A method, system, device and medium for grid-forming converter control

Abstract

The application provides a converter control method and device integrated with network construction. The method comprises the following steps: obtaining the current grid current and the current active power deviation of the converter; inputting the grid current into an observer to obtain grid voltage information; converting the active power deviation into a power synchronization compensation quantity, superimposing or subtracting the power synchronization compensation quantity and the quadrature axis component in the grid voltage information to obtain a synchronization error signal; obtaining a coordinate transformation angle based on the synchronization error signal; obtaining a predicted electrical parameter based on the coordinate transformation angle, the grid current and a preset voltage vector; obtaining an evaluation function value corresponding to each preset voltage vector according to the predicted electrical parameter, a preset instruction reference value and a multi-layer evaluation function; generating a driving signal based on the target voltage vector with the minimum evaluation function value to control the switching devices of the converter; and through the configuration of different instruction reference values, the converter is compatible and runs between the network following mode and the network construction mode.

Description

Technical Field

[0001] This invention relates to the field of power electronics and converter control technology, and in particular to a converter control method, system, equipment, and computer-readable storage medium that are integrated with the grid and applied to scenarios such as energy storage grid connection and new energy power generation. Background Technology

[0002] Different types of energy storage systems typically have different application requirements for grid-side converters. Even within the same type of energy storage, grid-side converters require multiple control modes. For example, to adapt to different grid strengths and extend the lifespan of electrochemical energy storage, electrochemical energy storage currently requires grid-side converters to operate in grid-connected power mode, grid-connected power mode, grid-connected constant bus voltage mode, grid-connected constant bus voltage mode, grid-connected current-limiting mode, and grid-connected current-limiting mode. Another example is flywheel energy storage systems, which require grid-side converters to operate in grid-connected / grid-connected constant bus voltage mode to accelerate the flywheel motor from zero speed to a specified speed range; simultaneously, flywheel energy storage systems also require grid-side converters to operate in grid-connected / grid-connected power mode to ensure normal charging and discharging operation of the flywheel within a specified speed range. Furthermore, supercapacitor energy storage systems and other hybrid energy storage systems that combine multiple energy sources all need to support these multiple control modes to meet the needs of different application scenarios.

[0003] Existing grid-side converter control schemes all involve two control strategies: grid-following control and grid-connection control. In practical applications, the switching commands issued by the upper-level dispatching system rely on a priori assessment of the grid short-circuit ratio (i.e., grid strength). Existing grid-following / connection control schemes place grid-following requirements and grid-connection requirements into different control strategies and involve online switching logic between the two strategies. Summary of the Invention

[0004] In view of the current situation where grid-following control and grid-connecting control belong to different control strategies and rely on cumbersome switching logic, the technical problem to be solved by the present invention is: how to achieve compatible operation of the converter between grid-following mode and grid-connecting mode under a unified control framework, so as to meet the needs of multiple control modes in different scenarios.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a converter control method integrated with the grid, applied to a converter, the method comprising:

[0006] Obtain the current grid current and the current active power deviation of the converter;

[0007] The grid current is input to the grid voltage observer to calculate the estimated grid voltage information;

[0008] The active power deviation is converted into a power synchronization compensation amount, and the power synchronization compensation amount is superimposed or differentially calculated with the quadrature axis component in the estimated grid voltage information to obtain a synchronization error signal.

[0009] The coordinate transformation angle is obtained by performing an integral operation based on the synchronization error signal.

[0010] Based on the coordinate transformation angle and the grid current, a discrete-time prediction model of the converter is constructed. A preset voltage vector is input into the discrete-time prediction model for iterative solution to obtain the predicted electrical parameters of the converter under the action of the preset voltage vector.

[0011] A multi-level evaluation function is constructed, and the predicted electrical parameters and preset command reference values ​​are input into the multi-level evaluation function to calculate the evaluation cost corresponding to each preset voltage vector.

[0012] The target voltage vector with the lowest evaluation cost is selected, and a drive signal is generated based on the target voltage vector to control the operation of the switching devices of the converter.

[0013] By configuring different instruction reference values, the converter can operate compatiblely between grid-connected mode and grid-connected mode.

[0014] Furthermore, after obtaining the current grid current of the converter, the method further includes: obtaining the current DC-side voltage of the converter; the method of constructing a discrete-time prediction model of the converter based on the coordinate transformation angle and the grid current includes: constructing the discrete-time prediction model based on the coordinate transformation angle, the grid current, and the DC-side voltage; wherein the predicted electrical parameters include at least one of predicted grid current, predicted active power, and predicted reactive power.

[0015] Furthermore, the grid voltage observer is a reduced-order observer; the step of inputting the grid current to the grid voltage observer and calculating the estimated grid voltage information includes: constructing a reduced-order state-space model with the grid current as a known input variable; inputting the sampled grid current to the reduced-order state-space model, iteratively solving the discretized state equations, and outputting the estimated grid voltage information; wherein, the observer bandwidth of the reduced-order state-space model is configured to be higher than the grid fundamental frequency and lower than the switching frequency of the converter.

[0016] Further, the step of converting the active power deviation into a power synchronization compensation amount includes: multiplying the active power deviation by a preset hybrid synchronization coefficient to obtain the power synchronization compensation amount; wherein, the hybrid synchronization coefficient is used to adjust the active power response speed and damping characteristics of the converter under grid frequency disturbances.

