A method and system for active support of grid-forming converter based on predictive control

Abstract

The application discloses a network-constructing type converter active support method and system based on predictive control, which comprises the following steps: a discrete time prediction model and a parameter adaptive VSG control model are established, a virtual rotational inertia and a damping coefficient are dynamically adjusted, and an additional damping coefficient is introduced to optimize a control link. A virtual impedance and an adaptive virtual resistance are used to correct a voltage reference value and adapt to different working conditions. An MPC cost function containing multi-step prediction, a switch penalty term and a constraint is designed, and an optimal control quantity is solved. Power grid signals are collected, a reference value is generated by an adaptive VSG, and a PWM modulation signal is outputted by the MPC to control an inverter. The application can realize fast frequency response, accurate voltage tracking and reliable fault ride-through, and improves the stable operation capability of a microgrid system.

Description

Technical Field

[0001] This invention relates to the field of power electronic converter control and micro-energy system technology, specifically to an active support method for grid-connected converters based on predictive control that integrates adaptive parameters and multi-step prediction. It is applicable to the active support control of frequency and voltage of grid-connected inverters in microgrids, weak grids, and islanded power supply systems with a high proportion of renewable energy access. Background Technology

[0002] With the continuous increase in the penetration rate of new energy sources such as wind power and photovoltaics, the gradual retirement of synchronous generators has led to a significant reduction in grid inertia, causing even slight power fluctuations to trigger severe frequency oscillations. Grid-connected inverters (GFM) have become key devices for solving the stability problem of low-inertia grids due to their synchronous motor-like characteristics. Among them, virtual synchronous generator (VSG) control provides inertial support and frequency and voltage regulation capabilities to the system by simulating the electromechanical equations of synchronous generators, and is widely used in micro-energy systems.

[0003] Traditional VSG control strategies employ a hierarchical structure combining outer-loop power droop control and inner-loop voltage and current dual closed-loop control. This structure achieves power regulation and grid frequency support by simulating the rotor motion equations and droop characteristics of a synchronous generator. However, this structure has inherent technical limitations. Due to the virtual rotational inertia of the VSG... With damping coefficient Traditionally set to fixed values, VSGs cannot adaptively adjust to dynamic changes in grid frequency. This leads to problems such as frequency deviation exceeding allowable ranges and excessive frequency change rate during load surges. Furthermore, the traditional VSG's inner-loop voltage and current dual-closed-loop control strategy relies on precise PI parameter tuning, resulting in slow dynamic response and insufficient phase margin in weak grid scenarios. When symmetrical or asymmetrical grid faults occur, traditional VSG control strategies lack effective fault current suppression mechanisms, leading to excessively high output current distortion and excessive overcurrent overshoot, potentially damaging power devices. In addition, in multi-unit parallel operation scenarios, differences in the impedance parameters of actual distribution lines cause power distribution deviations, which can lead to overload operation of some units and system power oscillations, affecting the overall stability of the parallel system.

[0004] Common methods for suppressing secondary ripple current can be divided into hardware methods and control methods. Hardware methods include using large-capacity electrolytic capacitors to suppress bus voltage fluctuations and using additional circuits to absorb ripple current; both methods increase the system size and cost. The voltage-current dual-loop active suppression method proposed in foreign literature introduces inverter-side filter inductor current feedback, which is equivalent to increasing the closed-loop output impedance of the DC-DC converter across the entire output voltage frequency band, thus suppressing secondary ripple current. However, this reduces the crossover frequency of the outer voltage loop, leading to a deterioration in the system's dynamic performance during load changes.

[0005] Therefore, traditional VSG control strategies still have many technical shortcomings that need to be addressed. Summary of the Invention

[0006] This invention aims to solve the problem of fixed control parameters in traditional VSG (virtual rotational inertia of VSG). With damping coefficient The problem of VSG control performance being greatly affected by the fact that it is usually set to a fixed value. In order to improve VSG control performance, the technical solution of this invention provides an active support method and system for grid-type converters based on predictive control. By dynamically adjusting the virtual moment of inertia J and damping coefficient D, it fundamentally solves the problem that the traditional VSG control strategy cannot adaptively adjust according to the dynamic changes of the grid frequency due to the fixed control parameters, which leads to problems such as frequency deviation exceeding the allowable range and frequency change rate exceeding the standard in the case of sudden load changes.

