Grid-connected converter control method and system based on DLQR

Abstract

The invention discloses a DLQR-based grid-connected converter control method and system, and belongs to the field of converter control, and the method comprises the steps: determining a system state variable, solving a state feedback gain matrix based on a DLQR strategy, and determining a state variable control signal through the state feedback gain matrix; constructing a polyhedral uncertain state space model, and sending the error between the network access current and the network access current reference quantity into the polyhedral uncertain state space model to respectively obtain a fundamental wave state vector and a harmonic wave state vector; the fundamental wave state vector and the harmonic wave state vector are respectively weighted by a state feedback gain matrix to obtain a fundamental wave control signal and a harmonic wave control signal; and the state variable control signal, the fundamental wave control signal and the harmonic wave control signal are synthesized and optimized through a feedback loop, and an optimized control law is obtained. The method provided by the invention has good robust stability, harmonic suppression capability and dynamic performance, is convenient to design, and can effectively consider the dynamic performance and the harmonic suppression capability of the converter under the weak power grid.

Description

Technical Field

[0001] This invention relates to the field of converter control technology, and in particular to a grid-connected converter control method and system based on DLQR. Background Technology

[0002] In renewable energy grid-connected systems, the common coupling point exhibits time-varying grid impedance and rich voltage harmonic characteristics due to factors such as long-distance transmission and distribution lines and leakage inductance of step-up transformers. This seriously threatens the high-quality and stable operation of the interface converter between renewable energy and the grid. With the increase in microprocessor clock frequency, digital control delay reduction technologies such as multi-sampling and real-time sampling can effectively improve converter stability and its adaptability to the grid. However, the introduction of non-integer beat delays poses a significant challenge to the discrete-domain modeling of grid-connected systems and the application of robust control technologies under parameter perturbations. There is an urgent need to construct a stability analysis model and propose robust current control strategies for grid-connected systems with arbitrary beat delays under weak grid access.

[0003] In existing technologies, traditional methods for describing the impact of digital control delays on grid-connected converter modeling have significant limitations: First, while establishing a state-space model of an arbitrary time-delay system with transcendental functions in the time domain can characterize the stability constraint boundaries of the time-delay system through analytical eigenvalues, the time-domain analysis of high-order systems is extremely difficult. Second, constructing a complex frequency domain transfer function model of the time-delay system using exponential functions neglects the amplitude-frequency characteristics of the zero-order hold, leading to high conservatism in stability analysis and controller design. Third, establishing a discrete-domain transfer function model of the grid-connected system with integer time-delay based on the real displacement theorem and solving the problem of accurately describing the characteristics of arbitrary time delays in the discrete domain through improved Z-transforms involves complex residue calculations and is limited to applications of classical control theory. In summary, state-space modeling of grid-connected systems with arbitrary time-delays that take into account grid impedance perturbations has not been fully explored and is difficult to meet the application requirements of modern robust control technologies.

[0004] Furthermore, existing research on robust current control of grid-connected converters in weak grid environments is largely based on mathematical models of uncertain systems. This involves constructing matrix norm constraints to characterize the control objective and employing numerical analysis tools to obtain a robust, closed-loop stable controller. However, such methods cannot be directly applied to uncertain systems, and the state feedback gain matrix in the continuous domain is difficult to analyze when considering the effects of digital control delays. Therefore, existing linear quadratic regulators (LQRs) for uncertain systems struggle to simultaneously consider the dynamic response of grid-connected converters and their adaptability to weak grid environments. Robust current control of uncertain systems with arbitrary time delays requires further research. Summary of the Invention

[0005] The purpose of this application is to overcome the shortcomings of the prior art and provide a grid-connected converter control method and system based on DLQR, which considers a robust control strategy with arbitrary digital control delay to solve the problems of robust stability and current harmonic suppression of grid-connected converters under time-varying power grids.

[0006] In a first aspect, this application provides a grid-connected converter control method based on DLQR, comprising the following steps: The inductor current, grid-connected current, and capacitor voltage of the grid-connected converter are collected to obtain system state variables. The state feedback gain matrix is ​​solved based on the discrete-time linear quadratic regulator (DLQR) strategy. The system state variables are weighted by the first state feedback gain matrix to obtain the state variable control signal. The grid-connected current is compared with a grid-connected current reference value to obtain an error signal; A polyhedral uncertain state space model is constructed, and the error signal is fed into the polyhedral uncertain state space model. After processing by the polyhedral uncertain state space model, the fundamental wave state vector and the harmonic state vector are obtained respectively. The fundamental state vector is weighted by the second state feedback gain matrix to obtain the fundamental control signal; The harmonic state vector is weighted by the third state feedback gain matrix to obtain the harmonic control signal; The state variable control signal, the fundamental wave control signal, and the harmonic control signal are synthesized to obtain a control law. The control law is then optimized through a feedback loop, and the optimized control law is transmitted to the drive module. The drive module generates a pulse width modulation signal to drive the switching devices in the grid-connected converter.

[0007] Optionally, the inductor current, grid-connected current, and capacitor voltage of the grid-connected converter are collected to obtain system state variables. The state feedback gain matrix is ​​solved based on the DLQR strategy. The system state variables are then weighted using the first state feedback gain matrix to obtain the state variable control signal, including: Collect the inductor current, grid-connected current, and capacitor voltage of the grid-connected converter to obtain the system state variables; Set the basic parameters of the grid-connected converter, and obtain the perturbation range of the uncertain parameters based on the basic parameters; Based on the perturbation range of the basic parameters and the uncertain parameters, a closed-loop space model of the uncertain state under arbitrary-shot digital control delay is constructed. Set up scalar and weight matrices, determine the system output based on the weight matrices, and construct a DLQR cost function based on the weight matrices and the system output; An optimization problem and linear matrix inequality constraints are established with the goal of minimizing the DLQR cost function. The optimization problem is solved using simulation tools. It is then determined whether the optimization problem has a solution. If there is no solution, the scalar is adjusted and the optimization problem is solved again. If there is a solution, it is further determined whether the obtained result satisfies the linear matrix inequality constraints. If not, the scalar is adjusted and the optimization problem is solved again. If it satisfies, the state feedback gain matrix is ​​obtained. The state feedback gain matrix includes a first state feedback gain matrix, a second state feedback gain matrix, and a third state feedback gain matrix. The system state variables are weighted by the first state feedback gain matrix to obtain the state variable control signal.

[0008] Optionally, the DLQR cost function J The expression is: in, Represents the mathematical expectation. This represents the number of discrete-time steps used to calculate the cost function. Indicates system output, express k augmented state variables at time t, Represents the control law. Q This is the first weight matrix. R This is the second weight matrix.

[0009] Optionally, the linear matrix inequality constraint is: in, F The matrices related to system state transitions and transformations. G A matrix relating control inputs to system state. L To characterize the symmetric positive definite class matrix of system performance / stability constraints, S i For the convex polyhedron i A symmetric positive definite matrix with vertices S j For the convex polyhedron j A symmetric positive definite matrix with vertices It is the identity matrix. For the convex polyhedron i An augmented open-loop matrix with vertices, For the convex polyhedron i The control input matrix of each vertex For the convex polyhedron i The perturbation input matrix of each vertex. μ As the first scalar, σ For the second scalar, δ It is the third scalar. , , n This represents the number of resonant controllers in the state-space model of the resonant controller. Q This is the first weight matrix. R This is the second weight matrix. The parameter is uncertain.

