A model predictive current control method for Vienna rectifier

Abstract

The application discloses a model predictive current control method of a Vienna rectifier, and designs a cascade control structure of an outer ring based on a super-spiral sliding mode voltage controller and an inner ring based on a model predictive current controller of a Luenberger disturbance observer; a super-spiral sliding mode controller is designed according to a voltage loop model of the Vienna rectifier to perform voltage loop control of the Vienna rectifier; an error signal of a reference bus voltage square term and an actual bus voltage square term is sent into the super-spiral controller to obtain an active power reference value; the reactive power reference value is given in the super-spiral sliding mode voltage controller; a Luenberger observer is established in a d-q current coordinate system; a disturbance value estimated by the Luenberger observer is compensated into the model predictive current controller; and an optimal voltage vector is obtained after traversal optimization. The application combines the model predictive control with the high-order sliding mode control, and the advantages of the two control methods are integrated; the application can inhibit the interference, has good robustness, and improves the response speed of the whole system.

Description

Technical Field

[0001] This invention belongs to the field of power electronic converter control technology, and relates to a model predictive current control method for Vienna rectifiers, especially a model predictive current control method for Vienna rectifiers based on a super-spiral sliding mode controller and a Luneburg observer. Background Technology

[0002] Currently, sliding mode variable structure control and model predictive control have received widespread attention and research in the field of power electronics. Sliding mode control, with its advantages of simple structure, strong robustness, and fast convergence speed, is widely used. Its application in the voltage loop of Vienna rectifiers ensures the stability of the rectified voltage. Model predictive control, due to its simple principle, fast response, and ease of multi-objective optimization, has also attracted widespread attention. Its application in the current loop of Vienna rectifiers can quickly track the given current output from the voltage loop. Existing schemes commonly use both sliding mode control and model predictive control, but only one is typically employed as the control method. Traditional sliding mode control suffers from problems such as complex sliding surface design, complex reaching law construction, and chattering. Model predictive control also faces robustness issues arising from the complex operating conditions of Vienna rectifiers under wind power and photovoltaic power generation. Summary of the Invention

[0003] The purpose of this invention is to provide a model predictive current control method for Vienna rectifiers. In a cascaded structure of outer loop based on superspiral sliding mode voltage control and inner loop based on Luneburg observer predictive current control, the advantages of predictive model control and high-order sliding mode control are combined to achieve robustness and speed of system response.

[0004] This invention discloses a model predictive current control method for a Vienna rectifier, which designs a cascaded control structure with an outer loop based on a superspiral sliding mode voltage controller and an inner loop based on a Luneburg disturbance observer model predictive current controller, including the following steps:

[0005] Step 1: Design a super spiral sliding mode controller based on the Vienna rectifier voltage loop model to control the voltage loop of the Vienna rectifier. Send the error signal between the reference bus voltage square term and the actual bus voltage square term into the super spiral controller to obtain the active power reference value. The reactive power reference value is given in the super spiral sliding mode voltage controller.

[0006] Step 2: In the dq current coordinate system, establish a Luneburger observer, compensate the disturbance value estimated online by the Luneburger observer into the model predictive current controller, and obtain the optimal voltage vector after traversal optimization.

[0007] The superspiral sliding mode voltage controller was designed using the following steps:

[0008] (1) Establish the Vienna rectifier voltage loop model:

[0009]

[0010] Among them, V dc For DC side voltage, i d and i q Let i be the component of the grid-side current on the dq axis. dc For load current, S d and S q For control signals;

[0011] (2) The voltage loop model is rewritten as follows:

[0012]

[0013] in, p = e d i d +e q i q e d and e q These are the components of the grid-side voltage on the dq axis, where d = 2V. dc i dc ;

[0014] (3) Select the voltage deviation signal as the sliding surface, and then send the error signal between the reference bus voltage square term and the actual bus voltage square term into the super-spiral sliding controller:

[0015] s0=z * -z (3)

[0016] Among them, V dc * For reference voltage,

[0017] (4) Determine the relative order of the sliding surface and the voltage loop model:

