Energy storage millisecond transient simulation and control method and system for low-voltage transformer area

Abstract

The invention provides an energy storage millisecond transient simulation and control method and system for a low-voltage transformer area, and the method comprises the steps: building a three-phase electromagnetic transient microscopic model of the low-voltage transformer area, and giving an equivalent operator; compressing the fast variable to a rotation coordinate, and performing synchronous transformation on the equivalent operator to obtain a rotated microscopic model; analyzing the rotated microscopic model in the short window, extracting macroscopic driving force of a fast variable by using a kernel function, and generating an intermediate state in the short window; based on the intermediate state in the short window, the slow variable is directly propelled, the fast variable under the rotating coordinate is propelled according to the macroscopic driving force, the macro step is self-adapted, and a macroscopic track is formed; and identifying disturbance on the macroscopic trajectory, quantifying voltage deviation and an equivalent power gap, mapping the voltage deviation and the equivalent power gap into an energy storage instruction, writing back the microscopic model, and performing closed-loop verification. Through collaborative design of millisecond transient simulation and rapid control, voltage drop suppression, power fluctuation compensation and user power utilization non-inductive crossing in a transformer area transient process are realized.

Description

Technical Field

[0001] This invention relates to the field of power systems and their automation technology, specifically to a method and system for millisecond-level transient simulation and control of energy storage in low-voltage distribution areas. Background Technology

[0002] Currently, transient simulations of low-voltage distribution areas largely rely on traditional electromagnetic transient simulation tools (such as PSCAD / EMTDC and MATLAB / Simulink). While these tools can ensure high accuracy in electromagnetic transient calculations, they typically employ fixed small step sizes (microseconds) for full-state solutions. In low-voltage distribution areas with multiple nodes and multiple disturbances, the computational load is extremely high, making it difficult to meet the requirements for millisecond-level near real-time simulations. Consequently, the simulation results cannot provide timely decision-making support for energy storage control. Furthermore, the models are not optimized to address the characteristics of three-phase imbalance and component parameter dispersion in low-voltage distribution areas, resulting in a gap between simulation efficiency and practical application requirements. In other words, existing electromagnetic transient simulation technology for low-voltage distribution areas faces the core problem of balancing accuracy and real-time performance. Traditional simulation tools (such as PSCAD / EMTDC) often use microsecond-level fixed small step sizes for full-state solutions to ensure the accuracy of electromagnetic transient calculations. However, low-voltage distribution areas (especially residential distribution areas) often contain multiple nodes and multiple impact loads (such as electric vehicle charging piles and distributed photovoltaic inverters). In such scenarios, the computational load of full-state small step size simulation increases exponentially, making it difficult to complete the simulation and output results within a millisecond time scale. This results in the simulation data not being able to provide timely decision support for energy storage control. At the same time, traditional simulations have not optimized models for the characteristics of three-phase imbalance and line parameter dispersion in low-voltage distribution areas, further reducing simulation efficiency and practical application adaptability.

[0003] Current low-voltage distribution area energy storage control technologies are mostly based on steady-state or slow-dynamic model designs, such as conventional droop control and PID regulation. These control strategies are mainly designed for scenarios with slow load changes. They are not fast enough to respond to millisecond-level transient disturbances (such as multiple electric vehicles closing at the same time or sudden changes in the output of distributed power sources). They cannot quickly identify and quantify voltage deviations and equivalent active power gaps during transient processes. Furthermore, after the control commands are generated, there is a lack of closed-loop verification with the micro-transient model, which can easily lead to control overshoot or insufficient compensation. It is difficult to achieve rapid voltage recovery and seamless transition of distribution areas under transient disturbances, thus affecting the power quality of users. In other words, existing low-voltage distribution area energy storage control technology is insufficient in its ability to respond to millisecond-level transient disturbances. Current mainstream control strategies (such as droop control and conventional PID regulation) are mostly based on steady-state or slow-dynamic model design, which can only cope with scenarios with slow load changes. When millisecond-level transient disturbances occur in the distribution area (such as multiple electric vehicles closing at the same time or a sudden drop in the output of distributed power sources), existing control cannot quickly identify the trigger time of the disturbance, nor can it accurately quantify the voltage deviation and equivalent active power gap of key nodes. This results in the energy storage system being unable to start the compensation mechanism in time, with delayed voltage drop suppression and untimely power fluctuation compensation, which in turn affects the normal operation of sensitive loads (such as household appliances and precision equipment) in the distribution area.

[0004] As can be seen from the above, in the existing technology, the transient simulation of low-voltage distribution areas and the energy storage control links are disconnected and lack collaborative design. After the energy storage control commands are generated, they are not verified in a closed loop through a micro-transient model. Instead, the control effect is judged based on experience or a simplified model. This may lead to parameter inaccuracies in the control commands. For example, when compensating for power gaps, overcompensation may occur, leading to voltage overshoot, or insufficient compensation may result in the voltage failing to recover to the target range. Furthermore, the mismatch between the commands and the micro-transient characteristics may cause secondary disturbances in the system, reducing the reliability and safety of energy storage control and failing to ensure the stability of the distribution area operation during transient processes. Existing technologies cannot achieve seamless transition for users under transient disturbances in low-voltage distribution areas. Due to the lag in transient simulation, slow control response, and disconnect between simulation and control, voltage fluctuations and power gaps in distribution areas last for a long time after encountering transient disturbances. Users can clearly perceive power anomalies (such as flickering lights and delayed equipment start-up and shutdown), which does not meet the core requirements for power quality optimization in low-voltage distribution areas. Especially for residential distribution areas, it cannot meet users' requirements for power comfort and continuity. There is an urgent need to achieve rapid suppression of transient disturbances and seamless transition for users through technological innovation. Summary of the Invention

[0005] To address the aforementioned shortcomings in existing technologies, this invention provides a method and system for millisecond-level transient simulation and control of energy storage in low-voltage distribution areas.

[0006] According to a first aspect of the present invention, a method for millisecond-level transient simulation and control of energy storage in low-voltage distribution areas is provided, comprising: A three-phase electromagnetic transient micro-model of a low-voltage distribution area is established to obtain fast and slow variables, and an equivalent operator is given. The fast variable is compressed into the rotating coordinates, and the equivalent operator is synchronously transformed into the rotating coordinates to obtain the rotated micro model. The rotated micro-model is analyzed within a short window, and the macroscopic driving force of fast variables is extracted using a kernel function to generate an intermediate state within the short window. Based on the intermediate state within the short window, the slow variable is directly advanced, and the fast variable under the rotating coordinate is advanced according to the macro driving force, and the adaptive macro step is formed to form a macro trajectory; Disturbances are identified on the macroscopic trajectory, and voltage deviations and equivalent power gaps are quantified. The voltage deviation and equivalent power gap are mapped to energy storage commands, and the microscopic model is written back for closed-loop verification.

