A virtual synchronous machine control method and device introducing model predictive control

Abstract

The present application relates to the technical field of power grid-connected control, and particularly relates to a virtual synchronous machine control method and device introducing model predictive control, which comprises the following steps: collecting three-phase voltage and current of a grid-connected point, and calculating electromagnetic power and reactive power; then, through active power-frequency loop and reactive power-voltage loop simulating the characteristics of a synchronous generator, a virtual angular velocity and a reference electromotive force are obtained to synthesize a reference voltage vector; a discrete current prediction model of a two-level converter is established, all eight switching states are enumerated and the next period output current is predicted, the optimal switching signal is obtained through cost function optimization, and is output to a converter driving circuit. Through the present application, the pain points of existing VSG parameter setting complexity, response lag and link redundancy are effectively solved, while the core ability of power grid inertia support is retained, and the system dynamic performance and working condition adaptability are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of power grid connection control technology, and in particular to a virtual synchronous machine control method and apparatus that incorporates model predictive control. Background Technology

[0002] The construction of my country's new power system is accelerating, with the installed capacity and grid connection share of renewable energy sources such as photovoltaics and wind power continuing to rise, resulting in a significant increase in the power system's electronic characteristics. Unlike traditional grids dominated by synchronous generators, new energy units connect to the grid through power electronic converters, which cannot provide natural mechanical inertia and damping support. This leads to a decrease in the overall inertia level of the system, a weakening of its disturbance immunity, and a significant increase in the risk of grid frequency and voltage stability issues, becoming a key problem that urgently needs to be addressed in the construction of the new power system.

[0003] Currently, mainstream virtual synchronous machine control schemes generally adopt the classic architecture of "outer loop power control loop + inner loop PI voltage and current tracking loop + SPWM modulation stage". Among them, the control effect of the PI inner loop is highly dependent on manually tuned control parameters, and the parameter adaptation range is narrow. Under complex operating conditions such as changes in grid impedance, switching between strong and weak grids, and sudden changes in power commands, stability problems such as control overshoot and system oscillation are prone to occur. At the same time, the PI closed-loop regulation has inherent response lag characteristics, making it difficult to achieve fast and accurate tracking of reference quantities, thus limiting the dynamic response performance of the system. The additional SPWM modulation stage further lengthens the control link, which not only increases the implementation complexity of the algorithm, but also introduces additional control delays, further restricting the improvement of virtual synchronous machine control performance. Summary of the Invention

[0004] This invention provides a virtual synchronous machine control method that incorporates model predictive control, which can effectively solve the problems in the background art.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A virtual synchronous machine control method incorporating model predictive control, the method comprising: S1: Collect the three-phase voltage and three-phase current at the grid connection point, calculate the instantaneous active power and instantaneous reactive power, and obtain the electromagnetic power and reactive power after low-pass filtering; S2: Compare the electromagnetic power with the given active power reference value, calculate the virtual angular velocity by simulating the active-frequency loop of the synchronous generator rotor motion equation, and obtain the reference phase by integrating the virtual angular velocity; S3: Compare the reactive power with the given reactive power reference value, and calculate the reference electromotive force by simulating the reactive-voltage loop of the synchronous generator excitation characteristics; S4: Based on the reference electromotive force and reference phase, synthesize the reference voltage vector in the three-phase stationary coordinate system; S5: Initialize the system parameters of model predictive control and establish the current prediction model of the two-level converter in the discrete time domain; S6: Enumerate all eight switching state combinations of the two-level converter, and predict the three-phase output current of the next control cycle for each switching state. S7: Define the cost function, calculate the cost function value for each switch combination, select the switch combination that minimizes the cost function value as the optimal switch signal, output it to the power switch drive circuit of the converter, and return to step S1 at the next sampling time.

[0006] Further, obtaining the electromagnetic power and reactive power includes: At the same sampling moment, the instantaneous values ​​of the three-phase voltage and the three-phase current at the grid connection point are acquired synchronously. The Park transformation is used to transform the three-phase stationary coordinate system to the dq synchronous rotating coordinate system to obtain the dq axis voltage and current; Instantaneous active power and instantaneous reactive power are calculated based on instantaneous power theory. After being filtered by a low-pass filter with a preset cutoff frequency, the electromagnetic power and reactive power are obtained.