[0007] Therefore, the present invention provides the following technical solution:

[0008] The present invention provides an active support method for a grid-type converter based on predictive control, which performs the following steps in each sampling period:

[0009] The system collects the speed deviation of the grid-type converter system, the speed change rate of the virtual synchronous machine, and the current and voltage signals of the power grid.

[0010] Based on the adaptive VSG control model, the virtual moment of inertia J and the damping coefficient D are dynamically adjusted.

[0011] The adaptive VSG control model is based on the speed deviation. and the rate of change of speed of the virtual synchronizer The grid-type converter system is divided into acceleration mode and deceleration mode; if In acceleration mode, the virtual moment of inertia J is increased to suppress the rate of change of the system output frequency; if In deceleration mode, the virtual moment of inertia J is reduced to accelerate the frequency recovery to the rated value; if the speed deviation is greater than or equal to the speed deviation threshold... Then increase the damping coefficient Otherwise, the damping coefficient Set to the preset steady-state damping coefficient; t is time;

[0012] The dynamically adjusted virtual moment of inertia J and damping coefficient D are introduced into the virtual synchronous generator VSG control strategy, and the VSG control and MPC control output PWM modulation signal to control the inverter.

[0013] Optionally, if the grid-type converter system is in acceleration mode, then the virtual moment of inertia... If the grid-type converter system is in deceleration mode, then the virtual moment of inertia... ;

[0014] in, The steady-state moment of inertia coefficient, , and All are adjustment factors; t represents the rotational speed of the virtual synchronous generator; t represents time.

[0015] Increase damping coefficient hour, , The steady-state damping coefficient is... This is the damping adjustment coefficient.

[0016] Optionally, the dynamically adjusted virtual moment of inertia J and damping coefficient D are introduced into the virtual synchronous generator VSG control strategy. The output voltage reference value and output current reference value are obtained through VSG control, specifically including:

[0017] Based on the dynamically adjusted virtual moment of inertia J and damping coefficient D, the virtual phase angle is obtained through an active-frequency droop control loop. And the internal potential amplitude E is obtained through the reactive power-voltage droop control loop;

[0018] Using the virtual phase angle And the internal potential amplitude E, introduce virtual impedance and adaptive virtual resistance to correct the output voltage reference value;

[0019] Under normal operating conditions, the original virtual resistor is used; under fault conditions, an adaptive virtual resistor is used instead of the original virtual resistor. for: , The internal potential amplitude of the VSG, The current limiting threshold is... This refers to the output current amplitude of the inverter. This represents the minimum amplitude of the three-phase voltage.

[0020] Optionally, there is an active power deviation in the active-frequency droop control loop. An additional damping coefficient is introduced after the calculation node. Construct the following closed-loop transfer function :

[0021]

[0022] In the formula, Here, is the PI ring proportionality coefficient, and s is the complex parameter variable of the Laplace transform;

[0023] Among them, active power deviation After closed-loop transfer function After calculation, output angular velocity compensation amount angular velocity compensation Combined with grid-side angular velocity Generate optimized virtual synchronizer speed Finally, after integration, the corrected virtual phase angle is obtained. .

[0024] Optionally, the MPC predictive control uses the zero-order hold method to discretize and obtain a discretized predictive model:

[0025]

[0026] in, and These are the matrix coefficients, and their corresponding values ​​are as follows: and ;matrix ,matrix ; Matrix coefficients The element value; Matrix coefficients The element value; The sampling period; For filtering inductors, For inductor parasitic resistance, For filtering capacitors;

[0027] These are the inverter-side filter inductor currents at time k+1 and time k, respectively, during discrete sampling. These are the filter capacitor voltages at time k+1 and time k, respectively, during discrete sampling; Let k be the grid-side current at time k, which is discretely sampled. Let k be the inverter bridge arm output voltage at time k.

[0028] Optionally, the model for the cost function J in the MPC predictive control is:

[0029]

[0030] in, These are the weighting coefficients. For the future The predicted value of the filter capacitor voltage at time k, where variable k represents the current discrete sampling time. These represent the switching states of the j-th bridge arm at time k+1 and time k, respectively, with values ​​of 0 or 1; j=1,2,3 correspond to the three-phase bridge arms of the inverter, respectively. For overvoltage and overcurrent constraints, The switching penalty coefficient is... This is the output voltage reference value. This is the reference value for the inverter output current. These are the inverter-side filter inductor currents at time k+1 of the discrete sampling.

