A double closed loop discrete vector control method for PWM rectifier

Abstract

The application discloses a double closed loop discrete vector control method for a PWM rectifier, which comprises the following steps: obtaining an I 2 / U 2 model covering the relationship between the voltage and current of the AC and DC sides based on the mathematical model of the PWM rectifier; obtaining the given value of the active power in the kth period and the given value of the current vector of the current inner loop by means of equivalent transformation; obtaining the predicted value of the current vector in the k+2th period by means of Laplace inverse transformation; obtaining the control voltage vector based on the given value of the current vector of the current inner loop and the predicted value of the current vector, converting the control voltage vector modulation into the PWM control signal of the inverter device, and supplying the inverter device at the k+1th moment to follow the given current vector to implement current control. Thus, the lag caused by the integrator is eliminated, the robustness and dynamic performance of the system are enhanced, the inner loop current vector prediction control is effectively optimized, the grid-side current impact is reduced, and the grid-side current distortion caused by the grid-side voltage imbalance is improved.

Description

Technical Field

[0001] This invention relates to the field of vector control technology for PWM rectifiers, and more particularly to a dual closed-loop discrete vector control method for PWM rectifiers. Background Technology

[0002] Three-phase PWM rectifiers are widely used in industrial applications such as distributed energy, uninterruptible power supplies (UPS), active power filters (APF), regenerative motor drives, and microgrids. Compared to traditional diode rectifiers, three-phase PWM rectifiers can achieve sinusoidal AC current and, due to their high power factor, adjustable DC voltage, and bidirectional energy flow, can significantly reduce harmonic pollution on the AC side.

[0003] In existing technologies, some vehicle-mounted generators or wind turbines produce AC ripple that is too large to be directly supplied to electrical appliances. Therefore, a stable, efficient, and controllable PWM rectifier is needed to ensure a stable voltage input to the DC microgrid for control. Typical control strategies for three-phase PWM rectifiers typically employ vector control and direct power control.

[0004] Chinese patent application CN201610430475.3 discloses a vector control method for a three-phase voltage-source PWM rectifier based on fractional-order PI. This control method employs dual-closed-loop vector control. First, it applies an inner current loop control based on integer-order PI to the three-phase voltage-source rectifier. Then, it applies an outer voltage loop control based on fractional-order PI. Next, it uses SVPWM generation to form a dual-closed-loop vector control model for the three-phase voltage-source PWM rectifier based on fractional-order PI. Finally, it performs parameter tuning on the above vector control model. Chinese patent application CN201720217635.6 discloses a novel space vector control PWM rectifier system, in which the outer voltage loop uses fuzzy sliding film variable structure control. However, in both of these prior art technologies, the inner current loop uses PI control. The inner loop PI controller needs to perform Clark and dq transformations on the current vector before performing an inverse dq transformation for vector control. Therefore, the computational delay in practical applications will affect the control speed.

[0005] Chinese patent application CN201510864288.1 discloses a control method for a three-phase PWM rectifier based on a novel model predictive control. In this method, the inner current loop also uses current predictive control, while the outer loop employs a PI controller. However, this method suffers from a slow bus voltage response and a large current vector tracking error. Furthermore, the traditional PI controller's outer voltage loop predicts the current vector in the second cycle slowly, failing to fully utilize the speed of the inner current loop's predictive control. Summary of the Invention

[0006] To address some or all of the technical problems existing in the prior art, the present invention provides a dual closed-loop discrete vector control method for PWM rectifiers.

[0007] The technical solution of the present invention is as follows:

[0008] A dual-closed-loop discrete vector control method for a PWM rectifier, the method comprising:

[0009] S11: Based on the mathematical model of the PWM rectifier, I is obtained, which covers the voltage and current relationship between the AC and DC sides. 2 / U 2 Model;

[0010] S12: Based on the obtained I 2 / U 2 The model, through equivalent transformation, yields the active power given value for the k-th cycle and the current vector given value for the inner current loop;

[0011] S13: Based on the obtained I 2 / U 2 The model, using the inverse Laplace transform, yields the predicted current vector value for the (k+2)th cycle;

[0012] S14: Based on the current vector setpoint and current vector prediction value of the obtained current inner loop, the control voltage vector is obtained, the control voltage vector is modulated and converted into the PWM control signal of the inverter, and supplied to the inverter at time k+1 to follow the given current vector to implement current control.