[0018]

[0019] According to the definition of relative order, the relative order between the sliding surface and the voltage loop model is 1 at this time;

[0020] (5) Based on the characteristics of the super-spiral sliding mode, when the relative order between the sliding surface and the voltage loop model is 1, the reference value p of the active power output of the super-spiral sliding mode voltage controller can be directly obtained. * The expression is:

[0021] p * =-λ0|s0| 0.5sign(s0)-÷W0sign(s0)dt (4)

[0022] Where λ0 and W0 are the adjustable parameters in the superspiral sliding mode voltage controller.

[0023] The Luneburger observer is designed as follows:

[0024] (1) Establish the mathematical model of the current loop in the dq current coordinate system:

[0025]

[0026] Among them, e d and e q Let L be the grid-side voltage component on the dq axis, L be the system input inductance, R be the equivalent resistance, ω be the grid angular frequency, and i be the voltage component on the dq axis. d and i q Let u be the component of the grid-side current on the dq axis. d and u q f is the component of the voltage vector along the dq axis. d and f q This represents the component of the system disturbance on the dq axis.

[0027] (2) The disturbance in the current equation is expressed as:

[0028]

[0029] Where ΔR and ΔL are the differences between the actual resistance value and the nominal inductance value, respectively;

[0030] (3) The state equation of the current loop mathematical model is expressed as:

[0031]

[0032] Where x = [i d i q f d f q ] T y = [i d i q i d i q ] T u = [e d -e q u d -u q ] T ,

[0033]

[0034] (4) Based on the state equation of the current loop mathematical model, the expression for the Luneburg observer is:

[0035]

[0036] in, For the estimated values ​​of the state variables, By adjusting the values ​​of l1 and l2, the Romberg observer can be stabilized, at which point the observed value is equal to the actual value.

[0037] The model predictive current controller is designed as follows:

[0038] (1) Convert the given power into a given current for predictive voltage control. The conversion formula is as follows:

[0039]

[0040] in, and It is the component of the current reference value on the dq axis; p * It is an active power reference value, q * This is a reference value for reactive power;

[0041] (2) Euler's forward formula is:

[0042]

[0043] Among them, T s Let z(k) be the system control period, and z(k+1) be the system state at time k and time k+1, respectively.

[0044] (3) Establish a discrete prediction model for the Vienna rectifier based on Euler's forward formula:

[0045]

[0046] (4) Establish a cost function that includes current and the potential difference at the neutral point:

[0047]

[0048] Where γ is the weighting coefficient of the cost function, V c1 (k+1), V c2 (k+1) represents the voltage values ​​of the upper and lower capacitors at time k+1;

[0049] (5) Using the cost function as the evaluation criterion, all switching states are traversed and optimized to select the optimal switching state so that the system input current tracks the given current as accurately as possible, thereby obtaining the optimal voltage vector after traversal optimization.

[0050] Since the Vienna rectifier control structure is a cascaded inner and outer loop, the requirements for the controller differ between the inner and outer loops. In the voltage controller design, this invention utilizes the rectifier voltage model to construct a superspiral sliding mode voltage controller structure, maintaining the robustness of traditional sliding mode control while reducing chattering. In the current controller design in the dq coordinate system, a Luneburger observer is integrated into the model predictive control structure, establishing a predictive current control structure based on the Luneburger observer. This not only retains the advantage of fast response speed of model predictive control but also avoids model accuracy issues caused by parameter variations. This cascaded structure of outer loop superspiral sliding mode voltage control and inner loop predictive current control based on the Luneburger observer combines model predictive control with high-order sliding mode control, integrating the advantages of both control methods. Functionally, the outer loop superspiral sliding mode voltage controller can suppress interference, exhibits good robustness, and effectively stabilizes the bus voltage. The inner loop typically requires high controller bandwidth; therefore, the predictive current controller based on the Luneburger observer can quickly track the outer loop setpoint, improving the overall system response speed and achieving both robustness and speed in the system response. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the three-phase Vienna rectifier circuit topology in this invention;

