the order of the numerical integration method used; ρ is the safety factor.

[0012] Furthermore, set the maximum allowable step size. and minimum allowable step size When there is During generation, determine: like > Then make = ; like ≤ ≤ Then keep The value remains unchanged; like < Then make = .

[0013] Furthermore, the specific rules for triggering system dynamic features are as follows: Set thresholds for critical system state changes: Voltage surge threshold Energy storage state of charge change threshold ; At the start of each simulation step or at intermediate monitoring points, the voltage change rate is calculated in real time. and the rate of change of energy storage charge ; If satisfied > ,or > If a predefined discrete event is detected, the adaptive variable step size strategy is immediately interrupted, and the next simulation step size is forcibly set to the preset minimum safe step size. This step size lasts for a period of time until the various rates of change of the system fall below the threshold, and then the adaptive variable step size strategy is re-enabled.

[0014] Furthermore, an extreme scenario is set up. When the optimal strategy runs on the low-order reduction model to an extreme scenario, the system switches back to the high-order reduction model for fine-grained node verification.

[0015] The technical solution of this invention can achieve the following technical effects: This method constructs a family of full-order reference models and multi-fidelity reduced-order models, and designs an adaptive variable step size strategy to shorten the verification time of the cooperative control strategy without losing key dynamic characteristics, thus effectively improving the verification efficiency of the cooperative control strategy. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating a simulation method applicable to comparative verification of collaborative control of multiple types of adjustable resources; Figure 2 A logic diagram for a simulation method applicable to comparative verification of collaborative control of multiple types of adjustable resources. Detailed Implementation

[0018] 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.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0020] This invention relates to a simulation method applicable to comparative verification of collaborative control of multiple types of adjustable resources. The method divides the verification process into two stages: "rapid strategy selection and coarse parameter tuning" and "fine strategy evaluation and robustness testing," and matches models and simulation strategies of different fidelity to each stage. It mainly includes S1 to S4 main processes, such as... Figures 1-2 As shown, the specific details of each process are as follows: S1: Construct a full-order state-space model for multiple types of adjustable resources and a family of multi-fidelity reduced-order models: We construct full-order state-space models for various types of adjustable resources, including photovoltaic, wind power, and energy storage, and establish small-signal full-order state-space models for these resources to clearly reveal the internal state of the system, serving as a benchmark truth model for verification.

[0021] Based on the full-order state-space model, model reduction techniques are applied to generate a set of reduced-order models (high-order reduced-order models and low-order reduced-order models) with different fidelity and complexity, forming a multi-fidelity reduced-order model family. The high-order reduced-order models can be used in the "rapid strategy screening and parameter coarse-tuning" stage (stage one) when comparing and screening various cooperative control strategies, while the low-order reduced-order models can be used in the "refined strategy evaluation and robustness testing" stage (stage two) when comparing and screening various cooperative control strategies.

[0022] S2: Design an adaptive variable step size strategy: Based on the high-order and low-order reduced models obtained by model reduction techniques, the local truncation error LTE at each step during simulation is estimated using the embedded Runge-Kutta method. Then, the real-time estimation of LTE is used as the primary and core basis for step size adjustment to ensure that the accuracy of numerical integration is always within the preset tolerance range.

[0023] To enable faster response under sudden power system changes, a parallel triggering rule based on the state change rate is added to the aforementioned LTE criteria. This rule has higher priority and is used to capture instantaneous changes that may lag behind LTE estimates.

[0024] S3: Phased Verification Process of Collaborative Control Strategy The comparison and selection of multiple collaborative control strategies are divided into two stages: stage one (rapid strategy selection and coarse parameter tuning) and stage two (fine strategy evaluation and robustness testing).

[0025] Phase 1 (Quick Strategy Selection and Coarse Parameter Tuning): Multiple control strategies to be validated are run on a typical daily fluctuation scenario on a "high-order reduced-order model". Using adaptive variable step size, preliminary performance indicators for each strategy in terms of steady-state voltage deviation, dynamic response time, and transient voltage boost are quickly obtained. This stage allows for the rapid elimination of obviously inferior strategies and coarse tuning of key control parameters for potential strategies.

[0026] Phase Two (Detailed Strategy Evaluation and Robustness Testing): One or more preferred strategies selected in Phase 1 are run on the "low-order reduction model" using different scenario libraries (covering different wind and solar power outputs, load levels, and fault locations). This phase primarily evaluates the long-term statistical performance, robustness, and cost-effectiveness of the strategies. For a few extreme scenarios requiring special attention, the system can be switched back to the "high-order reduction model" for fine-grained node verification.

[0027] S4: Full-order model verification: For one or more of the final selected preferred cooperative control strategies, switch back to the full-order state-space model for verification simulation. During the verification simulation, select a few harsh operating conditions to confirm the reliability of the verification conclusions of the reduced-order model.