[0013] Optionally, in S11, based on the mathematical model of the PWM rectifier, I is obtained, which covers the voltage and current relationship between the AC and DC sides. 2 / U 2 The model includes: the mathematical model of the PWM rectifier includes an AC side model and a DC side model, wherein,

[0014] The AC side model is represented as:

[0015]

[0016] Represented in matrix form as follows:

[0017]

[0018] In the above formula, U A U B and U C Represented as the phase voltages of the three phases a, b, and c on the AC side, U O Represented as common terminal voltage, i a i b and i cLet L represent the phase currents of phases a, b, and c on the AC side, L represent the three-phase input inductance, R represent the equivalent series resistance, and e represent the phase currents of phases a, b, and c on the AC side. a e b e c u represents the three-phase sinusoidal AC voltage with equal amplitude on the grid side. a u b u c It is represented as the controlled voltage between the three phases a, b, and c and the common terminal, where e a e b e c They can be represented as follows:

[0019]

[0020] Among them, E m Indicates voltage amplitude. The initial phase of the three-phase voltage is represented by ω, the angular frequency is represented by t, and time is represented by t.

[0021] Through the Clack transformation of space vectors, the mathematical model of the AC side in the stationary two-phase α-β coordinate system is obtained:

[0022]

[0023] Among them, u α u β Let i be the phase voltage of phases α and β. α i β It is represented as the phase current of the α and β phases.

[0024] The DC-side model is represented as follows:

[0025]

[0026] Among them, U dc Represented as DC bus voltage, C represents DC side charging capacitor, R L Expressed as the equivalent load impedance on the bus side, S a S b and S c Let be the switching coefficients corresponding to the three bridge arms on the DC side. Then, the switching functions corresponding to the three bridge arms are:

[0027]

[0028] Based on the switching function, the relationship between the DC bus voltage and the controlled voltage is obtained as follows:

[0029]

[0030] Therefore, the DC-side model can be rewritten as follows:

[0031]

[0032] in,

[0033]

[0034] Applying the Clack transformation to the above equation and substituting it into the equations of the AC side model, we obtain:

[0035]

[0036] Let the magnitude of the space current vector The mathematical model for the DC side with respect to the square of the bus voltage and the square of the grid current vector is obtained as follows:

[0037]

[0038] Where γ is the current angle, and according to unity power factor control, the current and voltage phases coincide, and there exists

[0039] Optionally, S12: based on the obtained I 2 / U 2 The model, through equivalent transformation, yields the active power setpoint for the k-th cycle and the current vector setpoint for the inner current loop, including:

[0040] First Converting this to the equivalent active power P* on the DC side, we get the following equation:

[0041]

[0042] P* is used as the equivalent control quantity, let Take the sliding mold surface:

[0043]

[0044] Combining the exponential reaching law to reduce the jitter of the error signal reaching the sliding surface, we obtain the following formula:

[0045]

[0046] Where ε and K are both reaching law coefficients. Since the sampling of the actual system is discrete periodic, the sliding surface in the formula is discretized to obtain:

[0047]

[0048] Among them, T s Represented as discrete period, y d (k+1) represents the square of the expected value of the bus voltage. Based on the above formula, the active power setpoint for the k-th cycle is obtained:

[0049]

[0050] For the above equation, use either method one or method two to solve it. in,

[0051] Method 1 includes: using the difference to replace the differential to obtain: in, The current vector value from the previous cycle is given and treated as a constant in the equation. Solving the quadratic equation directly, discarding inconsistencies, yields:

[0052]