[0010] Optionally, a polyhedral uncertain state space model is constructed, and the error signal is fed into the polyhedral uncertain state space model. After processing by the polyhedral uncertain state space model, the fundamental wave state vector and harmonic state vector are obtained, including: The state-space model of the resonant controller is extended to arbitrary-time digital control to obtain a polyhedral uncertain state-space model, which includes the fundamental branch and the harmonic branch. The error signal is fed into the polyhedral uncertain state space model, and the fundamental state vector is obtained through the fundamental branch. The error signal is passed through the harmonic branch in the polyhedral uncertain state space model to obtain the harmonic state vector.

[0011] Optionally, the polyhedral uncertain state-space model is expressed as: in, for k The system state variables at time +1, for k System state variables at any given time, Let be the open-loop state transition matrix. For grid-connected converter voltage input matrix, For the grid voltage input matrix, This represents the external disturbance vector. For the mapping matrix, This is the output vector of the polyhedral uncertain state-space model. The parameter is uncertain.

[0012] Optionally, the optimized control law The expression is: in, , The first state feedback gain matrix, For feedback matrix; , This is the second-state feedback gain matrix. This is the third-state feedback gain matrix; For system state variables; This is the state vector after resonant control; This is a 1-row, 14-column state feedback gain matrix; for k Augmented state variables at time step.

[0013] Optionally, it also includes constructing a transfer function from the grid-connected current reference quantity to the grid-connected current and plotting a Bode plot, wherein the grid-connected current reference quantity... i gref ( k ) to grid current i g ( k transfer function G ref_ig ( z The expression for ) is: in, To augment the output matrix, For delay operators, It is the identity matrix. This is the state matrix of the closed-loop space model. As a reference input matrix, The parameter is uncertain.

[0014] Optionally, it also includes constructing a transfer function from grid voltage to grid current, calculating the output impedance of the grid-connected converter based on the transfer function from grid voltage to grid current, and plotting a Bode diagram; The grid voltage v g ( k ) to grid current i g ( k transfer function G vg_ig ( z The expression is: in, To augment the output matrix, For delay operators, It is the identity matrix. The state matrix of the closed-loop space model. The perturbation input matrix is... These are uncertain parameters; The output impedance of the grid-connected converter Z o The expression is: .

[0015] Secondly, this application also provides a DLQR-based grid-connected converter control system, applied to a DLQR-based grid-connected converter control method as described in any one of the first aspects. The DLQR-based grid-connected converter control system is connected to the grid-connected converter. The DLQR-based grid-connected converter control system includes: a drive module, a discrete linear quadratic regulator control module, a phase-locked loop (PLL) module, an inductor current acquisition module, a capacitor voltage acquisition module, a grid-connected current acquisition module, and a common coupling point voltage acquisition module. The first end of the drive module is connected to the first end of the grid-connected converter, and the second end of the drive module is connected to the first end of the discrete linear quadratic regulator control module. The second end of the discrete linear quadratic regulator control module is connected to the PLL module. The first end of the block; the first end of the inductor current acquisition module is connected to the second end of the grid-connected converter, and the second end of the inductor current acquisition module is connected to the third end of the discrete linear quadratic regulator control module; the first end of the capacitor voltage acquisition module is connected to the third end of the grid-connected converter, and the second end of the capacitor voltage acquisition module is connected to the fourth end of the discrete linear quadratic regulator control module; the first end of the grid-connected current acquisition module is connected to the fourth end of the grid-connected converter, and the second end of the grid-connected current acquisition module is connected to the fifth end of the discrete linear quadratic regulator control module; the first end of the common coupling point voltage acquisition module is connected to the fifth end of the grid-connected converter, and the second end of the common coupling point voltage acquisition module is connected to the second end of the phase-locked loop module.

[0016] This application provides a grid-connected converter control method and system based on DLQR, which has good robust stability, harmonic suppression capability, and dynamic performance. It can ensure the stable operation of the grid-connected converter under uncertain parameter conditions and achieve error-free tracking of the current reference. The output impedance of the grid-connected converter has a high amplitude response at the fundamental frequency, which can effectively suppress background harmonics of the power grid and reduce the distortion rate of the grid current. The control method of this application is easy to design and can effectively balance the dynamic performance and harmonic suppression capability of the converter under weak power grid conditions. It has strong engineering applicability and is easy to implement and debug.

[0017] To make the above-mentioned features and advantages of the invention more apparent and understandable, specific embodiments are described below, and detailed descriptions are provided in conjunction with the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the structure of a grid-connected converter control system based on DLQR provided in one embodiment of this application.

[0020] Figure 2 This is a control block diagram of a grid-connected converter control method based on DLQR provided in one embodiment of this application.

[0021] Figure 3 This is a flowchart of step S1 in a DLQR-based grid-connected converter control method provided in one embodiment of this application.

[0022] Figure 4 This is a flowchart of step S3 in a DLQR-based grid-connected converter control method provided in one embodiment of this application.

[0023] Figure 5 This is a closed-loop pole diagram of the control method under uncertain parameters in a grid-connected converter control method based on DLQR provided in one embodiment of this application.

[0024] Figure 6 for Figure 5 A magnified view of region D in the middle.

[0025] Figure 7 This is a Bode plot of the transfer function from the grid-connected current reference quantity to the grid-connected current in a grid-connected converter control method based on DLQR provided in one embodiment of this application.

[0026] Figure 8 This is a Bode plot of the output impedance of the grid-connected converter in a DLQR-based grid-connected converter control method provided in one embodiment of this application.

[0027] Figure 9 This is a steady-state waveform diagram of the single-phase grid-connected current in a grid-connected converter control method based on DLQR provided in one embodiment of this application.

[0028] Figure 10 This is a transient waveform diagram of single-phase grid-connected current in a grid-connected converter control method based on DLQR provided in one embodiment of this application.

[0029] Figure 11 This is a simulation waveform of the full-load grid-connected current of the grid-connected converter in the public power grid in a grid-connected converter control method based on DLQR provided in one embodiment of this application.

[0030] Figure 12 This is a simulation waveform of the grid-connected converter's full-load grid-connected current with a short-circuit ratio of 5 in a grid-connected converter control method based on DLQR provided in one embodiment of this application.

[0031] Figure 13 This is a simulation waveform of the full-load grid-connected current of the grid-connected converter in a DLQR-based grid-connected converter control method provided in one embodiment of this application, with a short-circuit ratio of 2.5.

[0032] Figure 14 The simulation waveform of the full-load grid-connected current of the grid-connected converter in the public power grid when adopting a quasi-proportional resonance plus harmonic compensation control strategy is shown.

[0033] Figure 15 The simulation waveform of the grid-connected converter with a short-circuit ratio of 5 when using a quasi-proportional resonance plus harmonic compensation control strategy is shown.