[0007] Preferably, a three-phase electromagnetic transient micro-model of the low-voltage distribution area is established to obtain fast and slow variables, and equivalent operators are given, including: Establish the differential equations of the three-phase network coupled with the device, define the state variables and matrices, and give the equivalent state operator and equivalent input operator; where: The differential equation of the three-phase network is expressed as: In the formula, These are, respectively, nodal capacitor array, nodal parallel admittance array, line inductance array, line resistance array, and topology correlation array; This refers to the three-phase node voltage. This refers to the three-phase line current. The voltage at the grid connection point. The current at the grid connection point constitutes the state. For fast variables, Inject current into the three phases. For the slow variable vectors of the device and control, As required by external expectations or given by higher levels, This represents the functional relationship between the device port current and the grid connection point voltage, slow variables, and commands. The equivalent state operator and the equivalent input operator are expressed as follows: Let the mass matrix be... = State derivative stiffness matrix ,state Injection item The differential equation can then be rewritten in first-order form as follows: In the formula, It is an equivalent state operator used to reflect the linear effect of the state itself on the derivative in three-phase coordinates; For equivalent input operators, used to inject terms The contribution converted to the state derivative; This is the inverse of the mass matrix; Establish a synchronization reference angle, expressed as: In the formula, The reference phase angle used for the rotating coordinates; It is a function that estimates the synchronization angle from the slow variable and the grid connection point voltage.

[0008] Preferably, the fast variable is compressed into a rotating coordinate system, and the equivalent operator is simultaneously transformed to the rotating coordinate system to obtain the rotated microscopic model, including: Define an intermediate state such that the slow variable passes through and the fast variable is rotated, establishing a compression transformation of the fast variable; where: , , In the formula, The intermediate state vector is composed of slow variables. With the fast variable after rotation obtained by piecing together; Indicates compression mapping; For fast variables The result after applying a coordinate transformation; For the injected item The result after applying the same coordinate transformation is used to ensure that the input and the state are in the same coordinate system; This is a block diagonal rotation matrix used to perform coordinate transformations on the voltage and current subvectors, respectively. Its parameter is the reference phase angle used for the rotation coordinates. ; The equivalent operator is synchronously transformed to rotated coordinates and expressed as: , In the formula, Equivalent state operator in rotated coordinates ; This indicates that the equivalent state operator Transform from three-phase coordinate similarity to rotating coordinates; This represents the Coriolis term introduced by the angular velocity of the coordinate rotation, where for The derivative with respect to time; Equivalent input operator in rotated coordinates By equivalent input operator It is obtained through isomorphic transformation.

[0009] Preferably, the rotated microscopic model is analyzed within a short window, and the macroscopic driving force of fast variables is extracted using a kernel function to generate an intermediate state within the short window, including: Set the short window as follows: , In the formula, This refers to the length of a microscopic short-time window; The time points for convolution and evaluation are located within this short window; This serves as the anchor time between the current short window and the macro step. Within the short window, the differential equation in the rotating coordinates is solved, and the kernel function convolution average is applied to the envelope of the fast variable to obtain the macroscopic driving force of the fast variable, generating the intermediate state within the short window.

[0010] Preferably, the step of solving the differential equation in the rotated coordinates within the short window and performing kernel function convolution averaging on the envelope of the fast variable to obtain the macroscopic driving force of the fast variable and generate the intermediate state within the short window includes: By controlling the state update and defect error within the short window, and solving the differential equation in the rotating coordinates within the short window, we obtain: In the formula, For the first Defect error estimate for each sub-step; The equivalent operator in rotated coordinates; To transform the first-order differential equation in rotating coordinates Expand to The coefficient vector obtained from the order; It is the infinite norm of a vector; The current sub-step size is determined by the error threshold. The inverse solution yields, The current substep size Lth power; By performing kernel function convolution averaging on the envelope of the fast variables, the macroscopic driving force is extracted, yielding: In the formula, For at any time Macroscopic driving force estimation for the envelope of fast variables; For the short window length The kernel function obtained by scaling; For temporal convolution of the trajectory of fast variables in rotated coordinates; Solving for the intermediate state within the short window yields: In the formula, This is the intermediate state at the end of the short window; This is the right endpoint of the short window.

[0011] Preferably, based on the intermediate state within the short window, the slow variable is directly advanced, and the fast variable in the rotated coordinate is advanced according to the macroscopic driving force, and an adaptive macrostep is performed to form a macroscopic trajectory, including: The short-window results are transformed to a macroscopic scale, advancing the envelopes of slow and fast variables within the intermediate states of the short window, and adaptively adjusting the macrostep using an error index to form a macroscopic trajectory; wherein: The short window result includes the defect error estimate of the substep, the coefficient vector of the equivalent operator expansion under the rotating coordinates, and the set of calculation results of the current substep step size obtained by inverse solution of the error threshold; The process of transforming the short-window results to a macroscopic scale and performing a macroscopic forward update yields: In the formula, The envelope of the fast variable in the rotated coordinates of the next macrostep; This represents the intermediate state of the fast variable at the end of the short window; For macroscopic effective step size Multiply by the macro driving force of the previous step; For the slow variable in the next macrostep; A vector field for slow variables; To reconstruct The physical state obtained after restoring to three-phase coordinates is used for evaluation. ; Calculate the macrostep adaptive error index and obtain: In the formula, It is the zero-order error index; For the first Estimation of the change of each component within the macrostep; For the first The order of magnitude of each component; For the reason Derived local truncation error estimate; It is a vector formed by concatenating the vector field of the slow variable with the macroscopic driving force of the fast variable.

[0012] Preferably, identifying disturbances on the macroscopic trajectory and quantifying voltage deviations and equivalent power gaps includes: Using macroscopic envelope direct quantization to calculate the voltage deviation and equivalent active power gap at key nodes, we obtain: In the formula, For a moment Voltage amplitude deviation; The target voltage value; This represents the magnitude of the voltage vector in the rotating coordinate system. For a moment The equivalent active power gap; This is the envelope representation of instantaneous work in rotated coordinates, where It is a coefficient between 1 and 2; It serves as a reference when there is no disturbance. The triggering criterion for the voltage deviation and equivalent active power gap at the key nodes is expressed as follows: In the formula, To trigger the event, in the most recent short window Internal criteria Maximum certainty; The original trigger index at time t; This represents the voltage deviation and active power gap at the trigger moment.

[0013] Preferably, the voltage deviation and equivalent power gap are mapped to energy storage commands, and the microscopic model is written back for closed-loop verification, including: The voltage deviation and equivalent power gap are directly mapped to the active, reactive, and current references of energy storage to obtain energy storage control commands: In the formula, Whether the goal is achieved or not; These are the baseline values ​​of active and reactive power before the disturbance. For voltage deviation and equivalent power gap; The reactive power-voltage sensitivity coefficient; Current reference in rotating coordinates; coefficient To map the power reference to the current reference by a ratio, where It is a coefficient between 0 and 1. Take the direct-axis voltage envelope at the moment of triggering; For current amplitude constraints, This is the maximum allowable current of the device; The energy storage command is written back to the microscopic model, and closed-loop verification is performed within a short window, yielding the following result: In the formula, For writing back the current reference to the three-phase coordinate system; To reconstruct the matrix; The three components are the rotation coordinates; The three-phase network differential equations after writing back; time interval This means verifying whether the voltage and power have returned to the target bandwidth within a short window, and then pressing S4 to continue macroscopic advancement.