[0053] Method two includes: using the fourth-order Runge-Kutta method to solve the problem; to simplify the expression, we take a = R. c = E m The formula for the active power setpoint in the k-th cycle is transformed to obtain:

[0054]

[0055] The fourth-order Runge-Kutta method for solving the above differential equation is given by the following equation:

[0056]

[0057] in:

[0058]

[0059] Optionally, S13: based on the obtained I 2 / U 2 The model, using the inverse Laplace transform, yields the predicted current vector value for the (k+2)th cycle, including:

[0060] Will I 2 / U 2 In the model's formula, the current state equation on the AC side is transformed using the inverse Laplace transform to obtain the discrete state equation:

[0061] i s (t0+t)=F(t)i s (t0)+G(t)u s (t0)+H(t,t0)

[0062] in, The time constant τ = L / R, γ = arctan(ωL / R), and Tr represents the rotation transformation matrix. The change in current during this time period can be expressed as:

[0063]

[0064] Where, Δi si The change vector Δi is represented by the change caused by the zero-input current response. su This is expressed as a change in voltage input response, Δi se Expressed as the change in rotational speed, i.e., the back electromotive force vector response, the above three terms can be converted into amplitude-phase angle expressions as follows:

[0065]

[0066]

[0067]

[0068] Let the time change in the above formula be t0 = T. s By using Taylor series expansion and truncating to quadratic terms, and then discretizing, we can obtain:

[0069]

[0070]

[0071]

[0072] According to the current vector i at time k s (k), by summing, the predicted value of the current vector in the (k+1)th cycle can be calculated as follows:

[0073]

[0074] The predicted current vector values ​​for the (k+1)th cycle obtained from the above formula Substituting this into the prediction equation for the next period, we can obtain the predicted current vector value for the (k+2)th period:

[0075]

[0076] Optionally, in S14, based on the obtained current vector setpoint and current vector prediction value of the inner current loop, a control voltage vector is obtained, the control voltage vector is modulated and converted into a PWM control signal for the inverter, and supplied to the inverter at time k+1 to follow the given current vector to implement current control, including:

[0077] In the formula for obtaining the predicted current vector value for the (k+2)th cycle, Based on the previous solution of the outer voltage loop, the hidden value in Δi is... su u in (k+2k+1) s (k+1) represents the final control voltage vector that needs to be obtained, u. s(k+1) acts on the PWM rectifier in the interval [k+1, k+2], tracking the given current vector, and can be simplified to:

[0078]

[0079] in,

[0080] The main advantages of the technical solution of this invention are as follows:

[0081] The method in this invention first derives the mathematical model of the PWM rectifier using Clark transform, revealing a squared relationship between the DC bus voltage and the AC grid-side current vector. This relationship is then used in conjunction with an exponential reaching law to design a sliding mode controller. Finally, the given current vector is obtained by solving differential equations, replacing the method of calculating the given current vector using PI control. This eliminates the lag caused by the integrator, enhances the robustness and dynamic performance of the system, and effectively optimizes the inner-loop current vector prediction control. This method, based on a novel voltage squared outer loop of sliding mode control combined with dual closed-loop control of the inner-loop current vector prediction, can reduce the grid-side current surge caused by power variations and improve grid-side current distortion caused by grid-side voltage imbalance. Attached Figure Description

[0082] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and constitute a part of this invention, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0083] Figure 1 This is a circuit topology diagram of a PWM rectifier according to one embodiment of the present invention;

[0084] Figure 2 This is a diagram illustrating the dual closed-loop control strategy of a PWM rectifier according to one embodiment of the present invention.

[0085] Figure 3a This is a timing diagram for a traditional dual closed-loop control system used in PWM rectifiers.

[0086] Figure 3b The following is a timing diagram for dual closed-loop control of a PWM rectifier according to one embodiment of the present invention;

[0087] Figure 4a The voltage and current dynamic response curves of a PWM rectifier under traditional PI control are shown.

[0088] Figure 4b The diagram shows the voltage and current dynamic response curves of the PWM rectifier under the control of the method in this embodiment.