[0007] Furthermore, the formulas for calculating the electromagnetic power and reactive power are as follows: ; Among them, P e Electromagnetic power; Reactive power; This is the cutoff frequency of the low-pass filter; For the complex frequency domain operator in the Laplace transform; , These are the d-axis and q-axis voltage components of the three-phase voltage at the grid connection point after Park transformation, in the dq synchronous rotating coordinate system. , These are the d-axis and q-axis current components in the dq synchronous rotating coordinate system after the inverter output three-phase current has undergone Park transformation.

[0008] Furthermore, the control equations for the active-frequency loop are as follows: ; in, ω is the virtual rotor angular velocity of the virtual synchronizer; J is the virtual moment of inertia; D is the virtual damping coefficient. This is the reference value for the active power of the virtual synchronous machine; Rated rotor angular velocity; This refers to the angular velocity of the power grid synchronization.

[0009] Furthermore, the control equations for the reactive-frequency loop are: ; in, This is the reference electromotive force for the virtual synchronizer; The reference electromotive force at the grid connection point; This refers to the reactive power droop factor. This is the reference value for reactive power of the virtual synchronous machine.

[0010] Furthermore, the cost function value is expressed as: ; in, The cost function value; This is a reference value for the three-phase output current of the inverter; This refers to the predicted output three-phase current of the inverter in the next control cycle under different switching state combinations.

[0011] Furthermore, by enumerating all eight switching state combinations of the two-level converter, and for each switching state, predicting the three-phase output current for the next control cycle, including: Define the three-phase bridge arm switching function for a two-level converter; Based on the aforementioned switching function, there are a total of 8 valid switching state combinations for the three-phase two-level converter, each combination corresponding to a unique bridge arm output three-phase voltage; A continuous domain mathematical model is established based on the converter main circuit topology with LC filter. Based on the continuous domain mathematical model and the active-frequency loop and reactive-voltage loop, the three-phase output currents are obtained.

[0012] Furthermore, the continuous domain mathematical model is expressed as: ; in, , , These are the output phase voltages of the three-phase bridge arms a, b, and c of the two-level converter, respectively. The filter inductor is the LC filter on the output side of the converter; This is the parasitic resistance of the filter inductor; , , These are the instantaneous values ​​of the three-phase currents a, b, and c output by the converter, respectively. , , These are the instantaneous phase voltage values ​​of phases a, b, and c on the grid side, respectively.

[0013] Furthermore, the switching function of the two-level converter is expressed as: ; in, This is the switching state function of the k-phase bridge arm of the two-level converter, where k is the phase sequence identifier, corresponding to the a-phase, b-phase, and c-phase three-phase bridge arms of the two-level converter, respectively.

[0014] A virtual synchronous machine control device incorporating model predictive control, employing any one of the methods described above, the device comprising: The main circuit structure includes a DC power supply, a two-level inverter, an LC filter, a line impedance, and an AC power grid. The output terminal of the DC power supply is electrically connected to the DC input terminal of the two-level inverter, and the AC output terminal of the two-level inverter is connected to the AC power grid via the LC filter and the line impedance. The control model includes a signal acquisition unit, a power calculation unit, a VSG outer loop control unit, a model prediction control unit, and a drive output unit. The signal acquisition unit is used to acquire voltage and current signals at the grid connection point. The power calculation unit is used to calculate the electromagnetic power and reactive power of the system. The VSG outer loop control unit is used to generate reference phase, reference electromotive force, and current reference values ​​for grid connection control. The model prediction control unit is used to optimize and obtain the optimal switching control signal. The drive output unit is used to output the optimal switching control signal to the two-level inverter.

[0015] The technical solution of this invention can achieve the following technical effects: This invention effectively solves the problems of high dependence on manual tuning of PI controller parameters, narrow adaptability to power grid conditions, and insufficient operational stability under complex conditions in existing virtual synchronous machine control schemes. It overcomes the inherent response lag defect of traditional PI closed-loop regulation, eliminates the additional SPWM modulation stage, and removes the disadvantages of high algorithm complexity and large control delay caused by control link redundancy. While fully retaining the inertia support and frequency and voltage regulation core capabilities of the virtual synchronous machine, it significantly improves the dynamic response performance and operational stability of the system.