[0034] Figure 16 The simulation waveform of the grid-connected converter under full load grid current with a short-circuit ratio of 2.5 is shown when a quasi-proportional resonance control strategy with harmonic compensation is adopted. Detailed Implementation

[0035] To make the objectives and technical solutions of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the described embodiments of this application without creative effort are within the scope of protection of this application.

[0036] In one embodiment, see Figure 1This application provides a DLQR-based grid-connected converter control system 1. The DLQR-based grid-connected converter control system 1 is connected to a grid-connected converter 2 and is used to control the grid-connected converter 2. The DLQR-based grid-connected converter control system 1 may include: a drive module 11, a discrete linear quadratic regulator control module 12, a phase-locked loop module 13, an inductor current acquisition module 14, a capacitor voltage acquisition module 15, a grid-connected current acquisition module 16, and a common coupling point voltage acquisition module 17. The first end of the drive module 11 is connected to the first end of the grid-connected converter 2, and the second end of the drive module 11 is connected to the first end of the discrete linear quadratic regulator control module 12. The second end of the discrete linear quadratic regulator control module 12 is connected to the phase-locked loop module 13. The first terminal of the inductor current acquisition module 14 is connected to the second terminal of the grid-connected converter 2, and the second terminal of the inductor current acquisition module 14 is connected to the third terminal of the discrete linear quadratic regulator control module 12; the first terminal of the capacitor voltage acquisition module 15 is connected to the third terminal of the grid-connected converter 2, and the second terminal of the capacitor voltage acquisition module 15 is connected to the fourth terminal of the discrete linear quadratic regulator control module 12; the first terminal of the grid-connected current acquisition module 16 is connected to the fourth terminal of the grid-connected converter 2, and the second terminal of the grid-connected current acquisition module 16 is connected to the fifth terminal of the discrete linear quadratic regulator control module 12; the first terminal of the common coupling point voltage acquisition module 17 is connected to the fifth terminal of the grid-connected converter 2, and the second terminal of the common coupling point voltage acquisition module 17 is connected to the second terminal of the phase-locked loop module 13.

[0037] As an example, grid-connected converter 2 can be a single-phase inductor-capacitor-inductor (LCL) type filter three-level grid-connected converter.

[0038] As an example, the grid-connected converter 2 may include a DC voltage module 21, a grid-connected converter module 22, an LCL filter module 23, a common coupling point 24, a grid impedance module 25, and a grid equivalent voltage source module 26. The first terminal of the DC voltage module 21 is connected to the first terminal of the grid-connected converter module 22, and the second terminal of the DC voltage module 21 is connected to the second terminal of the grid-connected converter module 22. The third terminal of the grid-connected converter module 22 is connected to the first terminal of the LCL filter module 23, and the second terminal of the LCL filter module 23 is connected to the first terminal of the common coupling point 24. The second terminal of the common coupling point 24 is connected to the first terminal of the grid impedance module 25, and the second terminal of the grid impedance module 25 is connected to the first terminal of the grid impedance module 26. The grid equivalent voltage source module 26 is connected; the third terminal of the common coupling point 24 is connected to the first terminal of the grid-connected converter control system 1 based on DLQR; the second terminal of the LCL filter module 23 and the first terminal of the common coupling point 24 are connected to the second terminal of the grid-connected converter control system 1 based on DLQR; the third terminal of the LCL filter module 23 is connected to the third terminal of the grid-connected converter control system 1 based on DLQR; the third terminal of the grid-connected converter module 22 and the first terminal of the LCL filter module 23 are connected to the fourth terminal of the grid-connected converter control system 1 based on DLQR; and the fourth terminal of the grid-connected converter module 22 is connected to the fifth terminal of the grid-connected converter control system 1 based on DLQR.

[0039] As an example, DC voltage module 21 may include a DC power supply. V dc .

[0040] As an example, the third terminal of the grid-connected converter module 22 can be an output terminal, and the output terminal of the grid-connected converter module 22 can be the bridge arm voltage output point. v inv .

[0041] As an example, LCL filter module 23 may include a first inductor. L 1. Second inductor L 2. First capacitor C f First Inductor L The first terminal of 1 is connected to the bridge arm voltage output point of the grid-connected converter module 22. v inv First Inductor L The second terminal of 1 is connected to the second inductor. L 2's first terminal, second inductor L The second end of 2 is connected to the first end of the common coupling point 24, and the first capacitor C f The first end is connected to the first inductor L The second terminal of 1 and the second inductor L 2's first terminal, the first capacitor Cf The second end is grounded.

[0042] As an example, the current flows through the first inductor L The current in 1 is the first inductor current. i L1 Flow through the second inductor L The current in 2 is the second inductor current. i L2 .

[0043] As an example, the first capacitor C f It can be a filter capacitor, a second inductor. L 2 can be the grid-side filter inductor, the first inductor. L 1 can be the inductor on the grid-connected converter side.

[0044] As an example, the current flows through the first capacitor C f The current is the filter capacitor current. i c First capacitor C f The voltage across the terminals is the voltage of the filter capacitor. v c .

[0045] As an example, the grid impedance module 25 may include grid impedance. L g Grid impedance L g The first end connects to the second end of the common coupling point 24, and the mains impedance is... L g The second end is connected to the grid equivalent voltage source module 26.

[0046] As an example, the power grid impedance L g That is, grid-side inductance, grid impedance L g With the second inductor L 2 can form a grid-side inductor L T .

[0047] As an example, the grid equivalent voltage source module 26 may include a grid equivalent voltage source. v g .

[0048] As an example, the inductor current acquisition module 14 acquires harmonic inductor current. i L The capacitor voltage acquisition module 15 acquires the voltage of the filter capacitor. v c The grid-connected current acquisition module 16 acquires the grid-connected current.i g The common coupling point voltage acquisition module 17 acquires the common coupling point voltage of common coupling point 24. v pcc The phase signal synchronized with the grid voltage is obtained through the phase-locked loop module 13. It provides synchronization information for modules such as the discrete linear quadratic regulator control module 12 to ensure phase matching during grid connection.

[0049] As an example, the discrete linear quadratic regulator control module 12 receives harmonic inductor current. i L Filter capacitor voltage v c Grid-connected current i g The discrete linear quadratic regulator control module 12 can also receive a phase signal synchronized with the grid voltage obtained from the phase-locked loop module 13. This feedback signal reflects the operating status of the main circuit. It is used to achieve synchronous control with the power grid.

[0050] As an example, the drive module 11 receives the control signal output by the discrete linear quadratic regulator control module 12, generates a pulse width modulation (PWM) signal, and uses it to drive the switching devices in the grid-connected converter module 22 to control the output of the grid-connected converter module 22.

[0051] As an example, the control signal output by the discrete linear quadratic regulator control module 12 may include harmonic inductor current. i L Filter capacitor voltage v c Grid-connected current i g .