[0017] Furthermore, the converter is a multilevel converter, and the multi-layer evaluation function includes an effective voltage vector evaluation function and a redundant voltage vector evaluation function. The calculation of the evaluation value corresponding to each preset voltage vector includes: inputting the predicted electrical parameters and the command reference value into the effective voltage vector evaluation function to calculate the first evaluation value of each effective voltage vector; filtering out the candidate voltage vector with the smallest first evaluation value; if the candidate voltage vector has a corresponding redundant voltage vector, then inputting the redundant voltage vector into the redundant voltage vector evaluation function to calculate the second evaluation value, and using the redundant voltage vector with the smallest second evaluation value as the target voltage vector.

[0018] Furthermore, the effective voltage vector evaluation function includes an active power control term, a voltage control term, and a current limiting constraint term. The calculation method for the first evaluation cost value includes: calculating the active power control term based on the deviation between the predicted active power and the active power reference value, combined with a first dynamic evaluation coefficient; calculating the voltage control term based on the deviation between the estimated grid voltage information and the voltage reference value, combined with a second dynamic evaluation coefficient; calculating the current limiting constraint term based on the predicted grid current and a preset current safety boundary, combined with a third dynamic evaluation coefficient; and weighted summing the three to obtain the first evaluation cost value. Each dynamic evaluation coefficient is dynamically configured according to the steady-state or transient operating conditions of the grid, and the current limiting constraint term uses a nonlinear penalty function.

[0019] Furthermore, the redundant voltage vector evaluation function includes a switching loss optimization term and a midpoint potential balance term, which are combined with the corresponding dynamic evaluation coefficients and weighted to obtain the second evaluation value.

[0020] Furthermore, the step of configuring different command reference values ​​to enable the converter to operate compatiblely between grid-following mode and grid-connected mode includes: when receiving a grid-following mode command, configuring the command reference value as an active power reference value and a reactive power reference value; when receiving a grid-connected mode command, configuring the command reference value as a voltage amplitude reference value and a DC-side voltage reference value; and dynamically and smoothly switching the weight of the command reference value during grid voltage dips, so that the converter can seamlessly transition from power tracking state to voltage support state.

[0021] This invention also provides a converter control system integrated with the grid, comprising: a signal acquisition module, a voltage observation module, a synchronous phase-locked loop module, a model prediction module, an evaluation and optimization module, and a drive output module, for implementing the method described in any of the above embodiments.

[0022] This invention also provides a converter device, including a converter main circuit, a sampling circuit, a processor, and a memory, wherein the processor executes a computer program in the memory to implement the method described in any of the above embodiments.

[0023] This invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in any of the above embodiments.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] (1) The control framework eliminates the need to design the network switching logic. This application incorporates the network switching and network construction control requirements into the same control framework, eliminating the cumbersome logic and unstable oscillation risk caused by the switching of two independent strategies in the traditional scheme.

[0026] (2) Multiple control modes and application scenarios can be compatible simply by switching the instruction reference value. The upper-level scheduling system can flexibly switch between multiple modes such as the power control mode of the grid and the constant bus voltage control mode by simply selecting different combinations of instruction reference values.

[0027] (3) Only grid current needs to be sampled, not grid voltage, reducing hardware costs. By estimating grid voltage in real time using a reduced-order observer, the use of voltage sensors is reduced, thus lowering system hardware costs and debugging complexity.

[0028] (4) Multi-level evaluation functions reduce the number of iterations and computational load. By using two-level evaluation functions to sequentially filter the effective voltage vector and redundant voltage vector, the number of iterations for model predictive control is significantly reduced.

[0029] (5) Dynamic adaptive evaluation coefficients improve dynamic and steady-state performance. The evaluation coefficients of each control objective are dynamically and adaptively adjusted according to the system operating conditions, and the weights of each control objective under different operating conditions are optimized in real time, thereby improving the dynamic response and steady-state accuracy of the system.

[0030] (6) Real-time current constraint protection is achieved through a current limiting function. The predicted grid current is constrained in real time through a nonlinear penalty mechanism to ensure that overcurrent can be effectively limited under various control modes, thus ensuring equipment safety.

[0031] (7) Redundancy vector secondary screening achieves midpoint potential balance and switching loss optimization. By performing secondary screening of redundant voltage vectors, the switching loss is dynamically optimized based on the junction temperature of the power transistor while ensuring midpoint potential balance, thereby improving the reliability and efficiency of the system. Attached Figure Description

[0032] Figure 1This is a flowchart illustrating the converter control method integrated with the grid provided in this application embodiment;

[0033] Figure 2 This is a schematic diagram of the integrated converter control system provided in the embodiments of this application. Detailed Implementation

[0034] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0035] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0036] This application provides a grid-connected converter control scheme applicable to various power electronic conversion devices, including grid-side converters, electrochemical energy storage converters, flywheel energy storage converters, hybrid energy storage converters, and new energy grid-connected inverters. This scheme integrates grid-connected control and grid-connected control into a unified control framework. Through the coordinated operation of a hybrid synchronous orthogonal phase-locked loop and model predictive multi-objective control, it achieves seamless compatibility between grid-connected and grid-connected modes and flexible switching between multiple control modes.