[0016] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating the virtual synchronous machine control method for introducing model predictive control; Figure 2 A schematic diagram illustrating the process for obtaining electromagnetic power and reactive power; Figure 3 A flowchart illustrating the process of predicting the three-phase output current for the next control cycle; Figure 4 A schematic diagram of a virtual synchronous machine control device for introducing model predictive control. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0021] Example 1: like Figure 1 As shown, this application provides a virtual synchronous machine control method incorporating model predictive control, the method comprising: S1: Collect the three-phase voltage and three-phase current at the grid connection point, calculate the instantaneous active power and instantaneous reactive power, and obtain the electromagnetic power and reactive power after low-pass filtering; Specifically, within each fixed sampling control cycle, the instantaneous values ​​of the three-phase grid voltage and the instantaneous values ​​of the three-phase grid current output by the inverter at the inverter grid connection point are synchronously collected by voltage and current sensors. Based on the collected voltage and current sampling values, the instantaneous active power and instantaneous reactive power of the system are calculated. The calculated instantaneous power is then filtered by a low-pass filter with a preset cutoff frequency to remove high-frequency harmonic components, thus obtaining the electromagnetic power and reactive power used for subsequent outer loop control.

[0022] S2: Compare the electromagnetic power with the given active power reference value, calculate the virtual angular velocity by simulating the active-frequency loop of the synchronous generator rotor motion equation, and obtain the reference phase by integrating the virtual angular velocity; Specifically, the electromagnetic power obtained after filtering is compared with the system's preset active power reference value to obtain the active power deviation value. This active power deviation value is then input into the active-frequency control loop of the simulated synchronous generator rotor motion equation. Through the operation and processing of the active-frequency control loop, the virtual rotor angular velocity of the virtual synchronous machine is calculated. The obtained virtual angular velocity is then integrated to obtain the reference phase used for subsequent coordinate transformation and reference quantity synthesis.

[0023] S3: Compare the reactive power with the given reactive power reference value, and calculate the reference electromotive force by simulating the reactive-voltage loop of the synchronous generator excitation characteristics; Specifically, the reactive power obtained after filtering is compared with the system's preset reactive power reference value to obtain the reactive power deviation value. This reactive power deviation value is then input into the reactive-voltage control loop of the simulated synchronous generator excitation regulation characteristics. Through the calculation and processing of the reactive-voltage control loop, the reference electromotive force amplitude of the virtual synchronous machine is calculated.

[0024] S4: Synthesize the reference voltage vector in the three-phase stationary coordinate system based on the reference electromotive force and the reference phase; Specifically, based on the reference phase obtained in step S2 and the reference electromotive force amplitude obtained in step S3, a three-phase reference voltage vector in the three-phase stationary coordinate system is synthesized through sine wave operation. This reference voltage vector is the target voltage waveform that the system expects the inverter to output.

[0025] S5: Initialize the system parameters of model predictive control and establish the current prediction model of the two-level converter in the discrete time domain; The system parameters required for model predictive control are initialized. At the same time, based on the main circuit topology and the initialized system parameters, a current prediction model of the two-level converter in the discrete time domain is established. This model can predict the inverter output current value in the next control cycle based on the current and voltage values ​​at the current sampling time.

[0026] S6: Enumerate all eight switching state combinations of the two-level converter, and predict the three-phase output current of the next control cycle for each switching state. Specifically, all eight switching state combinations of the two-level three-phase converter are enumerated. For each enumerated switching state combination, the three-phase output current of the inverter at the end of the next control cycle is predicted in the form of the discrete current prediction model established in step S5.

[0027] S7: Define the cost function, calculate the cost function value for each switch combination, select the switch combination that minimizes the cost function value as the optimal switch signal, output it to the power switch drive circuit of the converter, and return to step S1 at the next sampling time.

[0028] Specifically, in step S7, a cost function is predefined to evaluate the control effect of the switching state. The cost function has the core optimization objective of minimizing the current tracking error. For the predicted three-phase output current value corresponding to each switching state in step S6, the corresponding cost function value is calculated. After calculating the cost function for all eight switching states, all calculation results are compared, and the switching state combination that minimizes the cost function value is selected as the optimal switching signal for this control cycle. The optimal switching signal is directly output to the drive circuit of the power switching transistors of each bridge arm of the inverter, driving the power switching transistors to perform the corresponding on / off actions, completing the control flow of this control cycle. When the next sampling time arrives, the process returns to step S1 to enter the cyclic control of the next control cycle.

[0029] This invention effectively solves the pain points of existing VSG parameters, such as complex tuning, slow response, and link redundancy. While retaining its core capability of supporting power grid inertia, it significantly improves the dynamic performance and adaptability of the system to operating conditions.

[0030] As a preferred embodiment of the above, such as Figure 2 As shown, the electromagnetic power and reactive power are obtained, including: A10: At the same sampling moment, synchronously acquire the instantaneous values ​​of the three-phase voltage and the three-phase current at the grid connection point; A20: The Park transformation is used to transform the three-phase stationary coordinate system to the dq synchronous rotating coordinate system to obtain the dq axis voltage and current; A30: Based on instantaneous power theory, instantaneous active power and instantaneous reactive power are calculated. After filtering by a low-pass filter with a preset cutoff frequency, electromagnetic power and reactive power are obtained.