[0052] In the aforementioned grid-connected converter control system based on DLQR, the harmonic inductor current is acquired in real time through the inductor current acquisition module 14. i L The capacitor voltage acquisition module 15 captures the voltage of the filtered capacitor. v c 16. Grid current acquisition module monitors the grid current. i g Common coupling point voltage acquisition module 17 acquires the common coupling point voltage. v pcc The phase-locked loop module 13 generates a power grid synchronization phase signal. This allows the discrete linear quadratic regulator control module 12 to fully grasp the operating status of the main circuit, accurately achieve phase matching with the power grid, and avoid grid-connected phase mismatch problems. Through the comprehensive processing of multiple feedback signals by the discrete linear quadratic regulator control module 12, combined with the output control signal, the drive module 11 generates a pulse width modulation signal, which can accurately control the switching devices of the grid-connected converter module 22. With the filtering effect of the LCL filter module 23, current harmonics are effectively suppressed, the output characteristics of the grid-connected converter are optimized, and the grid-connected current is more stable. The overall grid-connected converter control system based on DLQR, through closed-loop design, can significantly improve control accuracy and grid-connected reliability, reduce interference to the power grid, and ensure stable operation under different operating conditions.

[0053] In another embodiment, please refer to Figure 2 This application also provides a grid-connected converter control method based on DLQR, which may include the following steps: steps S1 to S6.

[0054] Step S1: Collect the inductor current, grid-connected current, and capacitor voltage of the grid-connected converter to obtain the system state variables. Solve the state feedback gain matrix based on the DLQR strategy. Weight the system state variables through the first state feedback gain matrix to obtain the state variable control signal.

[0055] Step S2: Compare the grid-connected current with the grid-connected current reference value to obtain the error signal.

[0056] Step S3: Construct a polyhedral uncertain state space model. Input the error signal into the polyhedral uncertain state space model. After processing by the polyhedral uncertain state space model, the fundamental wave state vector and harmonic state vector are obtained respectively.

[0057] Step S4: The fundamental state vector is weighted by the second state feedback gain matrix to obtain the fundamental control signal.

[0058] Step S5: The harmonic state vector is weighted by the third state feedback gain matrix to obtain the harmonic control signal.

[0059] Step S6: The state variable control signal, fundamental wave control signal, and harmonic control signal are synthesized to obtain the control law. The control law is then optimized through a feedback loop. The optimized control law is then transmitted to the drive module, which generates a pulse width modulation signal to drive the switching devices in the grid-connected converter.

[0060] In the DLQR-based grid-connected converter control method of this application, by weighting the fundamental and harmonic state vectors through the corresponding state feedback gain matrices, independent and precise control of the fundamental and harmonic components can be achieved, significantly reducing the harmonic distortion rate of the grid-connected current and improving power quality. By synthesizing multi-dimensional control signals and driving the switching devices after closed-loop optimization, dynamic adjustment of the control law can be achieved, ensuring that the converter can still operate stably and efficiently under complex operating conditions, ultimately improving the overall control accuracy and reliability of the grid-connected converter.

[0061] In step S1, please refer to Figure 2 In step S1, the inductor current, grid-connected current, and capacitor voltage of the grid-connected converter are collected to obtain the system state variables. The state feedback gain matrix is ​​solved based on the DLQR strategy. The system state variables are weighted by the first state feedback gain matrix to obtain the state variable control signal.

[0062] For example, please refer to Figure 3 Step S1 may include the following steps: Step S11 to Step S16.

[0063] Step S11: Collect the inductor current, grid-connected current, and capacitor voltage of the grid-connected converter to obtain the system state variables.

[0064] Step S12: Set the basic parameters of the grid-connected converter, and obtain the perturbation range of the uncertain parameters based on the basic parameters.

[0065] Step S13: Construct a closed-loop spatial model of the uncertain state under arbitrary-shot digital control delay based on the perturbation range of the basic parameters and the uncertain parameters.

[0066] Step S14: Set the scalar and weight matrices, determine the system output based on the weight matrices, and construct the DLQR cost function based on the weight matrices and the system output.

[0067] Step S15: Establish an optimization problem and linear matrix inequality constraints with the goal of minimizing the DLQR cost function. Solve the optimization problem using simulation tools and determine whether the optimization problem has a solution. If there is no solution, adjust the scalar and solve the optimization problem again. If there is a solution, further determine whether the obtained result satisfies the linear matrix inequality constraints. If not, adjust the scalar and solve the optimization problem again. If it satisfies, obtain the state feedback gain matrix, which includes the first state feedback gain matrix, the second state feedback gain matrix, and the third state feedback gain matrix.

[0068] Step S16: Weight the system state variables using the first state feedback gain matrix to obtain the state variable control signal.

[0069] Specifically, in step S11, the inductor current acquisition module 14 is used to acquire the inductor current of the grid-connected converter 2 in real time. i L (i.e., harmonic inductor current) i L The capacitor voltage is acquired using the capacitor voltage acquisition module 15. v c (i.e., the voltage of the filter capacitor) v c The grid-connected current is collected through the grid-connected current acquisition module 16. i g , to inductor current i L Grid-connected current i g capacitor voltage v c The system state variables are obtained by integration. .

[0070] Furthermore, in step S12, the basic parameters of the grid-connected converter are set according to the actual situation to provide initial conditions for subsequent design. The basic parameters of the grid-connected converter may include the grid voltage. v g DC voltage V dc Grid frequency f 0. Switching frequency f sw Sampling frequency f s Inductor on the grid-connected converter side L 1. Filter capacitor C f , grid-side inductor L g Maximum grid impedance and minimum grid impedance.

[0071] Furthermore, uncertain parameters are obtained based on the basic parameters. Perturbation range L 2, L Tmax ], used to define the boundary of parameter uncertainty, where, L 2 is the grid-side filter inductor. L Tmax This represents the maximum value of the grid-side inductance.

[0072] As an example, the maximum value of the grid-side inductance. L Tmax The expression is: L Tmax =L 2+ Lgmax in, L 2 is the grid-side filter inductor. L gmax This represents the maximum value of the grid-side inductance.

[0073] As an example, uncertain parameters It can be set as the grid-side inductance in the basic parameters of the grid-connected converter. L g Filter capacitor C f Inverter-side inductor L 1, etc., one or more.

[0074] The following uses uncertain parameters For grid-side inductance L g Taking the perturbation as an example, this paper specifically introduces the grid-connected converter control method based on DLQR.

[0075] As an example, in step S13, based on the basic parameters and uncertain parameters of the grid-connected converter... Perturbation range L 2, L Tmax Construct a closed-loop spatial model that can describe the uncertain state under arbitrary digital control delay, in order to accurately characterize the dynamic behavior of grid-connected converters.

[0076] As an example, the expression for the closed-loop space model of the uncertain state under arbitrary digital control delay is as follows: in, for k Augmented state variables at time +1; for k Augmented state variables at time t; This is the state matrix of the closed-loop spatial model, used to describe the uncertain parameters. θ Lower augmented state variables ρ ( k From time k arrive k The closed-loop evolution law of +1; The perturbation input matrix; The external disturbance vector; The reference input matrix; for k Reference value of grid-connected current at any given time; For the output vector of the grid-connected converter; To augment the output matrix; The parameter is uncertain.