[0037] Combination Figure 1 and Figure 2The converter control system provided in this application embodiment includes a signal acquisition module 100, a voltage observation module 200, a synchronous phase-locked loop (PLL) module 300, a model prediction module 400, an evaluation and optimization module 500, and a drive output module 600. The signal acquisition module 100 samples system state variables such as grid current and DC-side voltage; the voltage observation module 200 estimates grid voltage information based on grid current; the synchronous PLL module 300 uses a hybrid synchronous orthogonal PLL to obtain coordinate transformation angles; the model prediction module 400 constructs a discrete-time prediction model and iteratively predicts electrical parameters under a preset voltage vector; the evaluation and optimization module 500 filters the optimal voltage vector through multi-layer evaluation functions; and the drive output module 600 generates drive signals based on the optimal voltage vector to control the switching devices of the converter main circuit. The signal flow relationships between these modules constitute a complete closed-loop control system.

[0038] The following is combined Figure 1 The implementation process shown provides a detailed explanation of the integrated grid-connected converter control method provided in this application.

[0039] Step 1: Obtain the deviation between grid current and active power.

[0040] At the start of each control cycle, the signal acquisition module 100 first acquires the current grid current of the converter. Specifically, it samples the three-phase grid current on the grid-connected side of the converter using current sensors. The sampling frequency is consistent with or an integer multiple of the converter's switching frequency to ensure the real-time performance of the control system. After coordinate transformation, the sampled grid current in the three-phase stationary coordinate system is used to obtain the grid direct-axis current and grid quadrature-axis current in the synchronous coordinate system.

[0041] At the same time, the system acquires the current active power deviation. The active power deviation Active power reference value The difference between the current active power prediction value and the current active power feedback value reflects the current active power tracking error of the system.

[0042] In this embodiment, the signal acquisition module 100 also acquires the current DC-side voltage of the converter. DC side voltage The voltage is obtained through sampling by a voltage sensor. It is used not only for constructing subsequent discrete-time prediction models but also for calculating dynamic evaluation coefficients in the redundant voltage vector evaluation function. In systems with grid voltage sampling hardware, the grid voltage can also be sampled synchronously. In this case, the estimated grid voltage can be used as a redundant system state variable for fault-tolerant control, improving system reliability.

[0043] Step 2: Observation of grid voltage.

[0044] A key innovation of this application lies in the fact that it eliminates the need for physical sampling of the grid voltage. Instead, it uses a grid voltage reduction observer via a voltage observation module 200 to estimate the grid voltage in real time. This design effectively reduces system hardware costs and debugging complexity.

[0045] To understand the construction principle of a reduced-order observer, we first need to establish an inverter-grid model in synchronous coordinates. The electrical angular frequency of the inverter output is defined as... ,right Integrating the coordinates yields the position angle of the coordinate transformation. Position angle using coordinate transformation The inverter-grid model in the stationary coordinate system can be converted into a model in the synchronous coordinate system, as shown in equations (1) to (4): (1) (2) (3) (4)

[0046] In equations (1) to (4), and These are the direct-axis current and quadrature-axis current of the power grid in the synchronous coordinate system, respectively. It is the sum of the grid inductance and the inductance of the inverter output reactor. u and uq are the inverter output direct-axis voltage and inverter output quadrature-axis voltage in synchronous coordinate system, respectively. and These are the grid direct-axis voltage and grid quadrature-axis voltage in the synchronous coordinate system, respectively. The electrical angular frequency of the inverter output. This is the rated frequency of the power grid.

[0047] Due to the grid current and It can be directly sampled and used as a known input variable in this embodiment to construct a reduced-order state-space model with grid voltage as the state variable. Specifically, the grid voltage reduced-order observer is constructed using equations (1) to (4) as shown in equations (5) and (6): (5) (6)

[0048] In equations (5) and (6), and These are the estimated grid direct-axis voltage and the estimated grid quadrature-axis voltage, respectively. and These represent the direct-axis bandwidth and quadrature-axis bandwidth of the reduced-order observer, respectively. To achieve real-time estimation of the grid voltage, equations (5) and (6) are discretized to obtain discrete grid voltage reduced-order observers as shown in equations (7) and (8): (7) (8)

[0049] In equations (7) and (8), and These are the estimated grid direct-axis voltage and the estimated grid quadrature-axis voltage for the next time step. and These are the estimated grid direct-axis voltage and the estimated grid quadrature-axis voltage for the current frame, respectively. To control the cycle, The current frequency of the inverter output is the electrical angular frequency. and Let be the inverter output direct-axis voltage and inverter output quadrature-axis voltage in the current synchronization coordinate system, respectively, and id(k) and iq(k) be the grid direct-axis current and grid quadrature-axis current in the current synchronization coordinate system, respectively. and These are the direct-axis current and quadrature-axis current of the power grid in the previous synchronous coordinate system, respectively.

[0050] In this embodiment, the observer bandwidth (i.e., direct-axis bandwidth) of the reduced-order state-space model and cross-axis bandwidth The frequency is configured to be higher than the grid fundamental frequency and lower than the converter switching frequency. Preferably, to balance the noise suppression capability and dynamic performance of the discrete grid voltage reduction observer, its direct-axis bandwidth is... and cross-axis bandwidth The rated frequency of the power grid can be taken as the reference frequency. Five to ten times. This bandwidth configuration principle ensures that the observer can quickly track the fundamental component of the grid voltage while effectively filtering out high-frequency noise near the switching frequency.

[0051] Through iterative solution of the discretized state equations, the voltage observation module 200 can output estimated grid voltage information in real time during each control cycle, including the estimated grid direct-axis voltage. and estimated grid cross-axis voltage This provides accurate state variable inputs for subsequent phase-locked loop and model prediction.

[0052] Step 3: Power synchronization compensation and synchronization error signal.