[0014] Preferably, the reconstruction matrix is ​​established. ,include: By establishing and reconstructing the relationship, we obtain: In the formula, The original state; The mapping is used to reconstruct the intermediate state and restore it to the three-phase coordinate system. For slow variables, Rotation matrix The reverse, Rotation matrix The parameters, For fast variables in rotating coordinates.

[0015] According to a second aspect of the present invention, a millisecond-level transient simulation and control system for energy storage in low-voltage distribution areas is provided, comprising: The micro-model construction module is used to establish a three-phase electromagnetic transient micro-model of the low-voltage distribution area, obtain fast and slow variables, and provide equivalent operators. The coordinate transformation module is used to compress the fast variable into a rotating coordinate and synchronously transform the equivalent operator into the rotating coordinate to obtain the rotated micro model. The short-window microscopic solution module is used to analyze the rotated microscopic model within a short window and extract the macroscopic driving force of fast variables using a kernel function to generate intermediate states within the short window. The macro-propulsion module is used to directly propel the slow variable based on the intermediate state within the short window, propel the fast variable under the rotating coordinate according to the macro-driving force, and adaptively step on the macro to form a macro trajectory. A deviation calculation module is used to identify disturbances on the macroscopic trajectory and quantify voltage deviation and equivalent power gap. A closed-loop verification module is used to map the voltage deviation and equivalent power gap into energy storage commands and write back the microscopic model to perform closed-loop verification.

[0016] By adopting the above technical solution, the present invention has at least one of the following beneficial effects compared with the prior art: This invention constructs a unified electromagnetic transient microscopic model of a three-phase network in a low-voltage distribution area and an energy storage grid-connected device. It adopts a two-layer alternating advancement architecture of microscopic and macroscopic layers. After dividing the fast and slow variables, it combines rotational coordinate transformation (e.g., zero-sequence-direct axis-cross axis rotational coordinate) and short-window kernel function (e.g., convolution kernel function) to extract the macroscopic driving force. This not only ensures the high accuracy of electromagnetic transient simulation, but also greatly improves the computational efficiency through macroscopic large step advancement, effectively solving the problem that traditional simulation accuracy and real-time performance are difficult to balance.

[0017] This invention addresses transient impacts or faults in distribution transformer areas by rapidly identifying and quantifying voltage deviations and equivalent active power gaps at key nodes. These deviations are mapped to active, reactive, and direct-axis / quadrature-axis current references for energy storage. Simultaneously, it incorporates device constraints for limiting and prioritizing, and achieves closed-loop verification through reconstructing and writing back the microscopic model. This ensures a rapid response from the energy storage system to suppress voltage drops and power fluctuations. The entire solution achieves deep synergy between transient simulation and energy storage control, eliminating the need for pre-level reduction and complex partitioning. After the disturbance is resolved, it can automatically and smoothly switch back to normal operating mode, ultimately achieving a seamless transition for users under transient disturbances in low-voltage distribution transformer areas. Furthermore, it has low energy storage operating costs, balancing power quality optimization and system energy efficiency. Attached Figure Description

[0018] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a flowchart illustrating the millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas, according to a preferred embodiment of the present invention.

[0019] Figure 2 This is a schematic diagram of the components of a millisecond-level transient simulation and control system for energy storage in a low-voltage distribution area according to a preferred embodiment of the present invention.

[0020] Figure 3 This is a graph showing the voltage amplitude variation over time at a key node in a low-voltage distribution area in a specific application example of the present invention.

[0021] Figure 4 The graph shows the change of active power and energy storage compensation over time in a specific application example of the present invention.

[0022] Figure 5 This is an adaptive adjustment curve of the macroscopic step size in a specific application example of the present invention. Detailed Implementation

[0023] The embodiments of the present invention are described in detail below: These embodiments are implemented based on the technical solution of the present invention, and provide detailed implementation methods and specific operation processes. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention.

[0024] Existing low-voltage distribution area transient simulation and energy storage control technologies typically suffer from technical problems such as the contradiction between the accuracy and real-time performance of low-voltage distribution area transient simulation and the lag in energy storage control response. The low-voltage distribution area transient simulation and energy storage control are disconnected and lack coordinated design, which may lead to parameter inaccuracies in control commands, reducing the reliability and safety of energy storage control, failing to ensure the stability of distribution area operation during transient processes, and also failing to achieve seamless user transitions under low-voltage distribution area transient disturbances.

[0025] To address the aforementioned issues, one embodiment of the present invention provides a millisecond-level transient simulation and control method for energy storage in low-voltage distribution substations. This method focuses on the operation control and transient simulation of energy storage systems in low-voltage distribution substations. It aims to achieve voltage drop suppression, power fluctuation compensation, and seamless power consumption for users during transient processes through the coordinated design of millisecond-level transient simulation and rapid control. This provides support for the safe and stable operation of low-voltage distribution substations and the optimization of power quality, and is particularly suitable for residential or industrial and commercial low-voltage distribution substations with impulsive loads.

[0026] Specifically, such as Figure 1 As shown in the embodiment, the millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas provided by this embodiment may include: S1. Establish a three-phase electromagnetic transient micro-model of the low-voltage distribution area, obtain fast and slow variables, and give the equivalent operator; S2, compress the fast variable into the rotating coordinates, and simultaneously transform the equivalent operator to the rotating coordinates to obtain the rotated micro model; S3 analyzes the rotated micro-model within a short window and uses a kernel function to extract the macroscopic driving force of fast variables, generating intermediate states within the short window; S4, based on the intermediate state within a short window, directly advances the slow variable and advances the fast variable under the rotating coordinate according to the macro driving force, and adaptively steps the macro step to form a macro trajectory; S5 identifies disturbances on the macroscopic trajectory and quantifies voltage deviation and equivalent power gap; S6 maps the voltage deviation and equivalent power gap into energy storage commands and writes them back to the microscopic model for closed-loop verification.