[0089] Figure 5The diagram shows a comparison of the dynamic changes of the bus voltage from startup to stability, where the inner current loop uses the vector predictive controller proposed in this embodiment, and the outer voltage loop uses traditional PI and discrete sliding mode control respectively. Detailed Implementation

[0090] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0091] The technical solutions provided by the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0092] In related technologies, Chinese patent application number CN201410812646.X discloses a current prediction control method for surface-mounted permanent magnet synchronous motors. Based on this, the present invention applies this method to power electronic drive control and combines it with voltage loop control to obtain a more stable output voltage.

[0093] like Figures 1 to 5 As shown, in one embodiment of the present invention, a dual closed-loop discrete vector control method for PWM rectifiers is provided. This method is a novel dual closed-loop control method based on sliding mode control with a voltage squared outer loop and a current inner loop vector prediction. This method takes the current vector as the control target. It can reduce the grid-side current harmonics through current control, thereby reducing the harmonics of the current flowing into the bus. It can also reduce the fluctuation of the bus voltage due to load changes through voltage control, thereby improving the bus voltage response speed.

[0094] The dual closed-loop discrete vector control method for PWM rectifiers in this embodiment includes:

[0095] S11: Based on the mathematical model of the PWM rectifier, I is obtained, which covers the voltage and current relationship between the AC and DC sides. 2 / U 2 Model;

[0096] S12: Based on the obtained I 2 / U 2 The model, through equivalent transformation, yields the active power given value for the k-th cycle and the current vector given value for the inner current loop;

[0097] S13: Based on the obtained I 2 / U 2 The model, using the inverse Laplace transform, yields the predicted current vector value for the (k+2)th cycle;

[0098] S14: Based on the current vector setpoint and current vector prediction value of the obtained current inner loop, the control voltage vector is obtained, the control voltage vector is modulated and converted into the PWM control signal of the inverter, and supplied to the inverter at time k+1 to follow the given current vector to implement current control.

[0099] Specifically, the dual-closed-loop discrete vector control method for PWM rectifiers in this embodiment can be used for Figure 1 The circuit topology of the three-phase PWM rectifier is shown in the figure.

[0100] In step S11, based on the mathematical model of the PWM rectifier, I is obtained, which covers the voltage and current relationship between the AC and DC sides. 2 / U 2 The model includes:

[0101] The mathematical model of the PWM rectifier includes an AC side model and a DC side model, wherein...

[0102] The AC side model is represented as:

[0103]

[0104] Represented in matrix form as follows:

[0105]

[0106] In equations (1) and (2) above, U A U B and U C Represented as the phase voltages of the three phases a, b, and c on the AC side, U O Represented as common terminal voltage, i a i b and i c Let L represent the phase currents of phases a, b, and c on the AC side, L represent the three-phase input inductance, R represent the equivalent series resistance, and e represent the phase currents of phases a, b, and c on the AC side. a e b e c u represents the three-phase sinusoidal AC voltage with equal amplitude on the grid side. a u b u c It is represented as the controlled voltage between the three phases a, b, and c and the common terminal.

[0107] Among them, e a e b e c The three-phase input voltage, equivalent to that of a power grid or a three-phase generator, is usually considered to be a symmetrical voltage with equal amplitude and a phase difference of 120°, and can be expressed as follows:

[0108]

[0109] Among them, E m Indicates voltage amplitude. ω represents the initial phase of the three-phase voltage, ω represents the angular frequency, and t represents time.