[0053] The synchronous phase-locked module 300 is responsible for obtaining the inverter output electrical frequency. The corresponding coordinate transformation position angle θc. This embodiment uses a hybrid synchronous orthogonal phase-locked loop to achieve this function. Its core lies in introducing the power synchronization compensation amount into the input of the phase-locked loop, thereby realizing the coupling control between active power and grid frequency.

[0054] Specifically, the synchronous phase-locked module 300 first obtains the active power deviation in step one. Converted to power synchronization compensation amount The conversion method involves adjusting the active power deviation. Multiply by the preset hybrid synchronization coefficient ,Right now Among them, active power deviation Active power reference value With active power prediction The difference.

[0055] Hybrid Synchronization Coefficient It has a clear physical meaning: it is mainly used to unify the dimensions of multiple input variables of a hybrid synchronous quadrature phase-locked loop, and to compensate for the electrical angular frequency of the inverter output during dynamic processes. The difference between the actual electrical angular frequency and the frequency of the power grid. This is achieved by adjusting the hybrid synchronization coefficient. The size of the variable frequency drive (VFD) can be flexibly adjusted to control the active power response speed and damping characteristics of the converter under grid frequency disturbances, thereby achieving a balance between fast power point tracking and system stability.

[0056] Subsequently, the power synchronization compensation amount The cross-axis component in the grid voltage information estimated in step two The synchronization error signal is obtained by performing difference (or superposition) operations. The input of this hybrid synchronization quadrature phase-locked loop is the power synchronization quantity. With the estimated grid cross-axis voltage The difference.

[0057] Step 4: Integrate to obtain the coordinate transformation angle.

[0058] The synchronous phase-locked module 300 processes the synchronization error signal obtained in step three through a proportional-integral (PI) regulator and outputs the electrical angular frequency of the inverter. Then, the electrical angular frequency of the inverter output... By performing integration, the position angle of the coordinate transformation can be obtained. The coordinate transformation position angle The coordinate transformation between the stationary coordinate system and the synchronous coordinate system used in subsequent steps is a key parameter for constructing the discrete-time prediction model.

[0059] This phase-locked loop organically integrates the power synchronization mechanism with the traditional orthogonal phase-locked loop mechanism. It can accurately track the voltage phase of the grid in grid-following mode and achieve self-synchronization through power synchronization compensation in grid-building mode, thus being compatible with both grid-following and grid-building operation modes under a unified framework.

[0060] Step 5: Construct a discrete-time prediction model.

[0061] Model prediction module 400 is based on coordinate transformation angle A discrete-time prediction model for the converter is constructed using the grid current and the DC-side voltage obtained in step one. In this embodiment, the DC-side voltage is also incorporated. To participate in model building and improve prediction accuracy.

[0062] First, based on the inverter-grid model in the synchronous coordinate system shown in equations (1) and (2), the model is discretized using the forward Euler method to obtain the prediction models for the grid direct-axis current and grid quadrature-axis current in the synchronous coordinate system as follows: (9) (10)

[0063] In equations (9) and (10), and These are the predicted values ​​of the grid direct-axis current and grid quadrature-axis current in the next synchronous coordinate system, respectively. By traversing the preset voltage vector set, the values ​​corresponding to each preset voltage vector are... and By substituting the above model, the prediction of the grid current in the next cycle can be achieved.

[0064] Furthermore, combining equations (7) to (10), the prediction models for the active and reactive power of the grid-side converter can be obtained as follows: (11) (12)

[0065] In equations (11) and (12), and These are the predicted active power and reactive power values ​​for the next time period, respectively. The above predicted electrical parameters (including predicted grid current) , Predicted active power And predicting reactive power This will be used as input to subsequent multi-level evaluation functions to assess the control effect of each preset voltage vector.

[0066] In this embodiment, the DC side voltage The way in which the model is constructed is reflected in: the components of each voltage vector in the preset voltage vector set in the synchronous coordinate system. and This is achieved by relating the switching state of the converter to the DC-side voltage. This was obtained through combined calculations. Therefore, Accurate sampling directly affects the accuracy of the prediction model.

[0067] Step Six: Multi-level evaluation function and evaluation cost calculation.

[0068] The evaluation and optimization module 500 constructs multi-layer evaluation functions to comprehensively evaluate the control effect of each voltage vector in the preset voltage vector set. This step is the core of the entire control scheme, and will be explained in detail below using a general NPC three-level inverter as an example for the main circuit topology.

[0069] First layer: Overall architecture.

[0070] For an NPC three-level inverter, its main circuit contains 27 switching states, corresponding to 27 combinations of main circuit power transistor switching. These 27 switching states are mapped to 19 effective voltage vectors, defining a set... It includes 6 large vectors ( to (corresponding to 6 main circuit power transistor switching combinations), 6 medium vector ( to (corresponding to 6 main circuit power transistor switching combinations), 6 small vector ( to Each small vector corresponds to two main circuit power transistor switching combinations, therefore there are a total of 12 main circuit power transistor switching combinations) and zero vector ( (Corresponding to 3 main circuit power transistor switching combinations). Both the zero vector and the small vector are redundant vectors. Redundant vectors can be used to directly control multiple auxiliary control objectives such as NPC midpoint potential balance, switching losses, and common-mode voltage.