[0028] In S1, the construction of the full-order state-space model is specifically as follows: Establish a small-signal full-order state-space model that includes at least photovoltaic, wind power, and energy storage resources. The full-order state-space model is in the following form: ; ; In the formula, The rate of change of the state; For state variables, it represents the offset of each dynamic component in the system (such as generator rotor angle, energy storage SOC, photovoltaic array voltage, etc.) relative to the steady-state operating point, and its dimension is the order of the whole model; As input variables, they represent small disturbances imposed on the system from the outside, such as changes in light intensity, load fluctuations, or minor adjustments to control commands; These are measured variables, representing measurable or controllable quantities that need to be monitored, such as node voltage amplitude and line power. A is the system matrix, which describes the dynamic coupling relationship between the state variables within the system; B is the input matrix, which describes how the input variables affect the changes in the state variables; C is the output matrix, which describes how the state variables are mapped to the output variables; D is the feedforward matrix, which describes the direct influence of the input variables on the output variables.

[0029] In S1, the generation of higher-order and lower-order reduced-order models with different fidelity and complexity is specifically as follows: A high-order reduced-order model is generated by using the balance truncation method based on Hankel singular values. The high-order reduced-order model achieves order reduction by retaining the dominant dynamic mode that has the greatest impact on input and output behavior in the full-order state-space model and discarding the secondary modes. The singular perturbation theory is used to eliminate or simplify the ultrafast dynamic variables as fast variables in the high-order reduced model, and a low-order reduced model is constructed that retains only the slow dynamics related to minute-level voltage regulation and multi-resource power allocation.

[0030] The specific steps for generating a high-order reduced-order model are as follows: A balance truncation method based on Hankle singular values ​​is adopted to appropriately reduce the order of the full-order model while retaining all dominant dynamic modes strongly correlated with reactive power and voltage. This model can accurately reflect the transient voltage characteristics under large disturbances such as short-circuit faults, and can be used for rapid preliminary verification and screening of various cooperative control strategies in stage one. The method for reducing the order of the system based on Hankle singular values ​​is as follows: If the system model is in equilibrium, the Hankle singular values ​​provide a measure of state importance. That is, the state corresponding to the largest singular value is most affected by the control input, and the output is also most affected by changes in that state. Therefore, the state corresponding to the largest singular value has the greatest impact on the system's input-output behavior. Model order reduction is achieved by calculating the Gramian matrix and the magnitude of the Hankle singular values ​​of the equilibrium system. During order reduction, the order of the reduced system depends only on the number of Hankle singular values ​​to be retained, and the more Hankle singular values ​​retained, the smaller the error between the reduced model and the full-order model.

[0031] First, the previous steps have yielded the full-order state-space model: ; .

[0032] Define the controllable Gramian matrix of the system and observable Gramian matrix They are respectively: ; .

[0033] Establish the Lyapunov equation: ; ; Controllable Gramian matrix and observable Gramian matrix To establish solutions to the Lyapunov equations.

[0034] For controllable Gramian matrix and observable Gramian matrix Perform Cholesky decomposition and calculate and ; ; .

[0035] Singular value decomposition of the Cholesky product yields matrices U, Λ, and V satisfying the following relationship: .

[0036] The non-singular transformation matrix T is calculated as follows: .

[0037] Based on the nonsingular transformation matrix T, the full-order state-space model is transformed into: ; ; In the formula, , , , .

[0038] For the transformed full-order state-space model, the corresponding Gramian controllable Gramian matrix and observable Gramian matrix satisfy: ; In the formula, represents the Hankle singular value of the transformed full-order state-space model. The larger the Hankle singular value, the more important the corresponding state variable.

[0039] State variables are categorized into important state variables based on the magnitude of their Hankle singular values. and unimportant state variables The two parts, and the corresponding full-order state-space model, can be rewritten as: ; .

[0040] Will Setting it to zero, the resulting higher-order reduced-order model is: ; ; In the formula, ; ; ; Due to the existence of the nonsingular transformation matrix T, the state variables have changed. However, the input and output of the reduced-order model remain unchanged compared to the original full-order state-space model.

[0041] The specific steps for generating a low-order reduced-order model are as follows: Building upon the higher-order reduced-order model, singular perturbation theory is further employed to explicitly distinguish between fast and slow dynamics in the model. Some ultrafast states are eliminated or simplified as fast variables, resulting in a simplified slow-dynamic model that retains only those related to minute-level voltage regulation and multi-resource power allocation. This model can be used for steady-state and quasi-steady-state performance simulations over long time scales and in numerous scenarios. The implementation process of the singular perturbation theory application is as follows: For variables in a higher-order reduced-order model whose time scale has fast variable characteristics When performing fast dynamic reduction, the remaining variables are combined into normal rate variables. Then, the multi-timescale high-order reduction model is equivalent to: ; ; In the formula, These are singular perturbation parameters.