[0031] Specifically, at the beginning of each control cycle, the multi-channel ADC sampling module built into the digital controller is synchronously triggered to simultaneously sample the three-phase voltage and three-phase current sampling channels at the grid connection point. This ensures that the instantaneous sampled values ​​of the three-phase grid voltage and the inverter output three-phase grid current are obtained at the same time, avoiding phase errors introduced by the sampling time deviation of different channels, which would lead to deviations in subsequent power calculations. After sampling, the raw data is first preprocessed with software sliding mean filtering to filter out high-frequency switching noise introduced by the power switching transistors during the sampling process, improving the accuracy and stability of the sampled data. Subsequently, the preprocessed three-phase voltage and three-phase current data are processed... Coordinate transformation is performed. First, Clark transform is used to convert the voltage and current data in the three-phase stationary abc coordinate system into corresponding components in the two-phase stationary αβ coordinate system. Then, using the reference phase obtained in step S2 as the phase reference of the synchronous rotating coordinate system, Park transform is used to convert the voltage and current components in the two-phase stationary αβ coordinate system to the dq synchronous rotating coordinate system. Finally, the corresponding d-axis voltage component, q-axis voltage component, d-axis current component, and q-axis current component are obtained. The coordinate transformation prioritizes the equal amplitude transformation criterion to ensure the consistency of signal amplitude before and after the transformation, adapting to the subsequent power calculation and outer loop control operation logic. After the coordinate transformation is completed, based on instantaneous power theory, the obtained dq-axis voltage components and dq-axis current components are used to... The axis current component is calculated to obtain the instantaneous active power and instantaneous reactive power at the current sampling time. The calculated instantaneous active power and instantaneous reactive power are then sent to a low-pass filter with preset parameters for filtering. Here, a first-order Butterworth low-pass filter is preferred, and its cutoff frequency is preferably set in the range of 10Hz to 20Hz, with 15Hz being the optimal setting. This cutoff frequency range can effectively filter out the high-frequency power pulsation components caused by grid harmonics and switching actions, while avoiding the introduction of excessive phase lag due to the filtering stage, ensuring the dynamic response speed of the power signal, and not affecting the control performance of the subsequent active-frequency loop and reactive-voltage loop. After completing the filtering calculation in each sampling period, smooth and stable electromagnetic power and reactive power are output and sent to the active-frequency loop in step S2 and the reactive-voltage loop in step S3 for subsequent control calculations.

[0032] As a preferred embodiment of the above, the formulas for calculating electromagnetic power and reactive power are as follows: ; Among them, P e Electromagnetic power; Reactive power; This is the cutoff frequency of the low-pass filter; For the complex frequency domain operator in the Laplace transform; , These are the d-axis and q-axis voltage components of the three-phase voltage at the grid connection point after Park transformation, in the dq synchronous rotating coordinate system. , These are the d-axis and q-axis current components in the dq synchronous rotating coordinate system after the inverter output three-phase current has undergone Park transformation.

[0033] Specifically, the calculation of electromagnetic power and reactive power is based on the voltage and current components in a synchronous rotating coordinate system. It is achieved by combining instantaneous power theory with first-order low-pass filtering. The calculation process includes two core steps: instantaneous power calculation and low-pass filtering smoothing. In the instantaneous power calculation stage, the dq-axis voltage and current components obtained after Park transformation are used as inputs. According to instantaneous power theory, the active power component is obtained by adding the products of d-axis voltage and d-axis current and the products of q-axis voltage and q-axis current. The reactive power component is obtained by calculating the product of q-axis voltage and d-axis current and the difference between d-axis voltage and q-axis current. These are then multiplied by the coefficients of the three-phase system to obtain the instantaneous active and reactive power at the current sampling time. In the low-pass filtering stage, the calculated instantaneous power is input into a first-order low-pass filter. The transfer function of this filter includes a preset cutoff frequency parameter. Its function is to filter out high-frequency harmonic components and switching ripple interference in the instantaneous power signal, resulting in smooth and stable electromagnetic power and reactive power signals. This calculation method not only ensures the real-time performance of power calculation but also suppresses the impact of high-frequency interference on control performance through the filtering stage, ensuring the stable operation of the subsequent outer loop control stage.