[0027] In some preferred embodiments, the above-mentioned S1 may further include: Establish the differential equations of the three-phase network coupled with the device, define the state variables and matrices, and give the equivalent state operator and equivalent input operator; where: S11, the differential equation of the three-phase network, is expressed as: In the formula, These are, respectively, nodal capacitor array, nodal parallel admittance array, line inductance array, line resistance array, and topology correlation array; This refers to the three-phase node voltage. This refers to the three-phase line current. The voltage at the grid connection point. The current at the grid connection point constitutes the state. For fast variables, Inject current into the three phases. For the slow variable vectors of the device and control (phase-locked loop integral, power and current regulator integral, state of charge). For external expectations or given by the upper layer (such as power commands). This represents the functional relationship between the device port current and the grid connection point voltage, slow variables, and commands. S12, the equivalent state operator and the equivalent input operator, are expressed as: Let the mass matrix be... = State derivative stiffness matrix ,state Injection item The differential equation can then be rewritten in first-order form as follows: In the formula, It is an equivalent state operator used to reflect the linear effect of the state itself on the derivative in three-phase coordinates; For equivalent input operators, used to inject terms The contribution converted to the state derivative; This is the inverse of the mass matrix; S13, establish the synchronization reference angle, expressed as: In the formula, The reference phase angle used for the rotating coordinates; It is a function that estimates the synchronization angle (e.g., the output angle of a phase-locked loop) from the slow variable and the grid connection point voltage.

[0028] In some preferred embodiments, the above-mentioned S2 may further include: S21, define an intermediate state that allows the slow variable to pass through and the fast variable to rotate, establishing a compression transformation of the fast variable; where: , , In the formula, The intermediate state vector is composed of slow variables. With the fast variable after rotation obtained by piecing together; Indicates compression mapping; For fast variables The result after applying a coordinate transformation; For the injected item The result after applying the same coordinate transformation is used to ensure that the input and the state are in the same coordinate system; This is a block diagonal rotation matrix used to transform the voltage and current subvectors respectively, with the following parameters: ; The equivalent operator is synchronously transformed to the rotated coordinate system and expressed as: , In the formula, Equivalent state operator in rotated coordinates ; This indicates that the equivalent state operator Transform from three-phase coordinate similarity to rotating coordinates; This represents the Coriolis term introduced by the angular velocity of the coordinate rotation, where for The derivative with respect to time; Equivalent input operator in rotated coordinates By equivalent input operator It is obtained through isomorphic transformation.

[0029] In some preferred embodiments, the above-mentioned S3 may further include: S31, set the short window as follows: , In the formula, This refers to the length of a microscopic short-time window; The time points for convolution and evaluation are located within this short window; This is the anchor time between the current short window (micro window) and the macro step; S32 solves the differential equation in the rotating coordinates within the short window and performs kernel function convolution averaging on the envelope of the fast variable to obtain the macroscopic driving force and generate the intermediate state within the short window.

[0030] In some preferred embodiments, the above-mentioned S32 may further include: S321 controls the state update and defect error within the short window. The differential equation in the rotated coordinate system is solved within the short window, expressed as: In the formula, For the first Defect error estimate for each sub-step; The equivalent operator in rotated coordinates; To transform the first-order differential equation in rotating coordinates Expand to The coefficient vector obtained from the order; It is the infinite norm of a vector; The current sub-step size is determined by the error threshold. The inverse solution yields, The current substep size Lth power; S322, the kernel function convolution average of the fast variable envelope is used to extract the macroscopic driving force, expressed as: In the formula, For at any time Macroscopic driving force estimation for the envelope of fast variables; For the short window length The kernel function obtained by scaling; For temporal convolution of the trajectory of fast variables in rotated coordinates; S323, solving for the intermediate state within a short window, is expressed as: In the formula, This is the intermediate state at the end of the short window; This is the right endpoint of the short window.

[0031] In S3 above, the short window A microscopic short time window (preferably 4 milliseconds) is defined, and the solution of differential equations and kernel function convolution within the short window are completed within this millisecond-level window. The millisecond-level granularity of microscopic calculation is directly defined, and the millisecond-level time scale is realized by solving the microscopic problem through the short window.

[0032] In some preferred embodiments, the above-mentioned S4 may further include: The short-window results are transformed to a macroscopic scale, advancing the envelopes of slow and fast variables within the intermediate states of the short window, and adaptively adjusting the macrostep using an error index to form a macroscopic trajectory; where: S41, the short window result includes the defect error estimate of the substep obtained from the microscopic solution of the short window, the coefficient vector of the equivalent operator expansion under the rotated coordinates, and the calculation result set of the current substep step size obtained from the inverse solution of the error threshold; the short window result is transformed to the macroscopic scale and a macroscopic forward update is performed, as follows: In the formula, The envelope of the fast variable in the rotated coordinates of the next macrostep; This represents the intermediate state of the fast variable at the end of the short window; For macroscopic effective step size Multiply by the macro driving force of the previous step; For the slow variable in the next macrostep; A vector field for slow variables; To reconstruct The physical state obtained after restoring to three-phase coordinates is used for evaluation. ; S42, the macrostep adaptive error index is calculated as follows: In the formula, It is the zero-order error index (the upper bound of the relative rate of change of each component). For the first Estimation of the change of each component within the macrostep; For the first The magnitude of each component (add 1 to prevent the denominator from being too small). For the reason Derived local truncation error estimate; It is a vector formed by concatenating the vector field of the slow variable with the macroscopic driving force of the fast variable.

[0033] In some preferred embodiments, the above-mentioned S5 may further include: S51 employs macroscopic envelope direct quantization to calculate the voltage deviation and equivalent active power gap at key nodes. The macroscopic envelope refers to the aggregation of the defect error estimate obtained from the short-window microscopic solution, the coefficient vector of the equivalent operator expansion in the rotated coordinates, and the current sub-step size obtained from the inverse solution of the error threshold. This aggregates the state change profile and error boundary covering the macroscopic step size, used to rapidly advance the simulation at a macroscopic scale while ensuring accuracy. The result is: In the formula, For a moment Voltage amplitude deviation; The target voltage value; This represents the magnitude of the voltage vector in the rotating coordinate system. For a moment The equivalent active power gap; This is the envelope representation of instantaneous work in rotated coordinates, where It is a coefficient between 1 and 2, preferably ; It serves as a reference when there is no disturbance. S52, the triggering criterion for the critical node voltage deviation and equivalent active power gap, is expressed as follows: In the formula, To trigger the event, in the most recent short window Internal criteria Maximum certainty; The original trigger index at time t; This represents the voltage deviation and active power gap at the trigger moment.

[0034] In some preferred embodiments, the above-mentioned S6 may further include: S61 directly maps the voltage deviation and equivalent power gap to the active, reactive, and current references of energy storage, thus obtaining the energy storage control command: In the formula, Whether the goal is achieved or not; These are the baseline values ​​of active and reactive power before the disturbance. For voltage deviation and equivalent power gap; The reactive power-voltage sensitivity coefficient; Current reference in rotating coordinates; coefficient To map the power reference to the current reference by a ratio, where It is a coefficient between 0 and 1, preferably , Take the direct-axis voltage envelope at the moment of triggering; For current amplitude constraints, This is the maximum allowable current of the device; S62, write the energy storage command back to the microscopic model and perform closed-loop verification within a short window, obtaining: In the formula, For writing back the current reference to the three-phase coordinate system; To reconstruct the matrix; The three components are the rotation coordinates; The three-phase network differential equations after writing back; time interval This means verifying whether the voltage and power have returned to the target bandwidth within a short window, and then pressing S4 to continue macroscopic advancement.