[0077] As an example, k augmented state variables at time step The expression is: in, For system state variables, for k The state vector of the state-space model of the resonant controller at any given time.

[0078] As an example, the perturbation input matrix The expression is: in, It is a 4x1 grid voltage input matrix. It is a zero matrix with 10 rows and 1 column.

[0079] As an example, refer to the input matrix The expression is: in, This is the reference input matrix for the 10x1 resonant controller state-space model.

[0080] As an example, the augmented output matrix The expression is: in, for The mapping matrix.

[0081] As an example, the state matrix of a closed-loop space model The expression is: in, To augment the open-loop matrix, To control the input matrix, This is the state feedback gain matrix.

[0082] As an example, an augmented open-loop matrix The expression is: in, It is a 4x4 open-loop state transition matrix. This is the reference input matrix for the state-space model of the resonant controller. For the mapping matrix, This is the internal transition matrix of the 10x10 resonant controller state-space model.

[0083] As an example, control the input matrix The expression is: in, It is a 4x1 grid-connected converter voltage input matrix.

[0084] Furthermore, in step S14, scalar and weight matrices are set, which will affect the solution of the optimization problem and the performance of the final control strategy.

[0085] As an example, the scalar may include a first scalar. μ Second scalar σ Third scalar δ Among them, the first scalar μ It is a limit on the cost function, the second scalar. σ With the third scalar δ It is a limit on the state feedback gain matrix.

[0086] As an example, the weight matrix may include a first weight matrix. Q Second weight matrix R , where the first weight matrix Q Used to represent system state variables x m (k) The relative importance of the second weight matrix R Used to represent control laws u ( k The relative importance of ).

[0087] Furthermore, the system output can be set based on the weight matrix. z(k) The expression is: in, , , Let k be the augmented state variable at time k. For control laws, n This represents the number of resonant controllers in the state-space model of the resonant controller. Q This is the first weight matrix. R This is the second weight matrix.

[0088] Furthermore, based on the weight matrix and system output z(k) Constructing the DLQR cost function J DLQR cost function is used to evaluate system performance. J The expression is: in, Represents the mathematical expectation. This represents the number of discrete-time steps used to calculate the cost function. Indicates system output, express k augmented state variables at time t, This represents the control law, where Q is the first weight matrix. R This is the second weight matrix.

[0089] As an example, the DLQR cost function J It can be the H2 norm.

[0090] As an example, you can set the DLQR cost function. J Less than the first scalar μ This gives the grid-connected converter good harmonic suppression capabilities.

[0091] Furthermore, in step S15, for the given scalar and weight matrix, if a symmetric positive definite matrix exists... S i ∈ R 14×14 sum matrix G , L , F Then, minimize the DLQR cost function. J To establish an optimization problem ,in, R 14×14 Represents the real number field. i =1,2.

[0092] Furthermore, at all vertices of the convex polyhedron formed by the basic parameters of the grid-connected converter, the following linear matrix inequality (LIM) constraint must be satisfied: in, F This is a matrix related to system state transitions and transformations, used to describe the transition relationships between system state variables; G A matrix relating control inputs to system state. L To characterize the symmetric positive definite class matrix of system performance / stability constraints, S i For the convex polyhedron i A symmetric positive definite matrix with vertices S j For the convex polyhedron j A symmetric positive definite matrix with vertices It is the identity matrix. For the convex polyhedron iAn augmented open-loop matrix with vertices, For the convex polyhedron i The control input matrix of each vertex For the convex polyhedron i The perturbation input matrix of each vertex. μ As the first scalar, σ For the second scalar, δ It is the third scalar. , , n This represents the number of resonant controllers in the state-space model of the resonant controller. Q This is the first weight matrix. R This is the second weight matrix. The parameter is uncertain.

[0093] As an example, symmetric placeholders in linear matrix inequality constraints It can be set as the transpose of the top-right block matrix, i.e. ,in, A This represents the top-left corner block matrix. B This represents the top-right block matrix. C This represents the block matrix in the lower right corner.

[0094] Furthermore, simulation tools are used to solve the optimization problem to obtain preliminary control-related results. It is then determined whether the optimization problem has a solution; if not, the first scalar is... μ Increase the value of , re-optimize and solve, and observe if there is a solution. If there is still no solution, then increase the value of the second scalar. σ and the third scalar δ The value of is increased, and the optimization solution is re-run. If a solution exists, it is further determined whether the obtained result satisfies the linear matrix inequality constraint. If not, the first scalar is also increased. μ Increase the value of , re-optimize and solve, and observe if there is a solution. If there is still no solution, then increase the value of the second scalar. σ and the third scalar δ Increase the value of and re-optimize the solution; if it satisfies the condition, it means that the optimization problem and the linear matrix inequality constraint have a feasible solution at all vertices of the established convex polyhedron, and then the state feedback gain matrix can be obtained. K The expression is: in, A matrix relating control inputs to system state. It is the inverse matrix of the system state transition and transformation related matrices.

[0095] As an example, the state feedback gain matrix K Satisfy the robust stability of the arbitrary-shot control delay closed-loop digital system.

[0096] As an example, the state feedback gain matrix K Including the first state feedback gain matrix Second state feedback gain matrix Third-state feedback gain matrix ,in, i Representing the harmonic order, different state feedback gain matrices reflect the allocation of importance of different signals in the control process, enabling different control signals to participate in the synthesis of the final control law according to reasonable weights.

[0097] In one example, the optimization problem can be solved in the Matlab environment using the SeDuMi solver and the Yalmip toolbox.

[0098] Further, in step S16, the system state variables are... First state feedback gain matrix Weighting is performed to obtain the state variable control signal. .

[0099] In step S2, please refer to Figure 2 In step S2, the grid-connected current is compared with the grid-connected current reference value to obtain an error signal.

[0100] As an example, the grid connection current reference value should be set according to actual needs. i gref , will input current i g Reference quantity of grid-connected current i gref The comparison is performed to obtain the error signal.

[0101] In step S3, please refer to Figure 2 In step S3, a polyhedral uncertain state space model is constructed. The error signal is sent into the polyhedral uncertain state space model, and the fundamental wave state vector and harmonic state vector are obtained after processing by the polyhedral uncertain state space model.

[0102] For example, please refer to Figure 4 Step S3 may include the following steps: Step S31 to Step S33.

[0103] Step S31: Extend the state space model of the resonant controller to arbitrary pulse digital control delay to obtain a polyhedral uncertain state space model, which includes the fundamental branch and harmonic branch.

[0104] Step S32: Input the error signal into the polyhedral uncertain state space model and obtain the fundamental state vector through the fundamental branch.

[0105] Step S33: Obtain the harmonic state vector by passing the error signal through the harmonic branch in the polyhedral uncertain state space model.

[0106] Specifically, in step S31, the resonant controller state-space model The expression is: in, This is the internal transition matrix of the resonant controller's state-space model. for k The state vector after resonant control at time 1 This is the reference input matrix for the state-space model of the resonant controller. for k Reference value of grid-connected current at all times for k The current flowing into the grid at all times.

[0107] As an example, k The state vector after resonant control at any given time The expression is: in, , For the state components of the fundamental wave, , For harmonic state components, This represents the harmonic order.