[0042] Assuming the function on the right-hand side of the above equation does not contain ε, the higher-order reduced-order model can be linearized at the equilibrium point as follows: ; In the formula, , , , f and g respectively and The derivative matrix of .

[0043] assumed If it is not singular, then according to the singular perturbation theory, when ε is sufficiently small, the higher-order reduced-order model can be approximated by the following lower-order reduced-order model: ; And boundary layer systems: ; A lower-order reduced-order model can be used to replace a higher-order reduced-order model only when the state variables in the boundary layer system decay rapidly, the singular parameters are sufficiently small, and both the reduced-order system and the boundary system are stable.

[0044] In S2, the adaptive variable step size strategy is designed as follows: The local truncation error LTE at each step of the simulation was estimated using the embedded Runge-Kutta method; Let the current time be The current step size is The system state variables are After one-step integration using the embedded Runge-Kutta method, approximate solutions for the two next-step states are obtained: (Higher-order exact solution of system state variables). (Low-order exact solution of system state variables); Then the estimated value of the local truncation error in this step ; Pick The norm is used as the global error scalar estimate for this step. : .

[0045] Preset error tolerance upper limit and lower limit ; After each integration step using the embedded Runge-Kutta method, determine: like If the current accuracy is insufficient, reduce the simulation step size and return to re-integrate; like If the current accuracy is acceptable, then accept the result of this step and maintain the current simulation step size to proceed to the next calculation. like If the current accuracy is too conservative, then accept the result of this step and increase the simulation step size in the next step.

[0046] The specific calculation of the simulation step size is as follows: ; In the formula, For the new simulation step size; This is the current simulation step size or the step size of the last failed simulation. For expected tolerance; p is the order of the numerical integration method used; This is a safety factor used to avoid repeated recalculations or oscillations caused by slight fluctuations in error estimation.

[0047] Set the maximum allowable step size and minimum allowable step size Apply hard boundary constraints to all adjusted step sizes to prevent step size mismatch: Whenever During generation, determine: like > Then make = ; like ≤ ≤ Then keep The value remains unchanged; like < Then make = .

[0048] To enable faster response under sudden power system changes, a parallel triggering rule based on the state change rate is added to the aforementioned LTE criteria. This rule has higher priority and is used to instantly capture transient changes that may be delayed by LTE estimation. The specific system dynamic characteristic triggering rule is as follows: Set thresholds for critical system state changes: Voltage surge threshold Energy storage state of charge (SOC) abrupt change threshold At the beginning of each simulation step or at intermediate monitoring points, the voltage change rate is calculated in real time. and the rate of change of energy storage charge .

[0049] If satisfied > ,or > If a predefined discrete event is detected (such as "short circuit fault occurrence" or "large load switching"), the adaptive variable step size strategy based on LTE step size adjustment logic will be interrupted immediately, and the next simulation step size will be forcibly set to the preset minimum safe step size. This step size will last for a period of time until the various change rates of the system fall below the threshold, and then the adaptive variable step size strategy will be re-enabled.

[0050] Although this application has been described in conjunction with specific features and embodiments, it is apparent that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and accompanying drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application is intended to include such modifications and modifications.

Claims

1. A simulation method suitable for multi-type adjustable resource cooperative control comparison verification, characterized in that the steps of include: Construct a full-order state-space model for multiple types of adjustable resources, and generate high-order and low-order reduced-order models with different fidelity and complexity based on this full-order state-space model; An adaptive variable step size strategy is designed to dynamically adjust the simulation step size by estimating the local truncation error in real time and combining it with the system's dynamic characteristics triggering rules. Multiple control strategies to be verified are run on a typical daily fluctuation scenario on a high-order reduced-order model. Using adaptive variable step size, preliminary performance indicators of each strategy are obtained, and the preferred strategy is selected. The preferred strategy is then run on a low-order reduced-order model with different scenario libraries to obtain the long-term statistical performance, robustness, and economy of the preferred strategy. For the finally selected cooperative control strategy, switch back to the full-order state-space model for verification simulation to confirm the reliability of the reduced-order model verification conclusion.

2. The simulation method suitable for multi-type adjustable resource cooperative control comparison verification according to claim 1, characterized in that, The specific steps for constructing a full-order state-space model are as follows: Establish a small-signal full-order state-space model that includes at least photovoltaic, wind power, and energy storage resources. The full-order state-space model is in the following form: ； ； wherein is a rate of change of the state; is a measured variable; is a state variable; is an input variable; A is a system matrix; B is an input matrix; C is an output matrix; and D is a feedforward matrix.