[0052] Figure 2 This is a diagram of the super-helical sliding mode control structure in this invention;

[0053] Figure 3 This is a system control block diagram of the present invention;

[0054] Figure 4 This is a comparison chart of the voltage step response simulation results of the present invention and PI-MPC;

[0055] Figure 5 This is a comparison chart of the load mutation simulation results of the present invention and PI-MPC;

[0056] Figure 6 The results are the steady-state simulation results of the system of this invention. Detailed Implementation

[0057] To better understand the above-mentioned objects, features, and advantages of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other.

[0058] Numerous specific details are set forth in the following description to provide a thorough understanding of the invention. The described embodiments are merely some, not all, of the embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0059] 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 herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0060] The term "comprising" and any variations thereof in the specification, claims, and accompanying drawings of this invention are intended to cover a non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0061] Figure 1 This is a schematic diagram of the three-phase Vienna rectifier circuit topology of the present invention, wherein e a e b e c For three-phase input voltage; i a i b i c L1 represents the inductor current; L1, L2, and L3 are boost inductors, each with an inductance value of L; R1, R2, and R3 are equivalent resistances, each with a resistance value of R; D1 to D6 are six fast recovery diodes; C1 and C2 are DC-side filter capacitors; u C1 u C2 These are the voltages across the two capacitors; R L V is the load resistance; dc This is the DC side voltage; i dc The current is DC side, and S1, S2, and S3 are three sets of reverse series switches.

[0062] like Figure 3 As shown, this invention discloses a model predictive current control method for a Vienna rectifier. The method designs a cascaded control structure with an outer loop based on a superspiral sliding mode voltage controller and an inner loop based on a Luneburg disturbance observer model predictive current controller. The method includes the following steps:

[0063] Step 1: Design a super-spiral sliding mode controller based on the Vienna rectifier voltage loop model to control the Vienna rectifier voltage loop. Input the error signal between the reference bus voltage square term and the actual bus voltage square term into the super-spiral controller to obtain the active power reference value p. * reactive power reference value q * The given information is provided in the superspiral sliding mode voltage controller, such as... Figure 2 As shown, the superspiral sliding mode voltage controller is designed using the following steps:

[0064] (1) Establish the Vienna rectifier voltage loop model:

[0065]

[0066] Among them, V dc For DC side voltage, i d and i q Let i be the component of the grid-side current on the dq axis. dc For load current, S d and S q For control signals;

[0067] (2) The voltage loop model is rewritten as follows:

[0068]

[0069] in, p = e d i d +e q i q e d and e q These are the components of the grid-side voltage on the dq axis, where d = 2V. dc i dc ;

[0070] (3) Select the voltage deviation signal as the sliding surface, and then send the error signal between the reference bus voltage square term and the actual bus voltage square term into the super-spiral sliding controller:

[0071] s0=z * -z (3)

[0072] Among them, V dc * For reference voltage,

[0073] (4) Determine the relative order of the sliding surface and the voltage loop model:

[0074]

[0075] According to the definition of relative order, the relative order between the sliding surface and the voltage loop model is 1 at this time;

[0076] (5) Based on the characteristics of the super-spiral sliding mode, when the relative order between the sliding surface and the voltage loop model is 1, the reference value p of the active power output of the super-spiral sliding mode voltage controller can be directly obtained. * The expression is:

[0077] p*=-λ0|s0| 0.5 sign(s0)-∫W0sign(s0)dt (4)

[0078] Wherein, λ0 and W0 are the adjustable parameters in the super-spiral sliding mode voltage controller;

[0079] Step 2: In the dq current coordinate system, establish a Luneburg observer, compensate the disturbance value estimated online by the Luneburg observer into the model predictive current controller, and obtain the optimal voltage vector after traversal optimization.