[0108] As an example, the internal transition matrix of the state-space model of the resonant controller. The expression is: in, , , Angular frequency, The sampling period.

[0109] As an example, the reference input matrix of the resonant controller state-space model. The expression is: in, , , Angular frequency, The sampling period.

[0110] In one example, the harmonic order can be set. i = -5, 7, -11, 13.

[0111] As an example, k The state vector after resonant control at any given time It can include the fundamental state vector Harmonic state vector .

[0112] Furthermore, the state-space model of the resonant controller is... Extending this to arbitrary-time digital control, we obtain a polyhedral uncertain state-space model, expressed as: in, for k The system state variables at time +1, for k System state variables at any given time, Let be the open-loop state transition matrix. For grid-connected converter voltage input matrix, For the grid voltage input matrix, This represents the external disturbance vector. For the mapping matrix, This is the output vector of the polyhedral uncertain state-space model. The parameter is uncertain.

[0113] For example, please refer to Figure 2 The polyhedral uncertain state-space model can include the fundamental wave branch and harmonic branches. The fundamental wave branch includes the fundamental wave matrix operation step. Delay operator and fundamental state feedback matrix Harmonic branches include harmonic matrix operation steps. Fundamental state feedback matrix and harmonic state feedback matrix .

[0114] As an example, in step S32, the error signal is fed into a polyhedral uncertain state space model, where the fundamental branch of the polyhedral uncertain state space model performs targeted resonance control on the fundamental component of the error signal. Specifically, the fundamental component of the error signal undergoes fundamental matrix operation. Then, after the delay operator and fundamental state feedback matrix Feedback processing yields the fundamental state vector. .

[0115] As an example, in step S33, the harmonic branches in the polyhedral uncertain state-space model perform targeted resonance control on the harmonic components in the error signal. Specifically, the harmonic components in the error signal undergo harmonic matrix operation. Then, it passes through the fundamental state feedback matrix. and harmonic state feedback matrix Feedback processing yields the harmonic state vector. .

[0116] In step S4, please refer to Figure 2 In step S4, the fundamental state vector is weighted by the second state feedback gain matrix to obtain the fundamental control signal.

[0117] Specifically, please refer to Figure 2 Fundamental state vector After the second state feedback gain matrix Weighting is performed to obtain the fundamental control signal. .

[0118] In step S5, please refer to Figure 2 In step S5, the harmonic state vector is weighted by the third state feedback gain matrix to obtain the harmonic control signal.

[0119] Specifically, please refer to Figure 2 Harmonic state vector After the third state feedback gain matrix Weighting is performed to obtain the harmonic control signal. .

[0120] In step S6, please refer to Figure 2 In step S6, the state variable control signal, fundamental wave control signal, and harmonic control signal are synthesized to obtain the control law. The control law is then optimized through a feedback loop and transmitted to the drive module. The drive module generates a pulse width modulation signal to drive the switching devices in the grid-connected converter.

[0121] Specifically, please refer to Figure 2 , control the state variable signal Fundamental wave control signal Harmonic control signals The control law is synthesized and then optimized through a feedback loop.

[0122] As an example, a feedback loop may include a delay operator. and feedback matrix The control law passes through the delay operator. Generate intermediate state variables Intermediate variables are processed through the feedback matrix. The feedback control signal is obtained through processing. Feedback control signal With state variable control signals Fundamental wave control signal Harmonic control signals Together, they form a complete closed-loop control circuit, and the control law is optimized to obtain the optimized control law. The expression is: in, , The first state feedback gain matrix, For feedback matrix; , This is the second-state feedback gain matrix. This is the third-state feedback gain matrix; For system state variables; This is the state vector after resonant control; This is a 1-row, 14-column state feedback gain matrix; for k The state variable is augmented at each moment.

[0123] In one example, the 5th, 7th, 11th, and 13th fundamental frequencies can be selected. ,in, This is the second-state feedback gain matrix. The third-state feedback gain matrix is ​​the one for the 5th fundamental frequency. The third-state feedback gain matrix is ​​the one for the 7th fundamental frequency. The third-state feedback gain matrix is ​​for the 11th fundamental frequency. The third-state feedback gain matrix is ​​the 13th fundamental frequency.

[0124] Furthermore, the optimized control law The signal is transmitted to the drive module, which generates a pulse width modulation signal to drive the switching devices in the grid-connected converter.

[0125] In one embodiment, the robust stability, harmonic suppression capability, and dynamic performance of the control method can also be verified using Lyapunov stability criteria and closed-loop pole plots. Specifically, by analyzing the closed-loop pole distribution of all vertices in the polyhedral uncertain state-space model, if all poles are located within the unit circle, it indicates that the control method has robust stability within the set range of grid impedance variations.

[0126] In yet another embodiment, a grid-connected current reference quantity can also be constructed. i gref ( k ) to grid current i g ( k transfer functionG ref-ig ( z ), and plotted a Bode plot to verify the control method's ability to track the current reference.

[0127] As an example, the grid-connected current reference quantity i gref ( k ) to grid current i g ( k transfer function G ref_ig ( z The expression for ) is: in, To augment the output matrix, For delay operators, It is the identity matrix. This is the state matrix of the closed-loop space model. As a reference input matrix, The parameter is uncertain.

[0128] In another embodiment, the grid voltage can also be constructed. v g ( k ) to grid current i g ( k transfer function Based on grid voltage v g ( k ) to grid current i g ( k transfer function G vg_ig ( z Calculate the output impedance of the grid-connected converter. Z o Bode plots were drawn to verify the harmonic suppression capability of the control method.

[0129] As an example, grid voltage v g ( k ) to grid current i g ( k transfer function G vg_ig ( z The expression is: in, To augment the output matrix, For delay operators, It is the identity matrix. The state matrix of the closed-loop space model. The perturbation input matrix is... The parameter is uncertain.

[0130] As an example, the output impedance of a grid-connected converter Z o The expression is: in, For grid voltage v g ( k ) to grid current i g ( k The transfer function of ).

[0131] In one embodiment, the effective value of the grid voltage can be set. v g =110V, DC bus voltage V dc =350V, mains frequency f 0 = 50Hz, switching frequency f sw =12kHz, sampling frequency f s =24kHz, grid-connected converter side inductance L 1=1mH, filter capacitor C f =7uH, grid-side inductor L g =1mH, the maximum grid impedance is 5mH, and the minimum grid impedance is 0mH, as shown in Table 1.