[0071] This embodiment employs a two-layer evaluation function to sequentially filter effective voltage vectors and redundant voltage vectors. The multi-layer evaluation function architecture helps reduce the number of algorithm iterations and significantly decreases the computational load. The set of redundant vectors is defined as follows: Define the set of redundant vectors to be evaluated. The overall process of the two-layer screening is shown in equations (13) to (15): (13) (14) (15)

[0072] In equations (13) to (15), This represents the effective voltage vector evaluation function selected through iterative screening. The input voltage vector corresponding to the minimum value, function This is the evaluation function for redundant voltage vectors. This represents an indicator function used to determine whether to enable the second-level evaluation function. When Belongs to the redundant vector set When the indicator function outputs 1, the second-level evaluation function is enabled. Perform secondary filtering on redundant vectors; when Not a set When the indicator function outputs 0, it directly sets the time to 0. As the optimal output. This is the final selection of the optimal main circuit power transistor switching combination.

[0073] Second layer: Effective voltage vector evaluation function .

[0074] Effective voltage vector evaluation function It consists of three components: active power control term, voltage control term, and current limiting constraint term, and its mathematical expression is shown in equation (16): (16)

[0075] In equation (16), , and These are the dynamic evaluation coefficients for the active power control target (first dynamic evaluation coefficient), the voltage control target (second dynamic evaluation coefficient), and the current control target (third dynamic evaluation coefficient). and These are the active power reference value and the grid voltage reference value, respectively. This is the current limiting function. In the function... middle, and This not only achieves adaptive control of both active power and voltage, but also unifies the dimensions of the errors between the two control targets to facilitate result evaluation. Here, the dimension is taken as power unit, therefore the dynamic evaluation coefficient... It can be taken as the power protection value of the grid-side converter.

[0076] (a) Active power control items. Based on predicted active power. With active power reference value The deviation, combined with the first dynamic evaluation coefficient The dynamic evaluation coefficient of the active power control target is calculated. The mathematical expression is as follows: (17)

[0077] In equation (17), For the hybrid synchronization coefficient, This is the rated current value of the grid-side converter. From equation (17), it can be seen that... It will change with the inverter's output electrical frequency. Predicted cross-axis current of the power grid It changes with the change. During the transient process, when the active power deviation is large, The dynamic increase accelerates the dynamic response speed of active power.

[0078] (ii) Voltage control items. Based on estimated grid voltage information and voltage reference values. The deviation, combined with the second dynamic evaluation coefficient The dynamic evaluation coefficients of the voltage control target were calculated. The mathematical expression is as follows: (18)

[0079] In equation (18), This refers to the rated capacity of the grid-side converter. With the predicted value of reactive power It changes with the changes. When the system needs stronger voltage support... The dynamic increase strengthens the converter's ability to support grid voltage.

[0080] (III) Current Limitation Constraints. Based on the predicted grid current and the preset current safety boundary, combined with the third dynamic evaluation coefficient. Calculated. Current limiting function. The mathematical expression is as follows: (19)

[0081] In equation (19), This is the current protection value (i.e., the threshold of the current safety boundary) for the grid-side converter. This current limiting function employs a non-linear penalty mechanism: when the predicted magnitude of the grid current (i.e., ...) is... The square root of the value is less than or equal to the threshold. When the current limiting constraint term is zero, it does not interfere with the voltage vector selection under normal operating conditions; when the amplitude of the predicted grid current is greater than the threshold... At that time, the current limit constraint term outputs 1, combined with a large dynamic evaluation coefficient. This causes the evaluation value of the voltage vector to increase dramatically, thus naturally eliminating it during the optimization process. This piecewise nonlinear penalty mechanism achieves an exponential increase in penalty intensity, ensuring the reliability of overcurrent protection.

[0082] In this embodiment, , and A dynamic adaptive configuration method is adopted: when the power grid is detected to be in a steady-state condition, the [configuration method] is increased. and The weighting is adjusted to improve power point tracking accuracy and voltage control accuracy; when the grid is detected to be in a transient or fault condition, and the grid current is predicted to approach the current safety boundary, the weighting is dynamically increased. This makes the current limiting constraint term in the effective voltage vector evaluation function The central authority holds a dominant position, prioritizing equipment safety.

[0083] Third layer: Redundant voltage vector evaluation function .

[0084] If the effective voltage vector evaluation function Filtering results Redundant vector set to be evaluated (Right now If the vector is small or zero, then the redundant voltage vector evaluation function is enabled. For redundant vector sets A secondary selection is performed. Since each of the 6 small vectors corresponds to 2 switch combinations (12 in total), and the zero vector corresponds to 3 switch combinations, the secondary selection only requires 5 iterations, with minimal computational cost.

[0085] Redundant voltage vector evaluation function It includes switching loss optimization terms and midpoint potential balance terms, and its mathematical expression is as follows: (20)

[0086] In equation (20), and These are the dynamic evaluation coefficients for the number of switching operations (fourth dynamic evaluation coefficient) and the dynamic evaluation coefficients for the midpoint potential (fifth dynamic evaluation coefficient). This represents the number of times the switch operates between two consecutive cycles. and These are the positive half-bus voltage and the negative half-bus voltage.

[0087] (a) Optimization of switching loss The mathematical expression for the dynamic evaluation coefficient τ0 of the number of switching operations is as follows: (twenty one)

[0088] In equation (21), This is the junction temperature of the current power switch transistor. This is the junction temperature protection value for the converter. Let be the bus voltage of the converter. From equation (21), we can see that... With the current junction temperature of power switching transistors The value increases with the increase of junction temperature, which means that the weight of the switching loss optimization term is increased when the junction temperature rises. Redundant vectors with fewer switching operations are selected first to reduce switching losses, thereby effectively suppressing further rise in junction temperature and ensuring safe operation of power devices.