[0110] Through the Clack transformation of space vectors, the mathematical model of the AC side in the stationary two-phase α-β coordinate system is obtained:

[0111]

[0112] The DC-side model is represented as follows:

[0113]

[0114] Among them, U dc Represented as DC bus voltage, C represents DC side charging capacitor, R L Expressed as the equivalent load impedance on the bus side, S a S b and S c Let be the switching coefficients corresponding to the three bridge arms on the DC side. Then, the switching functions corresponding to the three bridge arms are:

[0115]

[0116] Based on the switching function, the relationship between the DC bus voltage and the controlled voltage is obtained as follows:

[0117]

[0118] Therefore, the DC-side model above can be rewritten as follows:

[0119]

[0120] in,

[0121]

[0122] Applying the Clack transformation to the above equation and substituting it into the equations of the AC side model, we obtain:

[0123]

[0124] Let the magnitude of the space current vector The mathematical model for the DC side with respect to the square of the bus voltage and the square of the grid current vector is obtained as follows:

[0125]

[0126] Where γ is the current angle, and according to unity power factor control, the current and voltage phases coincide. For example, if the initial position of the voltage vector is -90°, the initial value of γ should also be -90°.

[0127] Therefore, the mathematical model of the three-phase PWM rectifier becomes I, which covers the voltage and current relationship between the AC and DC sides. 2 / U 2 The model facilitates the subsequent design of the controller.

[0128] It is understandable that PWM rectifiers typically employ a control structure of an outer voltage loop and an inner current loop to obtain a robust DC bus voltage and a unity power factor AC current. In this embodiment, a voltage squared outer loop controller based on sliding mode control can be designed first, according to the mathematical model of the three-phase PWM rectifier, and then combined with an inner loop current prediction controller to achieve dual closed-loop control.

[0129] Figure 2 The overall control strategy in this embodiment is shown, wherein, It is the square of the bus voltage. The current vector setpoint provided to the inner current loop can be obtained by solving numerical differential equations after being output by the sliding mode controller.

[0130] From the mathematical model formula of the three-phase PWM rectifier, it can be seen that it contains the current vector |i s The square and linear terms of | make it impossible to directly establish the state equation to design a sliding mode controller. Therefore, it is necessary to design a sliding mode controller based on the differential equation.

[0131] In this embodiment, S12: based on the obtained I 2 / U 2 The model, through equivalent transformation, yields the active power setpoint for the k-th cycle and the current vector setpoint for the inner current loop, including:

[0132] First Converting this to the equivalent active power P* on the DC side, we get the following equation:

[0133]

[0134] P* is used as the equivalent control quantity, let Take the sliding mold surface:

[0135]

[0136] Combining the exponential reaching law to reduce the jitter of the error signal reaching the sliding surface, we obtain the following formula:

[0137]

[0138] Where ε and K are both reaching law coefficients. Since the sampling of the actual system is discrete periodic, the sliding surface in the formula is discretized to obtain:

[0139]

[0140] Among them, T s Represented as discrete period, y d (k+1) represents the square of the expected value of the bus voltage, and therefore does not change with the period. Based on the above formula, the active power setpoint for the k-th period is obtained:

[0141]

[0142] For the above equation, use either method one or method two to solve it. in,

[0143] Method 1 includes: using the difference to replace the differential to obtain: in, The current vector value from the previous cycle is given and treated as a constant in the equation. Solving the quadratic equation directly, discarding inconsistencies, yields:

[0144]

[0145] Method two includes: using the fourth-order Runge-Kutta method (RK4) to solve the problem. To simplify the expression, we take a = R. c = E m The formula (16) for the active power setpoint in the kth cycle is transformed to obtain:

[0146]

[0147] The fourth-order Runge-Kutta method for solving the above differential equation is given by the following equation:

[0148]

[0149] in:

[0150]

[0151] Of course, in addition to methods one and two mentioned above, other methods can be used to solve the problem in other implementations.

[0152] Therefore, the active power setpoint P for the kth cycle can be obtained. * (k), then substitute back into equation (14) to solve for the final control quantity using the differential equation.

[0153] Furthermore, in this embodiment, the current vector setpoint of the inner current loop is obtained, along with the predicted value obtained from current vector predictive control. Discretization allows us to calculate the final current control vector u. s (k+1).

[0154] In order to obtain the two physical quantities mentioned above, in this embodiment, it is first necessary to analyze the time from the initial time t0 to t 0+t The equation for the dynamic change of the current vector.