[0031] Optionally,

[0032] in, and These are the maximum allowable values ​​for the filter capacitor voltage and the inverter-side filter inductor current, respectively. and These represent the predicted values ​​of the filter capacitor voltage and the inverter-side filter inductor current, respectively, corresponding to the i-th prediction step size. and If any one of the above inequalities exists, then overvoltage and overcurrent protection will be implemented.

[0033] Optionally, a weighting coefficient can be set under normal operating conditions. When the inverter-side filter inductor current is detected , Set as , The set overcurrent protection safety threshold for the inverter. These are the minimum and maximum values ​​of the preset weighting coefficients, respectively.

[0034] The present invention also provides a control device based on the above method, comprising:

[0035] The acquisition module is used to acquire speed deviations, as well as current and voltage signals from the power grid.

[0036] The VSG control module is used to dynamically adjust the virtual moment of inertia J and damping coefficient D based on the adaptive VSG control model, and to introduce the dynamically adjusted virtual moment of inertia J and damping coefficient D into the VSG control strategy of the virtual synchronous generator, so as to control the output voltage reference value through VSG control.

[0037] The MPC control module is used to introduce the output voltage reference value into the MPC predictive control and solve the optimal control quantity by traversing various switching states of the inverter.

[0038] The PWM drive module is used to output a PWM modulation signal based on the optimal control quantity to control the operation of the inverter switching transistors.

[0039] The present invention also provides a control system, comprising:

[0040] The power supply is connected to the power grid via an inverter;

[0041] The aforementioned control device;

[0042] The inverter is connected to the control device and receives the PWM modulation signal generated by the control device.

[0043] Compared with the prior art, the present invention achieves the following progress and effects:

[0044] The active support method and system for grid-type converters based on predictive control disclosed in this invention features a core optimization: adaptive VSG control. This adaptive VSG control utilizes speed deviation... With rate of change The product sign dynamically adjusts the virtual moment of inertia. With damping coefficient This fundamentally solves the problem that traditional VSG control strategies cannot adaptively adjust to dynamic changes in grid frequency due to fixed control parameters, which leads to problems such as frequency deviation exceeding the allowable range and frequency change rate exceeding the standard in scenarios of sudden load changes.

[0045] Furthermore, this invention proposes two parts: improved APCL control and enhanced MPC control. The improved APCL control involves introducing an additional damping coefficient. Optimize the pole distribution of the active-frequency droop control loop, adjust the system damping ratio and inherent oscillation speed, and enhance the frequency dynamic response and active support capability without affecting steady-state characteristics; generate power and voltage reference benchmarks adapted to the operating conditions; Enhanced MPC control: based on a zero-order hold discretized system prediction model, combined with multi-step prediction (… , A multi-objective cost function is constructed using time-of-flight, switching penalty terms, and overvoltage / overcurrent constraints. The optimal PWM duty cycle d is then calculated by traversing the eight switching states of the inverter. Based on the duty cycle d, a triangular carrier wave is modulated using PWM to obtain the control signal for one bridge arm switch in the inverter. The control signal for the other bridge arm switch is complementary to this signal, and the inverter output is regulated using this control signal. Under fault conditions, the MPC automatically switches the weighting coefficients. (Increased from 3 to 30), combined with adaptive virtual resistance The updated voltage reference adapts to both symmetrical and asymmetrical faults without requiring fault type detection, ensuring reliable fault ride-through. This invention enhances frequency support flexibility and inertial damping adaptability through adaptive VSG, accelerates dynamic response and reduces switching losses through enhanced MPC, and optimizes active power link stability through improved APCL control. These three elements work synergistically to give the system stable support, efficient regulation, and robust operation capabilities.

[0046] In summary, this invention improves frequency support flexibility and inertial damping adaptability through adaptive VSG, accelerates dynamic response and reduces switching losses through enhanced MPC, and optimizes active power link stability through improved APCL control. The three work together to enable the system to have stable support, efficient regulation and robust operation capabilities. Attached Figure Description

[0047] Figure 1 This is a system block diagram of the improved APCL control module according to an embodiment of the present invention;

[0048] Figure 2 A block diagram of the VSG control model;

[0049] Figure 3 This is a block diagram of the active support and enhancement system for the virtual synchronous generator of the grid-connected inverter based on predictive control, according to an embodiment of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the application. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the application. It should be noted that the reference to "embodiment" herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The presentation of this phrase in various locations throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. Those skilled in the art will understand that the embodiments described herein can be combined with other embodiments. The term "and / or" as used herein refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations.