[0089] (ii) Midpoint potential balance term Midpoint potential dynamic evaluation coefficient The mathematical expression is as follows: (twenty two)

[0090] In equation (22), It is half the value of the converter's midpoint potential deviation protection. Hysteresis control logic is used: when the positive half-bus voltage and negative half bus voltage The absolute value of the difference in the hysteresis loop width error At that time, A value of 0 means that no midpoint potential control is performed at this time to avoid unnecessary switching actions; when the absolute value of the difference exceeds hour, Taking a larger value means that midpoint potential control is enabled at this time, and the midpoint potential is balanced by selecting an appropriate redundancy vector.

[0091] The two-layer evaluation function requires at most 24 iterations (19 for the first layer and 5 for the second layer). Each iteration uses the input voltage vector as the current pulse. and The aforementioned prediction models are used to make a single prediction of the grid voltage, grid current, active power, and reactive power for the next circuit. Compared to the traditional single-layer evaluation function, which requires traversing all 27 switching states, the multi-layer evaluation function in this embodiment significantly reduces the computational load.

[0092] Step 7: Select the target voltage vector and generate the driving signal.

[0093] Based on the selection results of the evaluation and optimization module 500, the drive output module 600 determines the target voltage vector with the minimum evaluation cost and its corresponding main circuit power transistor switching combination. Then, based on this optimal switching combination, a corresponding PWM drive signal is generated and output to each switching device (such as IGBT or SiC MOSFET) in the main circuit of the converter to control its conduction and turn-off, thereby outputting the target voltage vector in the next control cycle and realizing precise control of grid current, active power and reactive power.

[0094] Step 8: Multi-mode compatible operation.

[0095] A key advantage of this application is that by configuring different command reference values, the converter can operate compatiblely between grid-connected and grid-connected modes without modifying the underlying control algorithm or designing complex switching logic. Specifically, this embodiment supports command reference value configuration for the following four typical application scenarios:

[0096] Evaluation function based on effective voltage vector It can be seen that this scheme can be used for both grid-connecting and grid-building functions simultaneously. It incorporates grid-connecting / grid-building control requirements into a single control framework, employing both hybrid synchronous control and multi-objective dynamic control to address these requirements. Due to different application scenarios, the reference values ​​for the commands issued by the system to the grid-side converter will vary. This scheme is compatible with various grid-connecting and grid-building application scenarios by selecting appropriate active power reference values. Compared with the grid voltage reference value The source can be compatible with various application scenarios for different network connections and network structures. Therefore, the system can choose from:

[0097] (1) Optionally, active power reference values ​​can be issued. and grid voltage reference value Two instruction reference values ​​are provided for network construction requirements and support active power control modes for network construction.

[0098] (2) Optionally, active power reference values ​​can be issued. and reactive power reference value Furthermore, a reactive power controller is added to the outer loop, and the output of the outer loop reactive power controller is used as the grid voltage reference value. It is used for grid connection or grid construction needs, and supports power control modes for grid construction and grid connection.

[0099] (3) Option to issue bus voltage reference value and grid voltage reference value Two command reference values ​​are used, and a bus voltage controller is added to the outer loop. The output of the outer loop bus voltage controller is used as the active power reference value. It is used for grid construction requirements and supports constant bus voltage control mode for grid construction.

[0100] (4) Option to issue bus voltage reference value and reactive power reference value Two command reference values ​​are used, and a bus voltage controller and a reactive power controller are added to the outer loop. The two outer loop controllers are connected in parallel, and the output of the outer loop bus voltage controller is used as the active power reference value. The output of the outer loop reactive power controller is used as the grid voltage reference value. It is used for grid connection or grid construction needs, and supports constant bus voltage control mode for grid construction and constant bus voltage control mode for grid connection.

[0101] The four application scenarios mentioned above can be switched at any time; the system only needs to select and issue different instruction reference values. Additionally, the effective voltage vector evaluation function... Current limiting function in It can be seen that all the control modes under the above-mentioned grid connection / network structure can achieve real-time constraint on the grid current. The above four application scenarios can be switched at any time. It can be achieved simply by the system selecting and issuing different instruction reference values. There is no need to design the grid connection / network structure switching logic, making system debugging simpler.

[0102] Furthermore, during a grid voltage dip, this embodiment can dynamically and smoothly switch the weight of the command reference value. For example, when a grid voltage dip is detected, the system automatically increases the grid voltage reference value. Weighting and reducing the active power reference value The weighting of the converter allows for a seamless transition from power tracking to voltage support, providing reactive current injection to the grid to support grid voltage recovery and enabling fault ride-through. During this process, because the underlying control framework remains unchanged, the switching process is smooth and shock-free, effectively avoiding the unstable oscillations introduced by strategy switching in traditional solutions.

[0103] In addition, the evaluation function based on the effective voltage vector Current limiting function in As can be seen, all the control modes under the above-mentioned grid can achieve real-time constraint on grid current, ensuring that the converter output current will not exceed the safety boundary under fault conditions such as grid voltage drop, thus ensuring equipment safety.