[0155] In this embodiment, S13: based on the obtained I 2 / U 2 The model, using the inverse Laplace transform, yields the predicted current vector value for the (k+2)th cycle, including:

[0156] Will I 2 / U 2 In the model's formula, the current state equation on the AC side is transformed using the inverse Laplace transform to obtain the discrete state equation:

[0157] i s (t0+t)=F(t)i s (t0)+G(t)u s (t0)+H(t,t0) (21)

[0158] in, The time constant τ = L / R, γ = arctan(ωL / R), and Tr represents the rotation transformation matrix. The change in current during this time period can be expressed as:

[0159]

[0160] Where, Δi si The change vector Δi is represented by the change caused by the zero-input current response. su This is expressed as a change in voltage input response, Δi se This is expressed as the change in the vector response of the rotational speed, i.e., the back electromotive force.

[0161] It can be seen from formula (22) that Δi s It mainly consists of three independent terms Δi si , Δi su and Δi se Composed of these three independent terms, each consisting of the current vector i at time t0. s (t0), control voltage vector u s The grid-side voltage at time (t0) and t0 is determined.

[0162] In this embodiment, the above three items are converted into amplitude-phase angle expressions as follows:

[0163]

[0164]

[0165]

[0166] Let the time change in the above formula be t0 = T. s By using Taylor series expansion and truncating to quadratic terms, and then discretizing, we can obtain:

[0167]

[0168]

[0169]

[0170] According to the current vector i at time k s (k), by summing, the predicted value of the current vector in the (k+1)th cycle can be calculated as follows:

[0171]

[0172] The predicted current vector values ​​for the (k+1)th cycle obtained from the above formula Substituting this into the prediction equation for the next period, we can obtain the predicted current vector value for the (k+2)th period:

[0173]

[0174] Further, in the method of this embodiment, in S14, based on the obtained current vector setpoint and current vector prediction value of the inner current loop, a control voltage vector is obtained, the control voltage vector is modulated and converted into a PWM control signal for the inverter, and supplied to the inverter at time k+1 to follow the given current vector to implement current control, including:

[0175] In the formula for obtaining the predicted current vector value for the (k+2)th cycle, Based on the previous solution of the outer voltage loop, the hidden value in Δi is... su u in (k+2|k+1) s (k+1) represents the final control voltage vector that needs to be obtained, u. s (k+1) acts on the PWM rectifier in the interval [k+1, k+2], tracking the given current vector, and can be simplified to:

[0176]

[0177] in,

[0178] Therefore, in the method of this embodiment, for the PWM rectifier, sliding mode control combined with predictive control can respectively form voltage and current control of the PWM rectifier. Moreover, in the voltage outer loop sliding mode control, by using equivalent control, power can be output as an equivalent control quantity, and the corresponding current setpoint can be calculated.

[0179] Meanwhile, the method of this embodiment also adopts current vector prediction control in α-β coordinates and utilizes a mathematical model for predicting current obtained by finite truncation of Taylor series.

[0180] In the attached diagram:

[0181] Figure 3a In the traditional dual-loop control timing diagram shown, the sampled value (discrete value) is compared with the predicted value (given value) at a given time k+1 to obtain the error; then, the controller C, such as a PID controller, outputs the control variable u with the error. s (k+1) is applied to the controlled object G in the interval [k+1, k+2]. In this control method, the actual value is obtained at time k+2 to determine whether to follow the control result of the given value. Therefore, there is at least a one-cycle delay between sampling and control.

[0182] Figure 3b In the dual closed-loop control timing diagram of this embodiment shown, the control variable u s (k+1) At time k, the expected value at time k+2 and the derived predicted value u are mainly used. s (k+2) calculations can eliminate Figure 3a The delay shown in the figure.

[0183] Figure 4a and Figure 4b The voltage and current response curves at startup and during load surges are shown under both conventional PI control and the method described in this embodiment. In the comparative analysis with conventional PI control, the dynamic response at startup of 3kW and when the load suddenly increases from 3kW to 6kW clearly shows that the method in this embodiment can effectively suppress bus voltage fluctuations and reduce the current impact of the grid side on the system. The grid side voltage and current reach the same phase more quickly at startup, effectively reducing reactive power loss and achieving unity power factor control.