[0080] The Luneburger observer is designed as follows:

[0081] (1) Establish the mathematical model of the current loop in the dq current coordinate system:

[0082]

[0083] Where L is the system input inductance, R is the equivalent resistance, ω is the grid angular frequency, and i d and i q Let u be the component of the grid-side current on the dq axis. d and u q f is the component of the voltage vector along the dq axis. d and f q This represents the component of the system disturbance on the dq axis.

[0084] (2) The disturbance in the current equation is expressed as:

[0085]

[0086] Where ΔR and ΔL are the differences between the actual resistance value and the nominal inductance value, respectively;

[0087] (3) The state equation of the current loop mathematical model is expressed as:

[0088]

[0089] Where x = [i d i q f d f q ] T y = [id i q i d i q ] T u = [e d -e q u d -u q ] T ,

[0090]

[0091] (4) Based on the state equation of the current loop mathematical model, the expression for the Luneburg observer is:

[0092]

[0093] in, For the estimated values ​​of the state variables, By adjusting the values ​​of l1 and l2, the Romberg observer can be stabilized, at which point the observed value is equal to the actual value.

[0094] The model predictive current controller is designed as follows:

[0095] (1) Convert the given power into a given current for predictive voltage control. The conversion formula is as follows:

[0096]

[0097] in, and It is the component of the current reference value on the dq axis; p * It is an active power reference value, q * This is a reference value for reactive power;

[0098] (2) Euler's forward formula is:

[0099]

[0100] Among them, T s Let z(k) be the system control period, and z(k+1) be the system state at time k and time k+1, respectively.

[0101] (3) Establish a discrete prediction model for the Vienna rectifier based on Euler's forward formula:

[0102]

[0103] (4) Establish a cost function that includes current and the potential difference at the neutral point:

[0104]

[0105] Where γ is the weighting coefficient of the cost function, V c1 (k+1), V c2 (k+1) represents the voltage values ​​of the upper and lower capacitors at time k+1;

[0106] (5) Using the cost function as the evaluation criterion, all switching states are traversed and optimized to select the optimal switching state so that the system input current tracks the given current as accurately as possible, thereby obtaining the optimal voltage vector after traversal optimization.

[0107] Finally, to verify the performance of the STSM-MPC of this invention under the condition of a step change in reference voltage, a simulation platform was used to design a given voltage from 250V to 300V and then to 280V. Figure 4 Simulation results are shown. Compared to the PI-MPC method (65ms, 9.28V), the STSM-MPC method exhibits a faster recovery time (28ms) and smaller voltage overshoot (0.12V). Furthermore, the STSM-MPC method demonstrates better control performance under varying reference voltages.

[0108] To verify the performance of the STSM-MPC of this invention under a step load change, a simulation comparison experiment was designed, in which the load resistance changes from 100Ω to 150Ω in 0.3s and then returns to 100Ω in 0.45s. The simulation results are as follows. Figure 5 As shown, compared to the PI-MPC method (69ms, 4.92V), STSM-MPC has a faster recovery time (45ms) and less voltage overshoot (2.56V). Therefore, STSM-MPC exhibits better control performance.

[0109] Figure 6 The figure shows the simulation results of the invention in steady state. The voltage and current are in phase, which shows that the invention can achieve unity power factor.

[0110] The Vienna rectifier control structure is a cascaded inner and outer loop. In the voltage controller design, this invention utilizes the rectifier voltage model to construct a superspiral sliding mode voltage controller structure, maintaining the robustness of traditional sliding mode control while reducing chattering. In the current controller design in the dq coordinate system, a Luneburger observer is integrated into the model predictive control structure, establishing a predictive current control structure based on the Luneburger observer. This not only retains the fast response speed of model predictive control but also avoids model accuracy issues caused by parameter variations. This cascaded structure of outer loop superspiral sliding mode voltage control and inner loop predictive current control based on the Luneburger observer combines model predictive control with high-order sliding mode control, integrating the advantages of both control methods. Functionally, the outer loop superspiral sliding mode voltage controller can suppress interference, exhibits good robustness, and effectively stabilizes the bus voltage. The inner loop typically requires high controller bandwidth; therefore, the predictive current controller based on the Luneburger observer can quickly track the outer loop setpoint, improving the overall system response speed and achieving both robustness and speed in the system response.