[0034] As a preferred embodiment of the above, the control equation for the active power-frequency loop is: ; in, ω is the virtual rotor angular velocity of the virtual synchronizer; J is the virtual moment of inertia; D is the virtual damping coefficient. This is the reference value for the active power of the virtual synchronous machine; Rated rotor angular velocity; This refers to the angular velocity of the power grid synchronization.

[0035] Specifically, the active power reference value preset by the system is first calculated by subtracting the electromagnetic power obtained after filtering in step S1 to obtain the active power deviation. This deviation reflects the degree of mismatch between the inverter's current output power and the target power. Then, the deviation is divided by the rated rotor angular velocity to complete the per-unit processing, eliminating the scale differences between systems of different power levels and improving the versatility of the control algorithm. The per-unit active power deviation enters the adjustment loop composed of virtual moment of inertia and virtual damping coefficient. The virtual moment of inertia is used to simulate the mechanical inertia of the synchronous generator rotor. The larger its value, the slower the frequency change rate of the system during sudden changes in active power, which can effectively smooth out rapid fluctuations in grid frequency. The virtual damping coefficient simulates the damping winding characteristics of the synchronous generator, which can suppress frequency oscillations caused by active power deviation and improve the dynamic stability of the system. At the same time, the grid synchronous angular velocity is used as a feedforward reference to directly participate in the calculation, enabling the virtual angular velocity to quickly track the reference value of the grid frequency and avoid static frequency deviation. Finally, the virtual angular velocity obtained through the above steps includes both the grid frequency reference component and reflects the adjustment effect of virtual inertia and damping. This angular velocity is then fed into the integration stage, and the reference phase required for subsequent reference voltage synthesis is obtained through integration, thereby realizing the active-frequency regulation function of the virtual synchronous machine and providing active inertia support and primary frequency regulation capability for the grid.

[0036] As a preferred embodiment of the above, the control equation for the reactive power-frequency loop is: ; in, This is the reference electromotive force for the virtual synchronizer; The reference electromotive force at the grid connection point; This refers to the reactive power droop factor. This is the reference value for reactive power of the virtual synchronous machine.

[0037] Specifically, firstly, the system's preset reactive power reference value is compared with the reactive power obtained after filtering in step S1 to obtain the reactive power deviation. This deviation reflects the degree of mismatch between the inverter's current output reactive power and the target reactive power. Then, this reactive power deviation is multiplied by a preset reactive power droop coefficient. The magnitude of the droop coefficient determines the degree of influence of the reactive power deviation on voltage regulation and can be flexibly set according to the grid's voltage regulation requirements. For example, in scenarios with strong voltage support requirements, the droop coefficient can be appropriately increased to improve the sensitivity of voltage regulation. The multiplied deviation is then fed into the integration stage for processing. The integration stage eliminates the static voltage error caused by the reactive power deviation, ensuring the accuracy of voltage regulation and avoiding steady-state voltage deviation. Finally, the output of the integration stage is superimposed with the reference electromotive force (EMF) benchmark value at the grid connection point to obtain the final virtual synchronous machine reference EMF. This reference electromotive force serves as the amplitude benchmark for the subsequent synthesis of the three-phase reference voltage vector. When the grid voltage drops or rises, the reactive power-voltage loop automatically adjusts the magnitude of the reference electromotive force according to the reactive power deviation. For example, when the grid voltage drops, the inverter automatically increases the reactive power output to raise the grid connection point voltage; when the grid voltage is too high, the inverter absorbs reactive power to suppress the voltage rise, thereby achieving the same primary voltage regulation characteristics as a synchronous generator, providing active voltage support capability for the grid, and ensuring the stable operation of the grid connection point voltage.

[0038] As a preferred embodiment of the above, the cost function value is expressed as: ; in, The cost function value; This is a reference value for the three-phase output current of the inverter; This refers to the predicted output three-phase current of the inverter in the next control cycle under different switching state combinations.

[0039] Specifically, firstly, based on the three-phase reference voltage vector synthesized in step S4, combined with the filter parameters of the inverter main circuit and the grid voltage sampling value, the inverter output current reference value in the three-phase stationary coordinate system is calculated. This reference value is the target current waveform that the inverter needs to track in the current control cycle. Subsequently, for each switching state combination enumerated in step S6, combined with the established discrete current prediction model, the predicted value of the three-phase output current for the next control cycle is obtained. For each switching state, the error between the three-phase current reference value and the corresponding predicted value is calculated. The absolute values ​​of the errors of each phase are integrated to obtain the cost function value corresponding to the switching state. The smaller the value, the smaller the deviation between the predicted value and the reference value of the inverter output current in the switching state, and the better the control effect. By iterating and comparing the cost function values ​​of all eight switching states, the switching state combination with the smallest value is selected as the optimal drive signal for this control cycle and directly output to the inverter power switch tube, thereby achieving fast and accurate tracking of the reference current.