[0051] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0052] like Figure 1 and Figure 3 The control block diagram shown illustrates an active support method for a grid-type converter based on predictive control, provided by an embodiment of the present invention, comprising the following steps:

[0053] 1) Data sampling: Acquire speed signal, current signal, and voltage signal.

[0054] Specifically, within each sampling period, the inverter-side angular velocity (virtual synchronous machine angular velocity) is monitored via a phase-locked loop. ) and grid-side angular velocity Samples were taken separately, and the speed deviation was calculated. Rate of change of speed of virtual synchronous machine ; For the grid side angular velocity, f is the power grid frequency; ω represents the angular velocity of the virtual synchronizer; t represents time.

[0055] Three-phase current on the inverter side is collected using a high-precision Hall sensor. , , Three-phase voltage of filter capacitor , , and the three-phase current on the grid side , , After coordinate transformation, the current and voltage components along the dq axis are obtained, i.e., the current signal. , , ) and voltage signal ( , , )through After coordinate transformation, a two-phase stationary coordinate system is obtained. The grid-side current under With filter capacitor voltage Then, the real-time output active power is calculated using instantaneous power theory. With reactive power .

[0056] 2) Parameter adjustment: Based on the adaptive VSG control model, the virtual moment of inertia is dynamically adjusted according to the following rules. With damping coefficient The adaptive VSG control model proposes dividing the system into acceleration modes, such as when a sudden increase in load causes a decrease in speed and an increase in the rate of decrease. This can be addressed by increasing... Suppressing the rate of change of frequency (RoCoF) to avoid protection triggering; and deceleration mode, such as when a sudden load reduction causes a decrease in speed and a reduced rate of increase, by reducing... Accelerate the frequency recovery to the rated value.

[0057] In some embodiments, based on the collected rotational speed deviation Rate of change of speed of virtual synchronous machine The product partitioning system operates in the following mode: when At that time, the delineation system is in acceleration mode; when At that time, the delineation system was in deceleration mode.

[0058] It should be understood that this invention proposes to dynamically adjust the virtual moment of inertia based on the system's operating mode and the aforementioned rules. With damping coefficient However, in some embodiments, to improve control performance, the following preferred adjustment rules are also provided. It should be understood that these adjustment rules are preferred techniques of the present invention, but not the only feasible approach. Specifically:

[0059] The system is in acceleration mode, with virtual rotational inertia. Model Make adjustments to achieve Increase;

[0060] The system is in deceleration mode, with virtual rotational inertia. use Make adjustments to achieve Decrease.

[0061] In the formula, The steady-state moment of inertia coefficient, , and All are adjustment factors.

[0062] Damping coefficient Based on speed deviation threshold Adjustments need to be made, namely:

[0063] when When, the preferred damping coefficient is... Model Increase damping to suppress oscillations;

[0064] when When, damping coefficient This is used to ensure steady-state accuracy. The steady-state damping coefficient is... This is the damping adjustment coefficient.

[0065] In summary, virtual moment of inertia The adaptive adjustment formula is:

[0066]

[0067] Damping coefficient The adaptive adjustment formula is:

[0068]

[0069] In practical implementation, the steady-state moment of inertia coefficient This is generally an empirical value; in this example, it is set to 0.05; adjustment factor , and These are also empirical values; in this embodiment, they are taken as 0.1, 0.02, and 0.03 respectively; speed deviation threshold. This is also an empirical value; in this embodiment, the value is taken as follows: Steady-state damping coefficient in this embodiment The value is 20, which is the damping adjustment coefficient. The value is 50. Therefore, when... At that time, through Increase damping to suppress oscillations; when At that time, keep This ensures that the system damping ratio remains within a stable range of 0.7 to 1.0.

[0070] 3) Based on the updated virtual moment of inertia With damping coefficient The virtual phase angle is obtained through the active power droop loop. And the internal potential amplitude is obtained through the reactive power droop loop. .

[0071] It should be understood that the use of rotational inertia With damping coefficient The virtual phase angle is obtained through the active power droop loop. And the internal potential amplitude is obtained through the reactive power droop loop. These are all mature technologies in VSG control technology. That is, in some embodiments of this invention, there are no restrictions on how rotational inertia is utilized. With damping coefficient VSG control can be implemented using existing technologies. The implementation process in this embodiment is as follows:

[0072] For the active-frequency droop control loop, the VSG control module receives the calculated active power value. Calculated reactive power After receiving the frequency feedback signal, the updated virtual synchronous machine speed is calculated using the active power droop loop equation, and the virtual phase angle is obtained by integrating the virtual synchronous machine speed. The equation for the active drooping loop is:

[0073]

[0074] in, This is a reference value for active power. For the grid side angular velocity, This represents the speed of the virtual synchronizer.