[0132] Table 1 Basic Parameters of Grid-Connected Converters Furthermore, the perturbation range of the uncertain parameters can be obtained based on the basic parameters. A closed-loop spatial model of the uncertain state under arbitrary-time digital control is constructed based on the basic parameters of the grid-connected converter and the perturbation range of the uncertain parameters. The system output is determined based on the weight matrix. A DLQR cost function is constructed based on the weight matrix and the system output. An optimization problem and linear matrix inequality constraints are established with the goal of minimizing the DLQR cost function. The optimization problem is solved using simulation tools. The system is then checked whether the optimization problem has a solution. If there is no solution, the scalar is adjusted and the optimization is repeated. If there is a solution, the system is further checked whether the obtained result satisfies the linear matrix inequality constraints. If not, the scalar is adjusted and the optimization is repeated. If the constraints are satisfied, the state feedback gain matrix is ​​obtained. The state feedback gain matrix is ​​shown in Table 2. Main state feedback gain matrix, This is the second-state feedback gain matrix. The third-state feedback gain matrix is ​​the one for the 5th fundamental frequency. The third-state feedback gain matrix is ​​the one for the 7th fundamental frequency. The third-state feedback gain matrix is ​​for the 11th fundamental frequency. The third-state feedback gain matrix is ​​the 13th fundamental frequency.

[0133] Table 2 State Feedback Gain Matrix As an example, Figure 5 This is the closed-loop pole plot of the control method under uncertain parameters. Figure 5 This diagram presents the distribution of poles in a closed-loop system within the complex plane, with different markers (crosses, arrows, etc.) used to distinguish different pole sets. , The annotations regarding the radius, and 0.2 ω s 0.4 ω s 0.6 ω s 0.8 ω s Frequency-related scales are used to locate the position of poles in the complex plane, and to determine the distribution range and pattern of poles in grid-connected converters under uncertain parameters. Figure 6 for Figure 5 A magnified view of region D allows for a clearer observation of the dense distribution and arrangement of closed-loop poles within a localized area of ​​region D, aiding in a more precise analysis of the subtle impact of the proposed strategy on the distribution of closed-loop poles under uncertain parameters. This can be achieved by determining whether the closed-loop pole distribution at all vertices of the polyhedral uncertain state-space model is located within the specified region. Figure 5 The robust stability of the grid-connected converter is determined within the unit circle shown. Figure 6 It can be seen that with the grid-side inductance L g As the voltage increases, the distribution of closed-loop poles changes continuously. The inequality in the linear matrix inequality constraint proposed in this application, based on Lyapunov stability, ensures that the closed-loop poles of all vertices of the polyhedral uncertain state-space model are inside the unit circle. This indicates that the grid-connected converter has good robust stability within the set range of grid impedance variation.

[0134] As an example, Figure 7 The Bode plot of the transfer function from the grid-connected current reference quantity to the grid-connected current, from... Figure 7 It can be seen that the closed-loop spatial model has good current reference tracking capability with an amplitude of 0dB and a phase angle of 0º at the fundamental frequency of 50Hz. Therefore, the optimal amplitude frequency response is 0dB and the phase frequency response is 0° at the power frequency.

[0135] As an example, Figure 8 Bode plot of the output impedance of the grid-connected converter, from Figure 8 It can be seen that the output impedance of the grid-connected converter Z o The high amplitude response at the 5th, 7th, 11th, and 13th fundamental frequencies indicates a stronger suppression capability for these harmonics. This suggests that the control system has better performance in terms of grid-side inductance. L g It has good suppression capability for background harmonics of the power grid within the defined perturbation range.

[0136] As an example, Figure 9 This is a steady-state waveform diagram of the single-phase grid-connected current. Figure 10 This is a transient waveform diagram of the single-phase grid-connected current. From... Figure 9 , Figure 10 It can be seen that the control method of this application has good current reference tracking capability and dynamic response performance.

[0137] As an example, Figure 11 This is a simulation waveform of the full-load grid-connected current of the grid-connected converter in the public power grid. Figure 12 The image shows the simulated grid-connected current waveform of the grid-connected converter at a short-circuit ratio (SCR) of 5 under full load. Figure 13 The image shows the simulated grid-connected current waveform of the grid-connected converter under full load with a short-circuit ratio of 2.5. (From...) Figure 11 , Figure 12 , Figure 13 It can be seen that the control method of this application is adapted to different power grid conditions, and the grid-side inductor... L g The grid-connected converter can output high-quality grid-connected current regardless of any variations within the design range.

[0138] In another embodiment, the control method proposed in this application is compared with the conventional quasi-proportional resonant plus harmonic compensator (PR+HC) control strategy under three different conditions, with the calibrated digital control delay time set in the quasi-proportional resonant plus harmonic compensator control strategy. T cal Equal to 0.5 sampling periods T s ,Right now T cal =0.5 T s . Figure 14 This is a simulation waveform of the full-load grid-connected current of a grid-connected converter in a public power grid when using a quasi-proportional resonant control strategy with harmonic compensation. Figure 15 The image shows the simulated waveform of the full-load grid-connected current when the grid-connected converter has a short-circuit ratio of 5, using a quasi-proportional resonant control strategy with harmonic compensation. Figure 16 This is a simulation waveform of the full-load grid-connected current of a grid-connected converter using a quasi-proportional resonant control strategy with harmonic compensation, at a short-circuit ratio of 2.5. From... Figure 14 , Figure 15 , Figure 16 It can be seen that under the three conditions, the total harmonic distortion (THD) of the grid-connected converter using the quasi-proportional resonance plus harmonic compensation control strategy is greater than that using the control method proposed in this application. The control method proposed in this application is superior to the traditional quasi-proportional resonance plus harmonic compensation control strategy, and realizes the coordinated optimization of multiple control objectives such as dynamic performance, harmonic suppression capability and robust stability of the grid-connected converter under arbitrary digital control delay.

[0139] In the DLQR-based grid-connected converter control method of this application, state variables are obtained by collecting the inductor current, grid-connected current, and capacitor voltage of the grid-connected converter. The state feedback gain matrix is ​​solved based on the DLQR strategy, and the state variables are weighted to obtain the state variable control signal. At the same time, the error signal between the grid-connected current and the reference quantity is sent to the polyhedral uncertain state space model, decomposed to obtain the fundamental and harmonic state vectors, and weighted by the corresponding gain matrix respectively. This enables independent and precise control of the fundamental and harmonic components, significantly reduces the harmonic distortion rate of the grid-connected current, and improves power quality. By combining linear matrix inequality constraints and Lyapunov stability criteria, the robust stability of the system can also be guaranteed. The control method of this application ensures that the closed-loop poles of all vertices of the system are always located within the unit circle within the designed range of grid impedance parameter variations, thus ensuring the stable operation of the grid-connected converter under uncertain parameter conditions. It achieves error-free tracking of the current reference, and the converter output impedance has a high amplitude response at the fundamental frequency, which can effectively suppress grid background harmonics and reduce the distortion rate of the grid current. It has strong engineering applicability and is easy to implement and debug.

[0140] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the accompanying drawings may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0141] 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.

[0142] Although this application has been disclosed above with reference to embodiments, it is not intended to limit this application. Anyone skilled in the art can make some modifications and refinements without departing from the spirit and scope of this application.