[0040] As a preferred embodiment of the above, such as Figure 3 As shown, in step S6, all eight switching state combinations of the two-level converter are enumerated, and for each switching state, the three-phase output current of the next control cycle is predicted, including: S61: Defines the three-phase bridge arm switching function for a two-level converter; S62: Based on the switching function, there are a total of 8 valid switching state combinations for the three-phase two-level converter, and each combination corresponds to a unique bridge arm output three-phase voltage; S63: Based on the converter main circuit topology with LC filter, establish a continuous domain mathematical model; S64: Based on the continuous domain mathematical model and the active-frequency loop and reactive-voltage loop, the output current of each three phase is obtained.

[0041] Specifically, first, step S61 is executed to define the switching function of the three-phase bridge arm of the two-level converter: when the upper bridge arm of a phase is on and the lower bridge arm is off, the switching function value of that phase is 1; otherwise, it is 0. Simultaneously, through hardware and software dual interlocking, complementary conduction of the upper and lower bridge arms of the same phase is ensured, avoiding the risk of shoot-through short circuits. Then, step S62 is executed. Based on the switching function definition, the three-phase bridge arm forms a total of 8 non-repeating effective switching state combinations, including two zero-vector combinations and six effective vector combinations. Each combination corresponds to a unique set of three-phase bridge arm output voltages, providing a complete voltage input set for subsequent current prediction. Next, step S63 is executed, based on the LC... A continuous-domain mathematical model is established for the converter main circuit topology of the filter. The bridge arm output voltage and grid voltage are used as inputs, and the filter inductor current and capacitor voltage are used as state variables. Non-ideal factors such as inductor parasitic resistance are also considered to accurately describe the dynamic relationship between various physical quantities. Finally, step S64 is executed to discretize the continuous-domain model using the forward Euler method to construct a discrete-time current prediction relationship. Combining the grid voltage and inductor current state variables at the current sampling time, the bridge arm voltage corresponding to each switching state is used as the model input to calculate the predicted value of the three-phase output current for the next control cycle. This provides complete data support for subsequent cost function optimization, ensuring that the prediction results accurately reflect the current change trend and guarantee the accuracy of control decisions.

[0042] As a preferred embodiment of the above, the mathematical model of the continuous domain is expressed as follows: ; in, , , These are the output phase voltages of the three-phase bridge arms a, b, and c of the two-level converter, respectively. The filter inductor is the LC filter on the output side of the converter; This is the parasitic resistance of the filter inductor; , , These are the instantaneous values ​​of the three-phase currents a, b, and c output by the converter, respectively. , , These are the instantaneous phase voltage values ​​of phases a, b, and c on the grid side, respectively.

[0043] Specifically, the continuous domain mathematical model in this preferred embodiment is constructed based on the actual topology of the LC filter on the output side of the converter, and fully describes the dynamic relationship between the bridge arm output voltage, grid voltage, and output current. In the model, the three-phase output phase voltage of the bridge arm is directly determined by the current switching state and serves as the driving input of the system; the three-phase phase voltage on the grid side is the back electromotive force provided by the public grid, and its value is obtained by real-time sampling; the filter inductor and its parasitic resistance constitute the impedance of the output circuit, which determines the current variation characteristics. From a physical perspective, the difference between the bridge arm output voltage and the grid voltage must simultaneously overcome the voltage drop of the parasitic resistance and the induced electromotive force of the filter inductor. The magnitude of the induced electromotive force is proportional to the rate of change of current. Therefore, the dynamic change of current is directly determined by the bridge arm voltage, grid voltage, inductance value, and parasitic resistance. In this embodiment, the model preferably considers the influence of inductor parasitic resistance to avoid prediction deviations caused by ideal models and make the model closer to actual working conditions. At the same time, the model provides a foundation for subsequent discretization processing. By discretizing the differential equations using the forward Euler method, the continuous dynamic relationship can be converted into a discrete recursive form. Combined with the current and voltage values ​​at the current sampling time, the predicted current value for the next control cycle can be calculated, providing a reliable theoretical basis for optimizing the switching state in model predictive control.

[0044] As a preferred embodiment of the above, the switching function of the two-level converter is expressed as follows: ; in, This is the switching state function of the k-phase bridge arm of the two-level converter, where k is the phase sequence identifier, corresponding to the a-phase, b-phase, and c-phase three-phase bridge arms of the two-level converter, respectively.