[0075] For the reactive power-voltage droop control loop, the virtual voltage is calculated using the reactive power droop loop equation. , virtual voltage With grid-side voltage Voltage error between Transfer function with PI controller Multiplying them together gives the magnitude of the internal potential. The reactive power droop loop equation is as follows:

[0076]

[0077] Transfer function of outer loop voltage PI controller The expression is:

[0078]

[0079] In the formula, It is the proportional gain of the PI controller. It is the integral coefficient of the PI controller. s is the complex parameter variable of the Laplace transform, and j is the imaginary unit symbol. For the virtual synchronizer speed, This is a reference value for reactive power. The system's rated voltage. This is the reactive power-voltage droop coefficient.

[0080] Reference Figure 2 In some embodiments, the APCL (Active Power Control Loop) control module system block diagram is further improved, specifically by introducing an additional damping coefficient. Optimize the original active power-frequency droop control loop, i.e., active power deviation An additional damping coefficient is introduced after the calculation node. The following closed-loop transfer function was constructed. :

[0081]

[0082] Additional damping coefficient in this embodiment The value is It should be understood that the values ​​can also be adjusted based on the model's performance. This is the PI ring proportionality coefficient; This refers to the angular velocity on the power grid side.

[0083] From a practical application perspective, active power deviation After closed-loop transfer function After calculation, the angular velocity compensation amount is obtained. Its expression is:

[0084]

[0085] in, This indicates that the angular velocity compensation generated by the improved APCL control loop is different from the speed deviation mentioned above for adaptive VSG parameter adjustment. The angular velocity compensation amount Angular velocity of the power grid side Superimposed to generate the optimized virtual synchronizer speed ,Right now Then, through integration, the corrected virtual phase angle is obtained. .

[0086] The above technical solution adjusts the additional damping coefficient. Optimizing the system pole distribution so that the dominant pole is located on the real axis significantly increases system damping and reduces power overshoot without altering the frequency droop characteristics, while also improving the phase margin to 128° in weak grid environments. In load surge scenarios, the improved APCL control reduces the active power response time from 0.85s to 0.42s and the frequency settling time from 1.06s to 0.6s, with no significant oscillations.

[0087] 4) Correct the output voltage reference value by introducing virtual impedance and adaptive virtual resistance to correct the output voltage reference value under different operating conditions. Details are as follows:

[0088] Under normal operating conditions, the corrected model is as follows:

[0089]

[0090] In the formula, For virtual resistance; For virtual reactance; This is the reference value for the inverter output voltage input to the inner MPC control loop after virtual impedance correction; The initial internal potential reference value generated by the droop control (i.e., the voltage reference value before correction); denoted as the grid-side output current, and j represents the imaginary unit symbol.

[0091] Under fault conditions, the corrected model is as follows:

[0092]

[0093] in, This is the corrected output voltage reference value under fault conditions. As an adaptive virtual resistor, it eliminates the need to detect fault types by sampling the minimum amplitude of the three-phase voltage across the filter capacitor. Calculate adaptive virtual resistance , The internal potential amplitude of the VSG, The current limiting threshold is... This refers to the inverter output current amplitude; it should be understood that an adaptive virtual resistor is introduced under fault conditions. Replace the original virtual resistor This is the preferred and feasible method of the present invention. The inverter output current reference value The axis component, which can be calculated by the desired current amplitude and virtual phase angle generated by the VSG control outer loop through conventional Park / Clark coordinate inverse transformation technology, is not limited in this invention, that is, it can be achieved by existing technology.

[0094] It should be noted that this invention assumes that the current operating condition is known to be either a normal operating condition or a faulty operating condition, and does not impose any limitations on the operating condition identification technology or the fault type.