[0104] Corresponding to the above method embodiments, this application also provides a converter control system integrated with the grid, such as... Figure 2As shown, the system includes: a signal acquisition module 100, used to acquire the current grid current of the converter; a voltage observation module 200, used to input the grid current to the grid voltage observer and calculate the estimated grid voltage information; a synchronization phase-locked loop module 300, used to acquire the current active power deviation, convert it into a power synchronization compensation amount, and calculate the synchronization error signal by combining it with the cross-axis component in the estimated grid voltage information, and integrate it to obtain the coordinate transformation angle; a model prediction module 400, used to construct a discrete-time prediction model based on the coordinate transformation angle and grid current, and perform iterative solution to obtain the predicted electrical parameters; an evaluation and optimization module 500, used to construct a multi-layer evaluation function to calculate the evaluation cost and screen the target voltage vector; and a drive output module 600, used to generate a drive signal based on the target voltage vector to control the operation of the switching devices. The specific functions and signal flow relationships of each module correspond one-to-one with the steps of the above method, and will not be repeated here.

[0105] This application also provides a converter device, including a converter main circuit (comprising a topology and switching devices), a sampling circuit (for sampling grid current and DC-side voltage, etc.), a processor, and a memory. The processor is connected to the sampling circuit and the converter main circuit, and the memory is connected to the processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps in any of the above method embodiments. The processor can be a digital signal processor such as a DSP, FPGA, ARM, or MCU, or it can be a combination of multiple processors.

[0106] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in any of the above method embodiments. The storage medium may be a tangible, non-volatile storage medium, such as ROM, EEPROM, flash memory, magnetic disk, or optical disk.

[0107] It should be noted that this embodiment uses an NPC three-level inverter as an example for detailed description, but the control scheme of this application is not limited to this topology. Those skilled in the art will understand that this scheme is also applicable to other topologies such as two-level converters, five-level converters, and modular multilevel converters (MMC), requiring only adjustments to the voltage vector set and the dimensions of the evaluation function according to the specific topology. Furthermore, the scheme of this application is applicable to various energy storage application scenarios such as electrochemical energy storage, flywheel energy storage, supercapacitor energy storage, and hybrid energy storage, as well as new energy grid-connected scenarios such as photovoltaic grid-connected inverters and wind power converters.

[0108] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0109] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0110] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0111] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0112] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application embodiment, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0113] The above descriptions are merely implementation methods of the embodiments of this application and do not limit the scope of the embodiments of this application. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art, under the guidance of this application, may make equivalent structural or procedural transformations based on the description and drawings of the embodiments of this application, or directly or indirectly apply them to other related technical fields, without departing from the spirit and scope of protection of the claims, and all such transformations are similarly included within the protection scope of the embodiments of this application.

Claims

1. A converter control method integrated with the grid, applied to a converter, characterized in that, The method includes: The current grid current and current active power deviation of the converter are obtained; the converter is a multilevel converter. The grid current is input to the grid voltage observer to calculate the estimated grid voltage information; The active power deviation is converted into a power synchronization compensation amount, and the power synchronization compensation amount is superimposed or differentially calculated with the quadrature axis component in the estimated grid voltage information to obtain a synchronization error signal. The coordinate transformation angle is obtained by performing an integral operation based on the synchronization error signal. Based on the coordinate transformation angle and the grid current, a discrete-time prediction model of the converter is constructed. A preset voltage vector is input into the discrete-time prediction model for iterative solution to obtain the predicted electrical parameters of the converter under the action of the preset voltage vector. A multi-level evaluation function is constructed, and the predicted electrical parameters and preset command reference values ​​are input into the multi-level evaluation function to calculate the evaluation cost corresponding to each preset voltage vector; the multi-level evaluation function includes an effective voltage vector evaluation function and a redundant voltage vector evaluation function; The target voltage vector with the lowest evaluation cost is selected, and a drive signal is generated based on the target voltage vector to control the operation of the switching devices of the converter. By configuring different instruction reference values, the converter can operate compatiblely between grid-connected mode and grid-connected mode; The step of inputting the predicted electrical parameters and preset command reference values ​​into the multi-layer evaluation function to calculate the evaluation cost corresponding to each preset voltage vector includes: The predicted electrical parameters and the command reference values ​​are input into the effective voltage vector evaluation function to calculate the first evaluation value of each effective voltage vector. Select the candidate voltage vector with the lowest evaluation value in the first evaluation generation; If the candidate voltage vector has a corresponding redundant voltage vector, then the redundant voltage vector is input to the redundant voltage vector evaluation function to calculate the second evaluation value, and the redundant voltage vector with the smallest second evaluation value is taken as the target voltage vector.

2. The method according to claim 1, characterized in that, After obtaining the current grid current of the converter, the method further includes: obtaining the current DC side voltage of the converter; The discrete-time prediction model of the converter, constructed based on the coordinate transformation angle and the grid current, includes: Based on the coordinate transformation angle, the grid current, and the DC-side voltage, the discrete-time prediction model is constructed; wherein, the predicted electrical parameters include at least one of the predicted grid current, predicted active power, and predicted reactive power.

3. The method according to claim 1, characterized in that, The grid voltage observer is a reduced-order observer; the step of inputting the grid current to the grid voltage observer and calculating the estimated grid voltage information includes: Construct a reduced-order state-space model with the grid current as a known input variable; The sampled grid current is input into the reduced-order state-space model, and the estimated grid voltage information is output by iteratively solving the discretized state equations. The observer bandwidth of the reduced-order state-space model is configured to be higher than the grid fundamental frequency and lower than the switching frequency of the converter.