[0111] Although embodiments and drawings of the present invention have been disclosed for illustrative purposes, those skilled in the art will understand that various substitutions, variations and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to the contents disclosed in the embodiments and drawings.

Claims

1. A model predictive current control method for a Vienna rectifier, characterized in that: The design of a cascaded control structure, with an outer loop based on a superspiral sliding mode voltage controller and an inner loop based on a model predictive current controller using a Luneburg disturbance observer, includes the following steps: Step 1: Design a super spiral sliding mode controller based on the Vienna rectifier voltage loop model to control the voltage loop of the Vienna rectifier. Send the error signal between the reference bus voltage square term and the actual bus voltage square term into the super spiral controller to obtain the active power reference value. The reactive power reference value is given in the super spiral sliding mode voltage controller. Step 2: In the dq current coordinate system, establish a Luneburger observer, compensate the disturbance value estimated online by the Luneburger observer into the model predictive current controller, and obtain the optimal voltage vector after traversal optimization.

2. The model predictive current control method for a Vienna rectifier according to claim 1, characterized in that: The superspiral sliding mode voltage controller was designed using the following steps: (1) Establish the Vienna rectifier voltage loop model: (1) in, DC side voltage and Let be the component of the grid-side current on the dq axis. For load current, and For control signals; (2) The voltage loop model is rewritten as follows: (2) in, , , and These represent the components of the grid-side voltage on the dq axis. ; (3) Select the voltage deviation signal as the sliding surface, so that the error signal between the reference bus voltage square term and the actual bus voltage square term is sent to the super-spiral sliding controller: (3) in, For reference voltage, ; (4) Determine the relative order of the sliding surface and the voltage loop model: (4) According to the definition of relative order, the relative order between the sliding surface and the voltage loop model is 1 at this time; (5) Based on the characteristics of the super-spiral sliding mode, when the relative order between the sliding surface and the voltage loop model is 1, the reference value of the active power output of the super-spiral sliding mode voltage controller can be directly obtained. The expression is: (4) in, , These are the adjustable parameters in the superspiral sliding mode voltage controller.

3. The model predictive current control method for a Vienna rectifier according to claim 1, characterized in that: The Luneburger observer is designed as follows: (1) Establish the mathematical model of the current loop in the dq current coordinate system: (5) in, and These represent the components of the grid-side voltage on the dq axis. R is the system input inductance, and R is the equivalent resistance. The angular frequency of the power grid. and Let be the component of the grid-side current on the dq axis. and The voltage vector components along the dq axis. and This represents the component of the system disturbance on the dq axis. (2) The disturbance in the current equation is expressed as: (6) in, , These are the differences between the actual resistance and inductance values ​​and their nominal values, respectively. (3) The state equation of the mathematical model of the current loop is expressed as: (7) in , , , (4) Based on the state equation of the current loop mathematical model, the expression for the Luneburg observer is: (8) in, , For the estimated values ​​of the state variables, By adjusting , The value of can bring the Romberg observer to a stable state, at which point the observed value is equal to the actual value; The model predictive current controller is designed as follows: (1) Convert the given power into a given current for predictive voltage control. The conversion formula is as follows: (9) in, and It is the component of the current reference value on the dq axis; It is the active power reference value. This is a reference value for reactive power; (2) Euler's forward formula is: (10) in, Let z(k) be the system control period, and z(k+1) be the system state at time k and time k+1, respectively. (3) Establish a discrete prediction model for the Vienna rectifier based on Euler's forward formula: （11） (4) Establish a cost function that includes current and potential difference at the neutral point: (12) in, These are the weighting coefficients of the cost function. , These are the values ​​of the upper and lower capacitor voltages at time k+1; (5) Using the cost function as the evaluation criterion, all switching states are traversed and optimized, and the optimal switching state is selected so that the system input current tracks the given current, thereby obtaining the optimal voltage vector after traversal optimization.

Images