[0095] 5) Based on the continuous-time mathematical model of the grid inverter and LCL filter (inductor-capacitor-inductor filter), the zero-order hold method is used for discretization to obtain... Predicted values ​​at specific times. Detailed analysis follows:

[0096] First, based on circuit laws, establish The system continuity model in the coordinate system is as follows:

[0097]

[0098] Therefore, the expression for the discretized prediction model is:

[0099]

[0100] In the formula, The three-phase currents on the inverter side are respectively , , and three-phase voltage through After coordinate transformation, a two-phase stationary coordinate system is obtained. The inverter-side filter inductor current and inverter bridge arm output voltage; This is the voltage across the filter capacitor; This represents the output current from the grid side; A and B are both matrices. and Both are coefficient matrices, and their calculations are as follows: and ,matrix ,matrix ,in, This refers to the parasitic resistance of the inductor. For the filter inductor, the value in this embodiment is [value missing]. , For the filter capacitor, the value in the embodiment is [value missing]. ; Coefficient matrix The element value; Coefficient matrix The element value; The sampling period is set to a value in this embodiment. where e is the natural base; These are the inverter-side filter inductor currents at time k+1 and time k, respectively, during discrete sampling. These are the filter capacitor voltages at time k+1 and time k, respectively, during discrete sampling; Let k be the grid-side current at time k, which is discretely sampled.

[0101] Specifically, the voltage of the filter capacitor at the current sampling time k Inverter-side filter inductor current and grid-side current All values ​​were obtained through sampling and constitute the initial state of the discretized prediction model; the eight switching states of the inverter correspond to the output voltages of the eight candidate bridge arms. The values ​​are successively substituted into the prediction model as model inputs. When predicting time k+1, the current state and candidate inputs are directly calculated. When predicting subsequent times such as k+2 and k+3, it is not necessary to obtain the future measured values ​​in advance. Instead, the predicted state calculated in the previous step is used as the new initial state to continue the recursion, and the same candidate switching state or candidate switching state sequence is used for calculation within the prediction domain. Therefore, the multi-step prediction results in the predicted values ​​of the filter capacitor voltage and the inverter-side filter inductor current under each candidate switching state. The cost function J calculates the evaluation value based on these predicted values, the output voltage reference value, the overvoltage and overcurrent constraints, and the switching penalty term. Finally, the switching state with the smallest evaluation value is selected as the optimal switching control signal for the current sampling period.

[0102] 6) In the design of the cost function of MPC (Model Predictive Control), dynamically adjust the weight coefficient of the current penalty term. Under normal operating conditions, the following settings are provided: Prioritize voltage tracking; when inverter-side filter inductor current is detected... ( When the inverter overcurrent protection safety threshold is set, Automatically switch to Prioritize current limiting. After fault clearance, adaptive virtual resistor... Automatically reset to zero, the MPC weight coefficients are synchronously restored to normal operating condition settings. Among these, These are the minimum and maximum values, respectively, both of which are empirical values.

[0103] 7) Implement based on this discretized prediction model , Multi-step prediction of grid-side current and inverter-side voltage at any given time, and then... , Substituting the grid-side current and inverter-side voltage into the cost function at any given time, the expression is as follows:

[0104]

[0105] in, These are weighting coefficients, and are empirical values; in this example, we take... It is 5. =1; This is the switching penalty coefficient; This involves overvoltage and overcurrent constraints. The inverter's eight switching states are iterated, and the cost function value for each state is calculated. The switching state that minimizes the cost function is selected, and the optimal switching control signal is directly output. ) Drive the inverter switching transistors to operate; For the future The predicted value of the filter capacitor voltage at time k represents the current discrete sampling time. This predicted value is obtained by recursively calculating the current sampled voltage, current, and the traversed inverter switching states into the system's discretized prediction model. These represent the switching states of the j-th bridge arm at time k+1 and time k, respectively, with values ​​of 0 or 1; j=1,2,3 correspond to the three-phase bridge arms of the inverter. Switching penalty term. It is used to characterize the number of changes in the candidate switch state relative to the current switch state, thereby suppressing unnecessary switching actions and reducing switching losses in the cost function.

[0106] For overvoltage and overcurrent constraints, the expression is:

[0107]

[0108] in, and These are the maximum allowable values ​​for the filter capacitor voltage and the inverter-side filter inductor current, respectively. Their values ​​can be preset based on the rated values ​​and protection margins of the filter capacitor, filter inductor, and inverter switching devices. and These represent the predicted values ​​of the filter capacitor voltage and the inverter-side filter inductor current for the i-th prediction step, respectively, corresponding to the values ​​mentioned earlier. and That is, if the predicted voltage or current value exceeds the limit under any prediction step size, Jlim is assigned the value of infinity.

[0109] 8) Based on the duty cycle d, the triangular carrier is PWM modulated to obtain the control signal of one switch in the DC converter. The control signal of the other switch in the DC converter is complementary to this control signal.