[0184] Figure 5The diagram illustrates the dynamic changes of the bus voltage from startup to stability under the conditions of using the vector predictive controller proposed in this embodiment for the inner current loop and the conventional PI and discrete sliding mode control for the outer voltage loop, respectively. By comparison, it is clear that the conventional PI control has a certain delay in predictive control of the inner current loop, while the discrete sliding mode control method in this embodiment can effectively improve the speed and robustness of the system.

[0185] The dual-closed-loop discrete vector control method for PWM rectifiers in this embodiment has the following advantages:

[0186] The method in this embodiment first derives the mathematical model of the PWM rectifier using Clark transform, revealing a square relationship between the DC bus voltage and the AC grid-side current vector. This relationship is then used in conjunction with an exponential reaching law to design a sliding mode controller. Finally, the given current vector is obtained by solving differential equations, replacing the method of calculating the given current vector using PI control. This eliminates the lag caused by the integrator, enhances the robustness and dynamic performance of the system, and effectively optimizes the inner-loop current vector prediction control. This method, based on a novel voltage square outer loop of sliding mode control combined with dual closed-loop control of current inner-loop vector prediction, can reduce the grid-side current surge caused by power variations and improve grid-side current distortion caused by grid-side voltage imbalance.

[0187] It is understandable that, as an implementation method, the Clark converter, current vector prediction controller, and PWM rectifier mentioned above are all implemented in software in STM32F407ZET6. The voltage and current sensors are CHV-25P Hall voltage sensor and CHB-100A current Hall sensor from Beijing Senshe Electronics Co., Ltd. The inverter is a three-phase rectifier bridge device of the inverter section of a 30KW crane power take-off power converter. The improved version is compared with the load under an 80Ω / 5mH resistive-inductive load.

[0188] The aforementioned "Clark transformation" refers to the 3 / 2 transformation, which means the transformation from a three-phase to a two-phase stationary coordinate system, and is a general term; "PWM" means pulse width modulation, and is also a general term.

[0189] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Additionally, the terms "front," "back," "left," "right," "upper," and "lower" in this document refer to the placement shown in the accompanying drawings.