3. The simulation method suitable for multi-type adjustable resource coordinated control comparison verification according to claim 2, characterized in that, The specific steps for generating higher-order and lower-order reduced-order models with different fidelity and complexity are as follows: A high-order reduced-order model is generated by using the balance truncation method based on Hankel singular values. The high-order reduced-order model achieves order reduction by retaining the dominant dynamic mode that has the greatest impact on input and output behavior in the full-order state-space model and discarding the secondary modes. The singular perturbation theory is used to eliminate or simplify the ultrafast dynamic variables as fast variables in the high-order reduced model, and a low-order reduced model is constructed that retains only the slow dynamics related to minute-level voltage regulation and multi-resource power allocation.

4. The simulation method suitable for multi-type adjustable resource coordinated control comparison verification according to claim 3, characterized in that, The specific steps for generating a high-order reduced-order model are as follows: Controllable Gramian matrix defining a system and observable Gramian matrix respectively ； ； Establish the Lyapunov equation: ； ； Controllable Gramian matrix and observable Gramian matrix To find solutions to the Lyapunov equations; For controllable Gramian matrix and observable Gramian matrix Perform Cholesky decomposition and calculate and ; ； ； Singular value decomposition of the Cholesky product is performed to compute matrices U, A, and V that satisfy the following relationship: ; The non-singular transformation matrix T is calculated as: ; Based on the non-singular transformation matrix T, the full-order state-space model is transformed into: ； ； In the formulae, , , , ; For the transformed full-order state-space model, the corresponding Gramian controllability Gramian matrix and the observability Gramian matrix satisfies: ； In the formula, is the Hankle singular value of the converted full-order state space model; State variables are categorized into important state variables based on the magnitude of their Hankle singular values. and unimportant state variables The two parts, and the corresponding full-order state-space model, can be rewritten as: ； ； Set zero, the resulting high-order reduced model is: ； ； In the formulae, ; ; ; .

5. The simulation method suitable for multi-type adjustable resource coordinated control comparison verification according to claim 4, characterized in that, The specific steps for generating a low-order reduced-order model are as follows: For variables in a higher-order reduced-order model whose time scale has fast variable characteristics When performing fast dynamic reduction, the remaining variables are combined into normal rate variables. Then, the multi-timescale high-order reduction model is equivalent to: ； ； In the formula, is a singular perturbation parameter; Assuming the function on the right-hand side of the above equation does not contain ε, the higher-order reduced-order model can be linearized at the equilibrium point as follows: ； wherein , , , are the derivative matrices of f and g with respect to and respectively. The lower-order reduction model is then: ； And boundary layer systems: 。 6. The simulation method suitable for multi-type adjustable resource coordinated control comparison verification according to claim 1, characterized in that, The specific design of the adaptive variable step size strategy is as follows: The local truncation error LTE at each step of the simulation was estimated using the embedded Runge-Kutta method; Let the current time be , the current step size be , and the system state variable be ; After one step integration using the embedded Runge-Kutta method, the approximate solution of the two next states is obtained: , ; then the local truncation error estimate for this step is ; Pick The norm is used as the global error scalar estimate for this step. : ； preset upper error tolerance limit and lower limit ; After each integration step using the embedded Runge-Kutta method, determine: If then reduce the simulation step size and return to reintegration; If then accept the result of this step and proceed to the next step with the current simulation step size; If then accept the result of this step and increase the simulation step size in the next step.

7. The simulation method suitable for multi-type adjustable resource coordinated control comparison verification according to claim 6, characterized in that, The specific calculation of the simulation step size is as follows: ； In the formula, For the new simulation step size; This is the current simulation step size or the step size of the last failed simulation. ρ is the expected tolerance; p is the order of the numerical integration method used; ρ is the safety factor.

8. The simulation method suitable for multi-type adjustable resource cooperative control comparison verification according to claim 6, characterized in that, Setting a maximum allowed step size and a minimum allowed step size ; when there is generation, determining that: If then make = 0 = 1 ; If ≤ ≤ then keep the value unchanged; If < then make = .

9. The simulation method for comparative verification of collaborative control of multiple types of adjustable resources according to claim 6, characterized in that, The specific rules for triggering system dynamic features are as follows: Setting system critical state change quantity threshold: voltage jump threshold , energy storage state of charge jump threshold ; At the beginning of each simulation step or at intermediate monitoring points, the voltage rate of change is calculated in real time and the energy storage charge rate ; If satisfied > ,or > If a predefined discrete event is detected, the adaptive variable step size strategy is immediately interrupted, and the next simulation step size is forcibly set to the preset minimum safe step size. This step size lasts for a period of time until the various rates of change of the system fall below the threshold, and then the adaptive variable step size strategy is reactivated.

10. The simulation method suitable for multi-type adjustable resource coordinated control comparison verification according to claim 1, characterized in that, Set up extreme scenarios. When the optimal strategy runs on the low-order reduction model to the extreme scenario, switch back to the high-order reduction model for fine-grained node verification.

Images