[0110] The control module iterates through the inverter's eight switching states, calculates the cost function value for each state, selects the optimal switching state to output the PWM duty cycle, and drives the inverter's switching transistors. , , The operation is as follows: the switching frequency is stabilized at 14kHz, and the dead time is set to... To avoid bridge arm straight-through. MPC weighting coefficient under fault conditions. The system automatically switches from 3 to 30, prioritizing current limiting to ensure that the fault current does not exceed 1.5 times the rated value and maintaining the sinusoidal waveform (i.e., THD < 2%). This mechanism does not require fault type detection, can adapt to symmetrical and asymmetrical faults, and can still maintain the minimum voltage support requirement in severe fault scenarios where the voltage drops to near zero, and resume steady-state operation within 0.22 seconds after the fault is cleared.

[0111] In some embodiments, The value is 0.05; , and The values ​​are 0.1, 0.02, and 0.03. The value is ; The value is 20; The value is 50. The value is ; The value is . The values ​​are 2.58. The value is . The value is 0.5. The value is 311V; The value is 10A.

[0112] It should be understood that the above-described cost function design is a preferred embodiment of the present invention, but not the only feasible technique. In other feasible embodiments, cost functions already existing in the art can be referenced, and discretized prediction models can be integrated for MPC control. The present invention does not impose specific limitations on this.

[0113] In some embodiments, the present invention also proposes a control device based on the above method, including a data acquisition module, a VSG control module, an MPC control module, and a PWM drive module connected in sequence or interconnected with each other.

[0114] The acquisition module is used to acquire speed deviations, as well as current and voltage signals from the power grid.

[0115] The VSG control module is used to dynamically adjust the virtual moment of inertia J and damping coefficient D based on the adaptive VSG control model, and to introduce the dynamically adjusted virtual moment of inertia J and damping coefficient D into the VSG control strategy of the virtual synchronous generator, thereby controlling the output voltage reference value through VSG control.

[0116] The MPC control module is used to introduce the output voltage reference value into the MPC predictive control and solve for the optimal control quantity by traversing various switching states of the inverter.

[0117] The PWM drive module is used to output a PWM modulation signal based on the optimal control quantity to control the operation of the inverter switching transistors.

[0118] It should also be understood that the specific implementation process of each module is described in the above method. This invention will not repeat it here. The above division of functional modules is only for illustrative purposes. In some embodiments, some functional modules can be combined and some functional modules can be separated. Each functional module can be implemented in software, hardware, or a combination of software and hardware. The software and hardware devices include, but are not limited to, general-purpose computer equipment, programmable gate arrays, digital signal processors, microprocessors and their corresponding programming or burning software.

[0119] In some embodiments, the present invention also proposes a control system, comprising: a power supply terminal, a control device, and an inverter, wherein the power supply terminal is connected to the power grid via the inverter; the inverter is connected to the control device and receives a PWM modulation signal generated by the control device.

[0120] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for active support of a grid-type converter based on predictive control, characterized in that, The following steps are performed in each sampling period: The system collects the speed deviation and virtual synchronous machine speed change rate of the grid-connected converter system, as well as the current and voltage signals of the power grid. Based on the adaptive VSG control model, the virtual moment of inertia J and the damping coefficient D are dynamically adjusted. The adaptive VSG control model is based on the speed deviation. and the rate of change of speed of the virtual synchronizer The grid-type converter system is divided into acceleration mode and deceleration mode; if In acceleration mode, the virtual moment of inertia J is increased to suppress the rate of change of the system output frequency; if In deceleration mode, the virtual moment of inertia J is reduced to accelerate the frequency recovery to the rated value; if the rotational speed deviates... Greater than or equal to the speed deviation threshold Then increase the damping coefficient Otherwise, the damping coefficient Set to the preset steady-state damping coefficient; t is time; The dynamically adjusted virtual moment of inertia J and damping coefficient D are introduced into the virtual synchronous generator VSG control strategy, and the VSG control and MPC control output PWM modulation signal to control the inverter.

2. The method according to claim 1, characterized in that: If the grid-type converter system is in acceleration mode, then the virtual rotational inertia... If the grid-type converter system is in deceleration mode, then the virtual moment of inertia... ; in, The steady-state moment of inertia coefficient, , and All are adjustment factors; t represents the rotational speed of the virtual synchronous generator; t represents time. Increase damping coefficient hour, , This is the steady-state damping coefficient. This is the damping adjustment coefficient.