[0190] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A dual-closed-loop discrete vector control method for PWM rectifiers, characterized in that, The method includes: S11: Based on the mathematical model of the PWM rectifier, I is obtained, which covers the voltage and current relationship between the AC and DC sides. 2 / U 2 Model; S12: Based on the obtained I 2 / U 2 The model, through equivalent transformation, yields the active power given value for the k-th cycle and the current vector given value for the inner current loop; S13: Based on the obtained I 2 / U 2 The model, using the inverse Laplace transform, yields the predicted current vector value for the (k+2)th cycle; S14: Based on the obtained current vector setpoint and current vector prediction value of the inner current loop, the control voltage vector is obtained. The control voltage vector is modulated and converted into a PWM control signal for the inverter, and supplied to the inverter at time k+1 to follow the given current vector to implement current control, wherein: In step S11, the mathematical model of the PWM rectifier includes an AC side model and a DC side model, wherein, The AC side model is represented as: ； Represented in matrix form as follows: ； In the above formula, U A , U B and U C Represented as the phase voltages of the three phases a, b, and c on the AC side. U O Represented as common terminal voltage, i a , i b and i c Let L represent the phase currents of phases a, b, and c on the AC side, L represent the three-phase input inductance, and R represent the equivalent series resistance. e a , e b , e c This represents a three-phase sinusoidal AC voltage with equal amplitude on the grid side. u a , u b , u c It is represented as the controlled voltage between the three phases a, b, and c and the common terminal, where, e a , e b , e c They can be represented as follows: ； in, E m Indicates voltage amplitude. Indicates the initial phase of the three-phase voltage. Represents angular frequency. Indicates time, Through the Clack transformation of space vectors, the mathematical model of the AC side in the stationary two-phase α-β coordinate system is obtained: ； The DC-side model is represented as follows: ； in, U dc Represented as DC bus voltage. C This is represented as the DC-side charging capacitor. R L This is expressed as the equivalent load impedance on the bus side. S a , S b and S c Let be the switching coefficients corresponding to the three bridge arms on the DC side. Then, the switching functions corresponding to the three bridge arms are: ； Based on the switching function, the relationship between the DC bus voltage and the controlled voltage is obtained as follows: ； Therefore, the DC-side model can be rewritten as follows: ； in, ； Applying the Clack transformation to the above equation and substituting it into the equations of the AC side model, we obtain: ； Let the magnitude of the space current vector The mathematical model for the DC side with respect to the square of the bus voltage and the square of the grid current vector is obtained as follows: ； Where γ is the current angle, and according to unity power factor control, the current and voltage phases coincide, and there exists ; S12: Based on the obtained I 2 / U 2 The model, through equivalent transformation, yields the active power setpoint for the k-th cycle and the current vector setpoint for the inner current loop, including: First Converted to DC-side equivalent active power P We obtain the following formula: ； P As an equivalent control quantity, let , Take the sliding surface: ； Combining the exponential reaching law to reduce the jitter of the error signal reaching the sliding surface, we obtain the following formula: ； in, , All of these are reaching law coefficients. Since the sampling in the actual system is all discrete periodic, the sliding surface in the formula is discretized to obtain: ； in, T s Represented as discrete periodicity, Expressed as the square of the expected value of the bus voltage, the active power setpoint for the k-th cycle is obtained according to the above formula: ； For the above equation, use either method one or method two to solve it. ,in, Method 1 includes: using the difference to replace the differential to obtain: ,in, The current vector value from the previous cycle is given and treated as a constant in the equation. Solving the quadratic equation directly, discarding inconsistencies, yields: ； Method two includes: using the fourth-order Runge-Kutta method to solve the problem; to simplify the expression, we take... , , The formula for the active power setpoint in the k-th cycle is transformed to obtain: ； The fourth-order Runge-Kutta method for solving the above differential equation is given by the following equation: ； in: ； S13: Based on the obtained I 2 / U 2 The model, using the inverse Laplace transform, yields the predicted current vector value for the (k+2)th cycle, including: Will I 2 / U 2 In the model's formula, the current state equation on the AC side is transformed using the inverse Laplace transform to obtain the discrete state equation: ； in, time constant τ = L / R , γ =arctan( ωL / R ), where Tr represents the rotation transformation matrix. , i s ( t 0 )for t 0 The current vector at time t, u s ( t 0 )for t 0 The control voltage vector at a given time, and the change in current during this time interval, can be expressed as: ； in, Represented as the change vector caused by the zero-input current response. This is represented as a change in voltage input response. Expressed as the change in rotational speed, i.e., the back electromotive force vector response, the above three terms can be converted into amplitude-phase angle expressions as follows: ； ； ； The time change in the above formula is taken as t0= T s By using Taylor series expansion and truncating to quadratic terms, and then discretizing, we can obtain: ； ； ； Based on the current vector at time k By summing, the predicted value of the current vector in the (k+1)th cycle can be calculated as follows: ； The predicted current vector values ​​for the (k+1)th cycle obtained from the above formula Substituting this into the prediction equation for the next period, we can obtain the predicted current vector value for the (k+2)th period: ； In step S14, based on the obtained current vector setpoint and current vector prediction value of the inner current loop, a control voltage vector is obtained. The control voltage vector is modulated and converted into a PWM control signal for the inverter, and supplied to the inverter at time k+1 to follow the given current vector and implement current control, including: In the formula for obtaining the predicted current vector value for the (k+2)th cycle, Based on the previous solution of the outer voltage loop, the hidden... In The final control voltage vector that needs to be obtained is... Applying a PWM rectifier within the interval [k+1, k+2], tracking a given current vector, and simplifying, we obtain: ； in, .

Images