[0035] In some preferred embodiments, S62 above establishes a reconstruction matrix. It can also further include: By establishing and reconstructing the relationship, we obtain: In the formula, The original state; The mapping is used to reconstruct the intermediate state and restore it to the three-phase coordinate system. For slow variables, Rotation matrix The reverse, Rotation matrix The parameters, For fast variables in rotating coordinates.

[0036] Based on the same inventive concept, an embodiment of the present invention also provides a millisecond-level transient simulation and control system for energy storage in low-voltage distribution areas.

[0037] Specifically, such as Figure 2 As shown, the millisecond-level transient simulation and control system for energy storage in low-voltage distribution areas provided in this embodiment may include: The micro-model construction module is used to establish a three-phase electromagnetic transient micro-model of the low-voltage distribution area, obtain fast and slow variables, and provide equivalent operators. The coordinate transformation module is used to compress fast variables into rotating coordinates and synchronously transform the equivalent operators into rotating coordinates to obtain the rotated micro model. The short-window micro-solution module is used to analyze the rotated micro-model within a short window and extract the macroscopic driving force of fast variables using a kernel function to generate intermediate states within the short window. The macro-propulsion module is used to directly propel slow variables based on intermediate states within a short window, and propel fast variables in rotating coordinates according to macro-driving forces, while adaptively taking macro steps to form a macro-trajectory. Deviation calculation module, which is used to identify disturbances on the macroscopic trajectory and quantify voltage deviation and equivalent power gap; The closed-loop verification module is used to map voltage deviation and equivalent power gap into energy storage commands and write back the microscopic model for closed-loop verification.

[0038] The specific implementation methods of each functional module constituting the system provided in the above embodiments of the present invention are further described in detail below.

[0039] I. Microscopic Model Construction Module, used to establish a three-phase electromagnetic transient microscopic model of the transformer area and provide equivalent operators.

[0040] The three-phase network and device are coupled into a unified differential equation, the state variables and matrices that will be used repeatedly are defined, and the equivalent state operator and equivalent input operator are given. (1) Differential equations of a three-phase network Left side of the equals sign Quality matrix = With state derivative The product of these represents the constraint of the network energy storage elements (capacitors, inductors) on the rate of change of state; The time derivatives of three-phase node voltages and three-phase line currents; The first term on the right side of the equals sign Stiffness matrix With state The product of resistance, admittance, and topology represents the linear effect of these factors on the state. The second item on the right The device injects three-phase current into the network and zero inputs the line equations. These are, respectively, nodal capacitor array, nodal parallel admittance array, line inductance array, line resistance array, and topology correlation array; : Slow variable vectors of devices and controls (phase-locked loop integral, power and current regulator integral, state of charge). : External expectations or upper-level specifications (such as power commands); The functional relationship between the device port current and the grid connection point voltage, slow variables, and commands.

[0041] (2) Equivalent state operator and equivalent input operator Rewrite the expression in (1) in first-order form. ,get: Equivalent state operator, reflecting the linear effect of the state itself on the derivative in three-phase coordinates; : Equivalent input operator, injecting items The contribution is converted into the state derivative; The inverse of the mass matrix; Stiffness matrix. (3) Synchronous reference angle : The reference phase angle used for rotating coordinates; : A function that estimates the synchronization angle from slow variables and grid connection point voltage (e.g., phase-locked loop output angle).

[0042] II. Coordinate Transformation Module: This module is used to compress fast variables into zero-orthogonal rotating coordinates and to transform equivalent operators synchronously.

[0043] Construct an "intermediate state", namely "slow variable direct pass + fast variable rotation", and transform the equivalent operator from three-phase coordinates to rotating coordinates for subsequent short window solution and convolution averaging.

[0044] (1) Compression Transformation and Intermediate State Definition , , . Left side of the equals sign : Intermediate state vector, composed of slow variables With the fast variable after rotation splicing; : Compression mapping; For fast variables The result after applying a coordinate transformation; : For injected items The result after applying the same coordinate transformation ensures that the input and the state are in the same coordinate system; The block diagonal rotation matrix performs zero-direct-orthogonal transformations on the voltage and current sub-vectors, respectively, with the following parameters: .

[0045] (2) Equivalent operators in rotated coordinates The time derivative of the rotated coordinates introduces additional terms, which can be summarized as: , Equivalent state operator in rotated coordinates; Transform the equivalent state operator from three-phase coordinate similarity to rotating coordinate; The "Coriolis term" introduced by the angular velocity of coordinate rotation, where for The derivative with respect to time; The equivalent input operator in rotated coordinates, by Obtained through isomorphic transformation; (3) Reconstructing relationships (step six: writing back to micro-level applications) Left side of the equals sign Original state; Reconstruct the mapping to restore the intermediate state to the three-phase coordinates; : The inverse of the rotation matrix.

[0046] The current short window (micro-window) and the anchor time of the macro step; from here, a micro-short window of length η is opened, and fine-grained solutions and convolutional averaging are performed within this window.

[0047] The short-window micro-solution module is used for micro-solution and kernel function convolution averaging within a short window to extract macro-driving forces. The differential equations in the rotating coordinates are solved precisely within a short window, and the macroscopic driving force is obtained by convolution averaging the envelope of the fast variable; at the same time, the substep size is controlled by the defect error. (1) Setting up a short window (to prepare for the next step of convolution and short window solution) , : Length of the microscopic short time window; The time points for convolution and evaluation are located within this short window.

[0048] (2) Status update and defect error control within the short window : No. Defect error estimate for each sub-step; , Step 2 provides the equivalent operator for rotating coordinates; , :Bundle Expand to The coefficient vector obtained from the order; : Vector infinity norm; : Step size (unknown quantity, determined by the error threshold on the right-hand side of the equation) (obtained by inverse solution) (3) Kernel function convolution averaging (converts the effect of fast variables on slow envelopes into macroscopic driving forces) At any moment Macroscopic driving force estimation for the envelope of fast variables; : Determined by the length of the short window The kernel function obtained by scaling; : Temporal convolution of the trajectory of fast variables in rotated coordinates; : The convolution evaluation point set in step two.

[0049] (3) The intermediate state at the end of this step : The intermediate state at the end of the short window; The right endpoint of the short window.

[0050] The macro-propulsion module is used for macro-level large-step advancement (slow variables are directly advanced, and fast variables are advanced according to macro-level driving forces) and adaptive macro-step. The short-window results are transformed to a macro scale, and the slow and fast envelopes of the intermediate states are advanced with large step sizes. The macro step size is then adaptively adjusted using an error index.

[0051] (1) Macro-forward update : The fast variable envelope (zero-orthogonal coordinates) for the next macrostep; : The intermediate state of the fast variable at the end of the short window; Macroeconomic effective step size Multiply by the macro driving force of the previous step; : Slow variables in the next macrostep; : Vector field of slow variables; : By reconstructing The physical state obtained after restoring to three-phase coordinates is used for evaluation. .