[0045] Specifically, for each phase arm of the converter, there are only two safe complementary operating states: when the upper arm power switch is on and the lower arm power switch is off, the switching function value of that phase is defined as 1; conversely, when the upper arm power switch is off and the lower arm power switch is on, the switching function value of that phase is defined as 0. To avoid a DC bus short circuit caused by simultaneous conduction of the upper and lower arms of the same phase, this embodiment preferably adopts a dual interlocking mechanism of hardware dead-time circuit and software logic to ensure that the switching signals of the upper and lower arms of the same phase always maintain a complementary relationship. A short dead time is set during state switching to eliminate the risk of arm shoot-through. Based on the above switching function definition, the switching state of the three-phase arms can be represented by the combination of the switching function values ​​of phases A, B, and C. Since each phase only has two values, 0 and 1, the three-phase combination forms a total of 8 non-repeating effective switching state combinations, including two zero-vector combinations with completely identical three-phase switching states, and six effective vector combinations with non-repeating switching states. Each combination of switching states corresponds to a unique set of three-phase output voltages for each bridge arm. Under a zero-vector combination, the output voltages of all three bridge arms are zero. Under an effective vector combination, the output voltages of the three bridge arms are determined by both the DC bus voltage and the switching function value. In the current prediction stage of model predictive control, the switching function combination is the direct input for calculating the bridge arm output voltages. By substituting the output voltages corresponding to different switching states into the continuous domain mathematical model, the three-phase output current for the next control cycle can be predicted. This provides complete input data support for the subsequent optimization calculation of the cost function, thereby achieving optimal control decision-making based on switch state enumeration.

[0046] Example 2: Based on the same inventive concept as the virtual synchronous machine control method that incorporates model predictive control in the aforementioned embodiments, such as... Figure 4 As shown, the present invention also provides a virtual synchronous machine control device incorporating model predictive control, comprising: The main circuit structure includes a DC power supply, a two-level inverter, an LC filter, a line impedance, and an AC power grid. The output terminal of the DC power supply is electrically connected to the DC input terminal of the two-level inverter, and the AC output terminal of the two-level inverter is connected to the AC power grid via an LC filter and a line impedance. The control model includes a signal acquisition unit, a power calculation unit, a VSG outer loop control unit, a model prediction control unit, and a drive output unit. The signal acquisition unit is used to acquire voltage and current signals at the grid connection point. The power calculation unit is used to calculate the electromagnetic power and reactive power of the system. The VSG outer loop control unit is used to generate the reference phase, reference electromotive force, and current reference value for grid connection control. The model prediction control unit is used to find the optimal switching control signal. The drive output unit is used to output the optimal switching control signal to the two-level inverter.

[0047] The control device described above in this invention can effectively implement the virtual synchronous machine control method that introduces model predictive control, and the technical effects it can achieve are as described in the above embodiments, and will not be repeated here.

[0048] Similarly, the above-mentioned optimization schemes for the system can also achieve the optimization effects corresponding to the methods in Embodiment 1, which will not be repeated here.

[0049] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A virtual synchronous machine control method incorporating model predictive control, characterized in that, The method includes: S1: Collect the three-phase voltage and three-phase current at the grid connection point, calculate the instantaneous active power and instantaneous reactive power, and obtain the electromagnetic power and reactive power after low-pass filtering; S2: Compare the electromagnetic power with the given active power reference value, calculate the virtual angular velocity by simulating the active-frequency loop of the synchronous generator rotor motion equation, and obtain the reference phase by integrating the virtual angular velocity; S3: Compare the reactive power with the given reactive power reference value, and calculate the reference electromotive force by simulating the reactive-voltage loop of the synchronous generator excitation characteristics; S4: Based on the reference electromotive force and reference phase, synthesize the reference voltage vector in the three-phase stationary coordinate system; S5: Initialize the system parameters of model predictive control and establish the current prediction model of the two-level converter in the discrete time domain; S6: Enumerate all eight switching state combinations of the two-level converter, and predict the three-phase output current of the next control cycle for each switching state. S7: Define the cost function, calculate the cost function value for each switch combination, select the switch combination that minimizes the cost function value as the optimal switch signal, output it to the power switch drive circuit of the converter, and return to step S1 at the next sampling time.