4. The method according to claim 1, characterized in that, The step of converting the active power deviation into a power synchronization compensation amount includes: The active power deviation is multiplied by a preset hybrid synchronization coefficient to obtain the power synchronization compensation amount; The hybrid synchronization coefficient is used to adjust the active power response speed and damping characteristics of the converter under grid frequency disturbances.

5. The method according to claim 1, characterized in that, The effective voltage vector evaluation function includes active power control terms, voltage control terms, and current limit constraint terms; The calculation method for the first evaluation cost value includes: The active power control item is calculated based on the deviation between the predicted active power and the active power reference value in the predicted electrical parameters, combined with the first dynamic evaluation coefficient. The voltage control term is calculated based on the deviation between the estimated grid voltage information and the voltage reference value, combined with the second dynamic evaluation coefficient. Based on the predicted grid current in the predicted electrical parameters and the preset current safety boundary, the current limit constraint term is calculated in combination with the third dynamic evaluation coefficient. The first evaluation value is obtained by weighted summation of the active power control term, the voltage control term, and the current limit constraint term.

6. The method according to claim 5, characterized in that, The dynamic configuration methods for the first dynamic evaluation coefficient, the second dynamic evaluation coefficient, and the third dynamic evaluation coefficient include: When the power grid is detected to be in a steady-state condition, the first dynamic evaluation coefficient and the second dynamic evaluation coefficient are increased. When the power grid is detected to be in a transient or fault condition, and the predicted power grid current approaches the current safety boundary, the third dynamic evaluation coefficient is dynamically increased so that the current limit constraint term occupies the dominant weight in the effective voltage vector evaluation function.

7. The method according to claim 5, characterized in that, The calculation of the current limitation constraint term based on the predicted grid current in the predicted electrical parameters and the preset current safety boundary, combined with the third dynamic evaluation coefficient, includes: The magnitude of the predicted grid current is compared with the threshold of the current safety boundary; When the magnitude of the predicted grid current is less than or equal to the threshold, the output of the current limit constraint term is zero; When the amplitude of the predicted grid current exceeds the threshold, a nonlinear penalty function is used to calculate the current limit constraint term. The penalty intensity of the nonlinear penalty function increases exponentially with the degree to which the amplitude of the predicted grid current exceeds the threshold.

8. The method according to claim 1, characterized in that, The redundant voltage vector evaluation function includes a switching loss optimization term and a midpoint potential balance term; The calculation method for the second evaluation cost value includes: Based on the current switching state of the multilevel converter and the predicted switching state at the next moment, the switching loss optimization term is calculated in combination with the fourth dynamic evaluation coefficient. The midpoint potential balance term is calculated based on the deviation of the upper and lower bridge arm capacitor voltages on the DC side of the multilevel converter, combined with the fifth dynamic evaluation coefficient. The second evaluation value is obtained by weighted summing of the switching loss optimization term and the midpoint potential balance term.

9. The method according to claim 1, characterized in that, The step of configuring different instruction reference values ​​to enable the converter to operate compatiblely between grid-connected mode and grid-connected mode includes: When a grid-connected mode command is received, the command reference value is configured as an active power reference value and a reactive power reference value. When a network configuration mode command is received, the command reference value is configured as a voltage amplitude reference value and a DC side voltage reference value; During grid voltage dips, the weights of the command reference values ​​are dynamically and smoothly switched, enabling the converter to seamlessly transition from power tracking to voltage support.

10. A converter control system integrated with the grid, characterized in that, include: The signal acquisition module is used to acquire the current grid current of the converter; the converter is a multi-level converter. The voltage observation module is used to input the grid current to the grid voltage observer and calculate the estimated grid voltage information. The phase-locked loop module is used to acquire the current active power deviation, convert the active power deviation into a power synchronization compensation amount, and superimpose or perform difference calculation on the power synchronization compensation amount and the quadrature axis component in the estimated grid voltage information to obtain a synchronization error signal. Based on the synchronization error signal, an integral calculation is performed to obtain the coordinate transformation angle. The model prediction module is used to construct a discrete-time prediction model of the converter based on the coordinate transformation angle and the grid current, and to input a preset voltage vector into the discrete-time prediction model for iterative solution to obtain the predicted electrical parameters of the converter under the action of the preset voltage vector. The evaluation and optimization module is used to construct a multi-level evaluation function. The predicted electrical parameters and preset command reference values ​​are input into the multi-level evaluation function to calculate the evaluation cost value corresponding to each preset voltage vector; and the target voltage vector with the minimum evaluation cost value is selected. The multi-layer evaluation function includes an effective voltage vector evaluation function and a redundant voltage vector evaluation function; A drive output module is used to generate a drive signal based on the target voltage vector to control the operation of the switching devices of the converter; By configuring different instruction reference values, the converter can operate compatiblely between grid-connected mode and grid-connected mode; The evaluation optimization module is further configured to input the predicted electrical parameters and the instruction reference values ​​into the effective voltage vector evaluation function to calculate the first evaluation value of each effective voltage vector; filter out the candidate voltage vector with the smallest first evaluation value; if the candidate voltage vector has a corresponding redundant voltage vector, input the redundant voltage vector into the redundant voltage vector evaluation function to calculate the second evaluation value, and use the redundant voltage vector with the smallest second evaluation value as the target voltage vector.

11. A converter device, characterized in that, include: The main circuit of the converter includes the topology and switching devices; A sampling circuit is used to sample the mains current. The processor is connected to the sampling circuit and the main circuit of the converter. A memory connected to the processor, the memory storing a computer program, wherein the processor executes the computer program to implement the method as described in any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 9.

Images