[0052] (2) Macrostep adaptive error index Zero-order error index (upper bound of the relative rate of change of each component); : No. Estimation of the change of each component within the macrostep; : No. The magnitude of each component (add 1 to prevent the denominator from being too small). :Depend on Derived local truncation error estimate; : A vector formed by splicing the slow variable vector field with the fast variable macroscopic driving force.

[0053] The deviation calculation module is used to identify disturbances and quantify the "voltage deviation and equivalent power gap" on the macroscopic trajectory. The voltage deviation and equivalent active power gap of key nodes are directly calculated using macroscopic envelopes, and the triggering time and values ​​are given, providing input for the next step of instruction synthesis.

[0054] (1) Voltage deviation and equivalent power gap :time Voltage amplitude deviation; : Target voltage value; the square root term is the magnitude of the voltage vector in zero-orthogonal coordinates; :time The equivalent active power gap; The first item on the right of the second row: Envelope representation of instantaneous active power in zero-orthogonal coordinates; If there is no disturbance, the reference is useful.

[0055] (2) Triggering criteria and recording : Triggering time, in the most recent short window Internal criteria Maximum certainty; The original triggering index at time t; , Voltage deviation and active power gap at the trigger moment.

[0056] The closed-loop verification module is used to map the gap into energy storage commands and write back the micro-closed-loop verification. The deviation is directly mapped to the active, reactive and current references of energy storage, and the recovery effect is verified within a short window by reconstructing and writing back the three-phase layer before continuing macroscopic advancement. (1) Power and current reference synthesis (with constraints) The goal may or may not be successful; Base values ​​of active and reactive power before the disturbance; Step 5 Output; Reactive power-voltage sensitivity coefficient; Current reference in zero-direct-orthogonal coordinate system; coefficient The ratio by which the power reference is mapped to the current reference. Take the direct-axis voltage envelope at the moment of triggering; The last equation is for current amplitude constraint. This is the maximum allowable current for the device.

[0057] (2) Write back to the three-phase layer and verify the closed loop within the short window. : Write back to the current reference in the three-phase coordinate system; The reconstruction matrix given in step two; the column vectors in parentheses are zero-direct-intersecting three components; The second equation is the rewritten microscopic electromagnetic transient equation. Same as step one; Time range Verify whether the voltage and power return to the target bandwidth within a short window, and then continue macroscopically according to step four.

[0058] It should be noted that the steps in the method provided by the present invention can be implemented using the corresponding components in the system. Those skilled in the art can refer to the technical solution of the system to implement the steps of the method, and can also refer to the technical solution of the method to implement the composition of the system. That is, the embodiments in the system and the embodiments in the method can be understood as preferred examples of each other, which will not be elaborated here.

[0059] The technical solution provided by the above embodiments of the present invention will be further described in detail below with reference to a specific application example.

[0060] In this specific application example, taking a residential transformer substation with a rated voltage of 400 kVA and a low-voltage side of 0.4 kV as an example, the main feeder is approximately 450 meters long, with an equivalent line resistance of approximately 0.28 ohms and an equivalent inductance corresponding to an inductive reactance of 0.35 ohms. The substation is equipped with a substation-level energy storage device with a rated capacity of 250 kW and a storage capacity of 500 kWh. The maximum output current of the converter is limited to 1.2 times the rated value, and the initial state of charge of the energy storage is 75%. The device-side control includes a phase-locked loop, a power regulation loop, and a current regulation loop. The implementation process is executed in six steps within the same simulation and control framework: First, the three-phase node voltage, line current, and device injection current are unified with the device's slow internal state into an electromagnetic transient micro-model. The mass matrix and stiffness matrix, uniquely determined by the transformer area parameters, are given, thereby obtaining the equivalent state operator and equivalent input operator. At the same time, the synchronization reference angle is derived from the device's phase-locked loop for subsequent rotation coordinates. Next, without altering the device information, the three-phase voltage, current, and injection quantity are mapped as a whole to a zero-sequence direct-axis-quadrature axis rotating coordinate system, forming an "intermediate state" (slow variables are passed through as is, while fast variables are expressed in rotating coordinates). The equivalent state operator and equivalent input operator from the previous step are simultaneously transformed to the rotating coordinate system to maintain consistency between the equations and coordinates. The current microscopic short window length is set to 4 milliseconds, and the convolution evaluation point is taken at the center of the window. Subsequently, within this 4-millisecond short window, the rotating coordinate equations are solved using high-precision micro-processes, with a defect error threshold of 1×10⁻⁶. -4 Based on this, the substep size is adaptively adjusted. At the end of the micro-process, the kernel function convolution average is performed on the fast trajectory of the intermediate state to obtain the macro driving force vector and output the intermediate state value at the end of the short window. The fourth step is to enter the macro propulsion: the macro step size is selected as 20 milliseconds (16 milliseconds of effective macro step after removing the short window). The fast envelope is promoted by the macro driving force and the slow variable is promoted by the device's own vector field. At the same time, the next macro step is adjusted online according to the zero-order error index to balance accuracy and speed. The fifth step involves disturbance identification and quantification on the macroscopic trajectory: The scenario is set as follows: at 0.2 seconds, two 70kW electric vehicle charging piles simultaneously close, creating an equivalent active power gap of approximately 35kW. The critical node voltage drops instantaneously from 1.00 per unit to 0.92 per unit. Within a short window, the system provides the trigger moment according to the smoothed version of the weighted index, recording the voltage deviation of -8% and the 35kW power gap, and extracting the direct-axis voltage envelope of approximately 190V (phase voltage amplitude) at the trigger moment. The sixth step directly synthesizes the above deviations into energy storage commands: While maintaining the current amplitude within 1.2 times the device's upper limit, a positive active power of 35kW and reactive power are generated, with priority voltage boosting based on the voltage sensitivity coefficient (initial value of 0.1 kvar / V, automatically fine-tuned after online correction of local line impedance and voltage sensitivity), mapped to the direct axis and... The quadrature-axis current reference is sent to the current regulation loop, and the reference current is reconstructed back into the three-phase system via rotating coordinates and injected into the microscopic equations for closed-loop verification. Simulation and control results show that: 22 milliseconds are required for the voltage to rise to 0.98 per unit from triggering; the 35 kW active power gap is filled within 15 milliseconds; the voltage overshoot does not exceed 2%; the energy storage overhead caused by short-time injection is approximately 0.005 kWh; and the current peak is close to but does not exceed the upper limit. After the disturbance is cleared, the algorithm automatically reverts to the "macrostep priority, short-window sparse" timing, and the macrostep size gradually recovers to the 20-30 millisecond range. The energy storage control smoothly switches from the "seamless crossing mode" back to the daily droop and power point tracking mode. The entire process does not require pre-order reduction or complex partitioning, ultimately achieving millisecond-level suppression of impulsive loads and seamless crossing on the user side within the residential low-voltage distribution area. Figure 3 The figure shown is a graph illustrating the voltage amplitude variation over time at key nodes in the low-voltage distribution area. Figure 4 The figure shows the variation of active power and energy storage compensation over time. Figure 5 The figure shown is an adaptive adjustment curve of the macroscopic step size.