2. The virtual synchronous machine control method incorporating model predictive control according to claim 1, characterized in that, Obtaining the electromagnetic power and reactive power includes: At the same sampling moment, the instantaneous values ​​of the three-phase voltage and the three-phase current at the grid connection point are acquired synchronously. The Park transformation is used to transform the three-phase stationary coordinate system to the dq synchronous rotating coordinate system to obtain the dq axis voltage and current; Instantaneous active power and instantaneous reactive power are calculated based on instantaneous power theory. After being filtered by a low-pass filter with a preset cutoff frequency, the electromagnetic power and reactive power are obtained.

3. The virtual synchronous machine control method incorporating model predictive control according to claim 2, characterized in that, The formulas for calculating electromagnetic power and reactive power are as follows: ； Among them, P e Electromagnetic power; Reactive power; This is the cutoff frequency of the low-pass filter; For the complex frequency domain operator in the Laplace transform; , These are the d-axis and q-axis voltage components of the three-phase voltage at the grid connection point after Park transformation, in the dq synchronous rotating coordinate system. , These are the d-axis and q-axis current components in the dq synchronous rotating coordinate system after the inverter output three-phase current has undergone Park transformation.

4. The virtual synchronous machine control method incorporating model predictive control according to claim 1, characterized in that, The control equations for the active-frequency loop are as follows: ； in, ω is the virtual rotor angular velocity of the virtual synchronizer; J is the virtual moment of inertia; D is the virtual damping coefficient. This is the reference value for the active power of the virtual synchronous machine; Rated rotor angular velocity; This refers to the angular velocity of the power grid synchronization.

5. The virtual synchronous machine control method incorporating model predictive control according to claim 1, characterized in that, The control equation for the reactive-frequency loop is: ； in, This is the reference electromotive force for the virtual synchronizer; The reference electromotive force at the grid connection point; This refers to the reactive power droop factor. This is the reference value for reactive power of the virtual synchronous machine.

6. The virtual synchronous machine control method incorporating model predictive control according to claim 1, characterized in that, The cost function value is expressed as: ； in, The cost function value; This is a reference value for the three-phase output current of the inverter; This refers to the predicted output three-phase current of the inverter in the next control cycle under different switching state combinations.

7. The virtual synchronous machine control method incorporating model predictive control according to claim 1, characterized in that, Enumerate all eight switching state combinations of the two-level converter, and for each switching state, predict the three-phase output current of the next control cycle, including: Define the three-phase bridge arm switching function for a two-level converter; Based on the aforementioned switching function, there are a total of 8 valid switching state combinations for the three-phase two-level converter, each combination corresponding to a unique bridge arm output three-phase voltage; A continuous domain mathematical model is established based on the converter main circuit topology with LC filter. Based on the continuous domain mathematical model and the active-frequency loop and reactive-voltage loop, the three-phase output currents are obtained.

8. The virtual synchronous machine control method incorporating model predictive control according to claim 7, characterized in that, The continuous domain mathematical model is expressed as follows: ； in, , , These are the output phase voltages of the three-phase bridge arms a, b, and c of the two-level converter, respectively. The filter inductor is the LC filter on the output side of the converter; This is the parasitic resistance of the filter inductor; , , These are the instantaneous values ​​of the three-phase currents a, b, and c output by the converter, respectively. , , These are the instantaneous phase voltage values ​​of phases a, b, and c on the grid side, respectively.

9. The virtual synchronous machine control method incorporating model predictive control according to claim 7, characterized in that, The switching function of the two-level converter is expressed as follows: ； in, This is the switching state function of the k-phase bridge arm of the two-level converter, where k is the phase sequence identifier, corresponding to the a-phase, b-phase, and c-phase three-phase bridge arms of the two-level converter, respectively.

10. A virtual synchronous machine control device incorporating model predictive control, employing the method described in any one of claims 1-9, characterized in that, The device includes: The main circuit structure includes a DC power supply, a two-level inverter, an LC filter, a line impedance, and an AC power grid. The output terminal of the DC power supply is electrically connected to the DC input terminal of the two-level inverter, and the AC output terminal of the two-level inverter is connected to the AC power grid via the LC filter and the line impedance. The control model includes a signal acquisition unit, a power calculation unit, a VSG outer loop control unit, a model prediction control unit, and a drive output unit. The signal acquisition unit is used to acquire voltage and current signals at the grid connection point. The power calculation unit is used to calculate the electromagnetic power and reactive power of the system. The VSG outer loop control unit is used to generate reference phase, reference electromotive force, and current reference values ​​for grid connection control. The model prediction control unit is used to optimize and obtain the optimal switching control signal. The drive output unit is used to output the optimal switching control signal to the two-level inverter.

Images