3. The method according to claim 1, characterized in that: The dynamically adjusted virtual moment of inertia J and damping coefficient D are introduced into the virtual synchronous generator VSG control strategy. Output voltage and output current reference values ​​are obtained through VSG control, specifically including: Based on the dynamically adjusted virtual moment of inertia J and damping coefficient D, the virtual phase angle is obtained through an active-frequency droop control loop. And the internal potential amplitude E is obtained through the reactive power-voltage droop control loop; Using the virtual phase angle The output voltage reference value is corrected by introducing virtual impedance and adaptive virtual resistance, along with the internal potential amplitude E. Under normal operating conditions, the original virtual resistor is used; under fault conditions, an adaptive virtual resistor is used instead of the original virtual resistor. for: , The internal potential amplitude of the VSG, The current limiting threshold is... This refers to the output current amplitude of the inverter. This represents the minimum amplitude of the three-phase voltage.

4. The method according to claim 3, characterized in that: Active power deviation in the active-frequency droop control loop An additional damping coefficient is introduced after the calculation node. The following closed-loop transfer function is constructed. : In the formula, Here, is the PI ring proportionality coefficient, and s is the complex parameter variable of the Laplace transform; Among them, active power deviation After closed-loop transfer function After calculation, output angular velocity compensation amount angular velocity compensation Combined with grid-side angular velocity Generate optimized virtual synchronizer speed Finally, through integration, the corrected virtual phase angle is obtained. .

5. The method according to claim 1, characterized in that: The MPC predictive control uses the zero-order hold method to discretize and obtain the discretized prediction model. in, and These are the matrix coefficients, and their corresponding values ​​are as follows: and , For the integral variable of time, the matrix ,matrix ; Matrix coefficients The element value; Matrix coefficients The element value; The sampling period is This refers to the parasitic resistance of the inductor. For filtering inductors, For filtering capacitors; These are the inverter-side filter inductor currents at time k+1 and time k, respectively, during discrete sampling. These are the filter capacitor voltages at time k+1 and time k, respectively, during discrete sampling; Let k be the grid-side current at time k, which is discretely sampled. Let k be the inverter bridge arm output voltage at time k.

6. The method according to claim 1, characterized in that: The model for the cost function J in the MPC predictive control is as follows: in, These are the weighting coefficients. For the future The predicted value of the filter capacitor voltage at time k, where variable k represents the current discrete sampling time. These represent the switching states of the j-th bridge arm at time k+1 and time k, respectively, with values ​​of 0 or 1; j=1,2,3 correspond to the three-phase bridge arms of the inverter, respectively. For overvoltage and overcurrent constraints, The switching penalty coefficient is... This is the output voltage reference value. This is the reference value for the inverter output current. These are the inverter-side filter inductor currents at time k+1 of the discrete sampling.

7. The method according to claim 6, characterized in that: Overvoltage and overcurrent constraints for: in, and These are the maximum allowable values ​​for the filter capacitor voltage and the inverter-side filter inductor current, respectively. and These represent the predicted values ​​of the filter capacitor voltage and the inverter-side filter inductor current, respectively, corresponding to the i-th prediction step size. and If any one of the above inequalities exists, then overvoltage and overcurrent protection will be implemented.

8. The method according to claim 6, characterized in that: Weighting coefficients are set under normal operating conditions. ; When the inverter-side filter inductor current is detected , Set as , The set overcurrent protection safety threshold for the inverter. These are the minimum and maximum values ​​of the preset weighting coefficients, respectively.

9. A control device based on the method of any one of claims 1-8, characterized in that, include: The acquisition module is used to acquire speed deviations, as well as current and voltage signals from the power grid. The VSG control module is used to dynamically adjust the virtual moment of inertia J and damping coefficient D based on the adaptive VSG control model, and to introduce the dynamically adjusted virtual moment of inertia J and damping coefficient D into the VSG control strategy of the virtual synchronous generator, so as to control the output voltage reference value through VSG control. The MPC control module is used to introduce the output voltage reference value into the MPC predictive control and solve the optimal control quantity by traversing various switching states of the inverter. The PWM drive module is used to output a PWM modulation signal based on the optimal control quantity to control the operation of the inverter switching transistors.

10. A control system, characterized in that: include: The power supply is connected to the power grid via an inverter; The control device as described in claim 9; The inverter is connected to the control device and receives the PWM modulation signal generated by the control device.

Images