[0061] Regarding the technical solutions provided in the above embodiments of the present invention, there are alternative solutions that can achieve the same inventive purpose. In the coordinate transformation stage, the traditional dq0 rotating coordinate system or the zero-sequence-direct-intersecting-axis rotating coordinate system can be used to achieve the coordinate transformation, as long as it can be ensured that the three-phase fast variables can be compressed into the rotating coordinate system to simplify the solution of the transient process. In the macroscopic driving force extraction stage, Gaussian kernel function, rectangular kernel function or convolution kernel function can be used. The core is to ensure that the trajectory of the fast variables within the short window can be effectively convolved and averaged to reflect the macroscopic effect. In terms of macroscopic step size adaptive adjustment, the index based on the first-order local truncation error can be used instead of the zero-order error index, as long as it can be adjusted according to the simulation accuracy. The demand is dynamically adjusted to balance efficiency and accuracy. In the energy storage command generation stage, the power value required for energy storage compensation can be calculated first and then indirectly converted into a current reference, rather than directly mapped to a direct-axis / quadrature-axis current reference. As long as reasonable control commands can be generated by combining voltage deviation, power gap and device constraints (such as current upper limit and state of charge), all alternative solutions must meet the core requirements of maintaining the accuracy of electromagnetic transient simulation, achieving millisecond-level transient response, and completing simulation and control closed-loop verification. They must also ensure that the invention aims to ultimately achieve transient disturbance suppression in low-voltage distribution areas and seamless user transition. They cannot deviate from the overall technical framework of micro-modeling-fast-slow separation-macro-propulsion-control closed loop.

[0062] The millisecond-level transient simulation and control method and system for energy storage in low-voltage distribution areas provided in the above embodiments of the present invention adopts a method for constructing a unified electromagnetic transient micro-model of the three-phase network and the energy storage grid-connected device in the distribution area. Specifically, the three-phase network and the device are coupled into a unified differential equation through a mass matrix and a stiffness matrix, and equivalent state operators and equivalent input operators are defined to achieve an accurate description of the electromagnetic transient characteristics of the distribution area. An "intermediate state" construction technique combining fast and slow variable partitioning and rotating coordinate transformation is used. This technique maintains the slow variables as they are, transforms the three-phase fast variables as a whole into a rotating coordinate system, and simultaneously transforms the equivalent operators, laying the foundation for subsequent layered solutions. A macroscopic driving force extraction method combining short-window microscopic solution and kernel function convolution averaging is used. By solving the differential equation in rotating coordinates with high precision within a millisecond-level short window and averaging the trajectory of fast variables using kernel function convolution, a connection between microscopic accuracy and macroscopic efficiency is achieved. A two-layer adaptive step-size advancement technology, based on defect error and zero-order error indices ("micro-macro"), ensures maximum macroscopic step size while maintaining simulation accuracy, thus improving real-time performance. This is achieved through a quantitative identification technology for voltage deviation and equivalent active power gap at key nodes in the distribution area under transient disturbances, and a control command generation method that maps these parameters to energy storage active and reactive power and direct-axis / quadrature-axis current references, combined with constraints (current upper limit, state of charge) for limiting and prioritizing. A collaborative technology that writes energy storage control commands back to the micro-model for closed-loop verification and dynamic switching achieves deep integration of simulation and control. Ultimately, this combination of technologies enables near-real-time computing capabilities while maintaining electromagnetic transient accuracy in low-voltage distribution areas, as well as millisecond-level suppression capabilities against impacts and short-term faults, achieving seamless distribution area transitions.

[0063] Any matters not covered in the above embodiments of the present invention are well-known in the art.

[0064] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various modifications or variations within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. A millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas, characterized in that, include: A three-phase electromagnetic transient micro-model of a low-voltage distribution area is established to obtain fast and slow variables, and an equivalent operator is given. The fast variable is compressed into the rotating coordinates, and the equivalent operator is synchronously transformed into the rotating coordinates to obtain the rotated micro model. The rotated micro-model is analyzed within a short window, and the macroscopic driving force of fast variables is extracted using a kernel function to generate an intermediate state within the short window. Based on the intermediate state within the short window, the slow variable is directly advanced, and the fast variable under the rotating coordinate is advanced according to the macro driving force, and the adaptive macro step is formed to form a macro trajectory; Disturbances are identified on the macroscopic trajectory, and voltage deviations and equivalent power gaps are quantified. The voltage deviation and equivalent power gap are mapped into energy storage control commands, and the microscopic model is written back for closed-loop verification.

2. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 1, characterized in that, A three-phase electromagnetic transient microscopic model of a low-voltage distribution area is established to obtain fast and slow variables, and equivalent operators are given, including: Establish the differential equations of the three-phase network coupled with the device, define the state variables and matrices, and give the equivalent state operator and equivalent input operator; where: The differential equation of the three-phase network is expressed as: In the formula, These are, respectively, nodal capacitor array, nodal parallel admittance array, line inductance array, line resistance array, and topology correlation array; This refers to the three-phase node voltage. This refers to the three-phase line current. The voltage at the grid connection point. The grid connection point current constitutes the state. For fast variables, Inject current into the three phases. For the slow variable vector of the device and control, As required by external expectations or given by higher levels, This represents the functional relationship between the device port current and the grid connection point voltage, slow variables, and commands. The equivalent state operator and the equivalent input operator are expressed as follows: Let the mass matrix be... = State derivative stiffness matrix ,state Injection item The differential equation can then be rewritten in first-order form as follows: In the formula, It is an equivalent state operator used to reflect the linear effect of the state itself on the derivative in three-phase coordinates; For equivalent input operators, used to inject terms The contribution converted to the state derivative; This is the inverse of the mass matrix; Establish a synchronization reference angle, expressed as: In the formula, The reference phase angle used for the rotating coordinates; It is a function that estimates the synchronization angle from the slow variable and the grid connection point voltage.

3. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 1, characterized in that, The fast variable is compressed into a rotating coordinate system, and the equivalent operator is simultaneously transformed to the rotating coordinate system to obtain the rotated microscopic model, including: Define an intermediate state such that the slow variable passes through and the fast variable is rotated, establishing a compression transformation of the fast variable; where: , , In the formula, The intermediate state vector is composed of slow variables. With the fast variable after rotation obtained by piecing together; Indicates compression mapping; For fast variables The result after applying a coordinate transformation; For the injected item The result after applying the same coordinate transformation is used to ensure that the input and the state are in the same coordinate system; This is a block diagonal rotation matrix used to perform coordinate transformations on the voltage and current subvectors, respectively. Its parameter is the reference phase angle used for the rotation coordinates. ; The equivalent operator is synchronously transformed to rotated coordinates and expressed as: , In the formula, Equivalent state operator in rotated coordinates ; This indicates that the equivalent state operator Transform from three-phase coordinate similarity to rotating coordinates; This represents the Coriolis term introduced by the angular velocity of the coordinate rotation, where for The derivative with respect to time; Equivalent input operator in rotated coordinates By equivalent input operator It is obtained through isomorphic transformation.

4. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 1, characterized in that, The rotated microscopic model is analyzed within a short window, and the macroscopic driving force of fast variables is extracted using a kernel function to generate intermediate states within the short window, including: Set the short window as follows: , In the formula, This refers to the length of a microscopic short-time window; The time points for convolution and evaluation are located within this short window; This serves as the anchor time between the current short window and the macro step. Within the short window, the differential equation in the rotating coordinates is solved, and the kernel function convolution average is applied to the envelope of the fast variable to obtain the macroscopic driving force of the fast variable, generating the intermediate state within the short window.

5. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 4, characterized in that, The process of solving the differential equation in the rotated coordinates within the short window, and performing kernel function convolution averaging on the envelope of the fast variable to obtain the macroscopic driving force of the fast variable, and generating the intermediate state within the short window, includes: By controlling the state update and defect error within the short window, and solving the differential equation in the rotating coordinates within the short window, we obtain: In the formula, For the first Defect error estimate for each sub-step; The equivalent operator in rotated coordinates; To transform the first-order differential equation in rotating coordinates Expand to The coefficient vector obtained from the order; It is the infinite norm of a vector; The current sub-step size is determined by the error threshold. The inverse solution yields, current substep size Lth power; By performing kernel function convolution averaging on the envelope of the fast variables, the macroscopic driving force is extracted, yielding: In the formula, For at any time Macroscopic driving force estimation for the envelope of fast variables; For the short window length The kernel function obtained by scaling; For temporal convolution of the trajectory of fast variables in rotated coordinates; Solving for the intermediate state within the short window yields: In the formula, This is the intermediate state at the end of the short window; This is the right endpoint of the short window.

6. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 1, characterized in that, Based on the intermediate state within the short window, the slow variable is directly advanced, and the fast variable in the rotated coordinate is advanced according to the macroscopic driving force, with adaptive macrostepping to form a macroscopic trajectory, including: The short-window results are transformed to a macroscopic scale, advancing the envelopes of slow and fast variables within the intermediate states of the short window, and adaptively adjusting the macrostep using an error index to form a macroscopic trajectory; wherein: The short window result includes the defect error estimate of the substep, the coefficient vector of the equivalent operator expansion under the rotating coordinates, and the set of calculation results of the current substep step size obtained by inverse solution of the error threshold; The process of transforming the short-window results to a macroscopic scale and performing a macroscopic forward update yields: In the formula, The envelope of the fast variable in the rotated coordinates of the next macrostep; This represents the intermediate state of the fast variable at the end of the short window; For macroscopic effective step size Multiply by the macro driving force of the previous step; For the slow variable in the next macrostep; A vector field for slow variables; To refactor The physical state obtained after restoring to three-phase coordinates is used for evaluation. ; Calculate the macrostep adaptive error index and obtain: In the formula, It is the zero-order error index; For the first Estimation of the change of each component within the macrostep; For the first The order of magnitude of each component; For the reason Derived local truncation error estimate; It is a vector formed by concatenating the vector field of the slow variable with the macroscopic driving force of the fast variable.

7. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 4, characterized in that, Disturbances are identified on the macroscopic trajectory, and voltage deviations and equivalent power gaps are quantified, including: Using macroscopic envelope direct quantization to calculate the voltage deviation and equivalent active power gap at key nodes, we obtain: In the formula, For a moment Voltage amplitude deviation; The target voltage value; This represents the magnitude of the voltage vector in the rotating coordinate system. For a moment The equivalent active power gap; This is the envelope representation of instantaneous work in rotated coordinates, where It is a coefficient between 1 and 2; It serves as a reference when there is no disturbance. The triggering criterion for the voltage deviation and equivalent active power gap at the key nodes is expressed as follows: In the formula, To trigger the event, in the most recent short window Internal criteria Maximum certainty; The original trigger index at time t; This represents the voltage deviation and active power gap at the trigger moment.

8. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 7, characterized in that, The voltage deviation and equivalent power gap are mapped to energy storage control commands, and the microscopic model is written back for closed-loop verification, including: The voltage deviation and equivalent power gap are directly mapped to the active, reactive, and current references of energy storage to obtain energy storage control commands: In the formula, Whether the goal is achieved or not; These are the baseline values ​​of active and reactive power before the disturbance. For voltage deviation and equivalent power gap; The reactive power-voltage sensitivity coefficient; Current reference in rotating coordinates; coefficient To map the power reference to the current reference by a ratio, where It is a coefficient between 0 and 1. Take the direct-axis voltage envelope at the moment of triggering; For current amplitude constraints, This is the maximum allowable current of the device; The energy storage control commands are written back to the microscopic model, and closed-loop verification is performed within a short window, yielding the following results: In the formula, For writing back the current reference to the three-phase coordinate system; To reconstruct the matrix; The three components are the rotation coordinates; The three-phase network differential equations after writing back; time interval This means verifying whether the voltage and power return to the target bandwidth within a short window, and then pressing S4 to continue macroscopic advancement.

9. The millisecond-level transient simulation and control method for energy storage in low-voltage distribution areas according to claim 8, characterized in that, Establish the reconstruction matrix ,include: By establishing and reconstructing the relationship, we obtain: In the formula, The original state; The mapping is used to reconstruct the intermediate state and restore it to the three-phase coordinate system. For slow variables, Rotation matrix The reverse, Rotation matrix The parameters, For fast variables in rotating coordinates.

10. A millisecond-level transient simulation and control system for energy storage in low-voltage distribution areas, characterized in that, include: The micro-model construction module is used to establish a three-phase electromagnetic transient micro-model of the low-voltage distribution area, obtain fast and slow variables, and provide equivalent operators. The coordinate transformation module is used to compress the fast variable into a rotating coordinate and synchronously transform the equivalent operator into the rotating coordinate to obtain the rotated micro model. The short-window microscopic solution module is used to analyze the rotated microscopic model within a short window and extract the macroscopic driving force of fast variables using a kernel function to generate intermediate states within the short window. The macro-propulsion module is used to directly propel the slow variable based on the intermediate state within the short window, propel the fast variable under the rotating coordinate according to the macro-driving force, and adaptively step on the macro to form a macro trajectory. A deviation calculation module is used to identify disturbances on the macroscopic trajectory and quantify voltage deviation and equivalent power gap. A closed-loop verification module is used to map the voltage deviation and equivalent power gap into energy storage control commands and write back the microscopic model to perform closed-loop verification.

Images