Claims

1. A DLQR-based grid-connected converter control method, characterized in that, The method comprises the following steps: Collecting the inductance current, grid-connected current and capacitor voltage of the grid-connected converter to obtain system state variables, solving a state feedback gain matrix based on a DLQR strategy, weighting the system state variables through the first state feedback gain matrix to obtain a state variable control signal; Comparing the grid-connected current with a grid-connected current reference to obtain an error signal; Constructing a polyhedral uncertain state space model, inputting the error signal into the polyhedral uncertain state space model, and obtaining a fundamental state vector and a harmonic state vector after processing by the polyhedral uncertain state space model; Weighting the fundamental state vector through a second state feedback gain matrix to obtain a fundamental control signal; Weighting the harmonic state vector through a third state feedback gain matrix to obtain a harmonic control signal; Synthesizing the state variable control signal, the fundamental control signal and the harmonic control signal to obtain a control law, optimizing the control law through a feedback loop, and transmitting the optimized control law to a driving module to generate a pulse width modulation signal to drive the switching devices in the grid-connected converter.

2. The DLQR-based grid-connected converter control method of claim 1, wherein, Collecting the inductance current, grid-connected current and capacitor voltage of the grid-connected converter to obtain system state variables, solving a state feedback gain matrix based on a DLQR strategy, weighting the system state variables through the first state feedback gain matrix to obtain a state variable control signal, comprising: Collecting the inductance current, grid-connected current and capacitor voltage of the grid-connected converter to obtain system state variables; Setting the basic parameters of the grid-connected converter, obtaining the perturbation range of uncertain parameters according to the basic parameters; Based on the basic parameters and the perturbation range of the uncertain parameters, constructing a closed-loop space model of uncertain states under arbitrary digital control delay; Setting a scalar and a weight matrix, determining the system output based on the weight matrix, and constructing a DLQR cost function according to the weight matrix and the system output; Establishing an optimization problem and a linear matrix inequality constraint with the goal of minimizing the DLQR cost function, solving the optimization problem using a simulation tool, determining whether the optimization problem has a solution, adjusting the scalar to solve the optimization problem again if there is no solution, further determining whether the obtained result meets the linear matrix inequality constraint if there is a solution, adjusting the scalar to solve the optimization problem again if it does not meet the linear matrix inequality constraint, and obtaining a state feedback gain matrix if it meets the linear matrix inequality constraint, wherein the state feedback gain matrix comprises a first state feedback gain matrix, a second state feedback gain matrix and a third state feedback gain matrix; Weighting the system state variables through the first state feedback gain matrix to obtain a state variable control signal.

3. The DLQR-based grid-connected converter control method of claim 2, wherein, The DLQR cost function J , is expressed as: wherein, denotes the mathematical expectation, denotes the number of discrete time steps used to compute the cost function, denotes the system output, denotes k the augmented state variable at time instant t, denotes the control law, Q is a first weight matrix, R is a second weight matrix.

4. The DLQR-based grid-connected converter control method of claim 2, wherein, The linear matrix inequality constraint is: wherein F is a matrix associated with system state transition, G is a matrix associated with control input and system state, L is a symmetric positive definite matrix associated with system performance or stability constraints, S i is a symmetric positive definite matrix associated with the i th vertex of a convex polytope, S j is a symmetric positive definite matrix associated with the j th vertex of a convex polytope, is an identity matrix, is an augmented open-loop matrix associated with the i th vertex of a convex polytope, is a control input matrix associated with the i th vertex of a convex polytope, is a disturbance input matrix associated with the i th vertex of a convex polytope, μ is a first scalar, σ is a second scalar, δ is a third scalar, , , n is a number of resonant controllers in a resonant controller state space model, Q is a first weight matrix, R is a second weight matrix, is an uncertain parameter.

5. The DLQR-based grid-connected converter control method of claim 1, wherein, Constructing a polyhedral uncertain state space model, inputting the error signal into the polyhedral uncertain state space model, and obtaining a fundamental state vector and a harmonic state vector after processing by the polyhedral uncertain state space model, comprising: The resonance controller state space model is extended to an arbitrary order digital control delay to obtain a polyhedral uncertain state space model, which includes a fundamental wave branch and a harmonic wave branch; The error signal is sent into the polyhedral uncertain state space model, and a fundamental wave state vector is obtained through the fundamental wave branch; The error signal is sent into the polyhedral uncertain state space model, and a harmonic wave state vector is obtained through the harmonic wave branch.

6. The DLQR-based grid-connected converter control method of claim 5, wherein, The polyhedral uncertain state space model is expressed as: where, is k the system state variable at time instant is k the system state variable at time instant is the open-loop state transition matrix, is the grid-connected converter voltage input matrix, is the grid voltage input matrix, is the external disturbance vector, is the mapping matrix, is the output vector of the polytopic uncertain state-space model, is the uncertain parameter.

7. The DLQR-based grid-connected converter control method of claim 1, wherein, Optimized control law The expression is: wherein, , is a first state feedback gain matrix, is a feedback matrix; , is a second state feedback gain matrix, is a third state feedback gain matrix; is a system state variable; is a state vector after resonance control; is a state feedback gain matrix of 1 row and 14 columns; is k is an augmented state variable at the time point.

8. The DLQR-based grid-connected converter control method of claim 1, wherein, Also included is constructing a transfer function of the grid current reference quantity to the grid current, and plotting a Bode plot of the grid current reference quantity i gref ( k ) to the grid current i g k ) to the grid current G ref_ig z ) is given by:​​ wherein is an augmented output matrix, is a delay operator, is an identity matrix, is a state matrix of the closed loop space model, is a reference input matrix, is an uncertain parameter.

9. The DLQR-based grid-connected converter control method of claim 8, wherein, Further comprising constructing a transfer function of a grid voltage to a grid current, calculating an output impedance of the grid-connected converter based on the transfer function of the grid voltage to the grid current, and drawing a Bode diagram; the grid voltage v g ( k ) to the grid current i g ( k ) transfer function G vg_ig ( z ) is wherein, is an augmented output matrix, is a delay operator, is an identity matrix, is a closed loop space model state matrix, is a disturbance input matrix, is an uncertain parameter; The grid-connected converter output impedance Z o The expression for: .

10. A DLQR-based grid-connected converter control system, characterized in that, The DLQR-based grid-connected converter control system is connected with the grid-connected converter, and comprises a driving module, a discrete linear quadratic regulator control module, a phase-locked loop module, an inductor current acquisition module, a capacitor voltage acquisition module, a grid current acquisition module, and a point of common coupling voltage acquisition module; a first end of the driving module is connected with a first end of the grid-connected converter; a second end of the driving module is connected with a first end of the discrete linear quadratic regulator control module; a second end of the discrete linear quadratic regulator control module is connected with a first end of the phase-locked loop module; a first end of the inductor current acquisition module is connected with a second end of the grid-connected converter; a second end of the inductor current acquisition module is connected with a third end of the discrete linear quadratic regulator control module; a first end of the capacitor voltage acquisition module is connected with a third end of the grid-connected converter; a second end of the capacitor voltage acquisition module is connected with a fourth end of the discrete linear quadratic regulator control module; a first end of the grid current acquisition module is connected with a fourth end of the grid-connected converter; a second end of the grid current acquisition module is connected with a fifth end of the discrete linear quadratic regulator control module; a first end of the point of common coupling voltage acquisition module is connected with a fifth end of the grid-connected converter; and a second end of the point of common coupling voltage acquisition module is connected with a second end of the phase-locked loop module.

Images