Two-level PWM rectifier fixed switching frequency model prediction direct power control method and device

By constructing a discrete domain mathematical model and adaptive superspiral sliding mode control in a two-level PWM rectifier, combined with the strategy of suppressing reactive power change, the complexity and slow response speed of the control strategy under the unbalanced power grid are solved, and the problem of "slow response" is achieved. More efficient current control and stability improvement.

CN119010637APending Publication Date: 2024-11-22ORDOS HAOHUA CLEAN COAL CO LTD +1
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411063599.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing two-level PWM rectifier control strategy under an unbalanced power grid has problems such as cumbersome extraction of positive and negative sequence components, difficult controller parameters to be tuned, and slow dynamic response speed. The finite set model prediction control has a long rolling optimization time and control power. The problem of large pulsation and irregular switching frequency.

Method used

The discrete domain PWM rectifier mathematical model is constructed under the two-phase stationary αβ coordinate system, defines new reactive power, uses an adaptive superspiral sliding mode controller to establish a DC-side power mathematical model, calculates the optimal voltage vector online through the cost function, and Apply a control strategy to suppress reactive power change rate to optimize switching frequency and voltage vectors.

Benefits of technology

It shortens the calculation time, improves the dynamic response speed, reduces the THD of the grid-side current, enhances the steady-state and dynamic performance of the system, and solves the problems of poor robustness and slow dynamic response speed of the PI controller.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119010637A_ABST
    Figure CN119010637A_ABST
Patent Text Reader

Abstract

The invention discloses a two-level PWM rectifier fixed switching frequency model prediction direct power control method and device, and the method comprises the steps: solving partial derivatives of voltage vectors under an alpha-beta coordinate system through a cost function, and carrying out the online calculation of an expression of an optimal voltage vector. Meanwhile, the influence of the optimal voltage vector on the actual reactive power is analyzed, a control strategy for reducing the change rate of the reactive power is provided, and the harmonic component of the grid-side current is effectively reduced. The self-adaptive super-spiral sliding mode controller is used for replacing a traditional PI controller to obtain an active power given value, compared with the PI controller, the self-adaptive super-spiral sliding mode controller is high in dynamic response speed and robustness, the fluctuation range of the active power given value is small, the steady-state and dynamic characteristics of the rectifier are effectively improved, and the service life of the rectifier is prolonged. And direct power control of low harmonic distortion of the two-level PWM rectifier under the unbalanced power grid is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of power electronics, and in particular relates to a method and a device for predicting direct power control of a two-level PWM rectifier using a fixed switching frequency model. Background Art

[0002] The PWM rectifier is essentially a device that converts AC power into DC power. It has the advantages of bidirectional energy flow, adjustable DC side output voltage and unit power operation. It is widely used in photovoltaic power generation systems, wireless charging devices for new energy vehicles and uninterruptible power supplies. The control strategy under unbalanced power grid is generally based on independent double closed-loop control of positive and negative sequence components, but it has disadvantages such as cumbersome extraction of positive and negative sequence components, difficult setting of controller parameters and slow dynamic response speed. Model predictive direct power control based on extended reactive power theory not only omits the extraction of positive and negative sequence components and phase-locked loop, but also speeds up the dynamic response speed and reduces the number of controllers. When using finite set model predictive control, its control logic is to traverse the cost function with a limited voltage vector so that the minimum voltage vector of the cost function value is the optimal voltage vector. However, finite set model predictive control has problems such as long rolling optimization time, large control power pulsation and non-fixed switching frequency. Summary of the invention

[0003] The object of the present invention is to provide a method and device for direct power control of a two-level PWM rectifier using a fixed switching frequency model prediction, so as to solve the above-mentioned problem.

[0004] The present invention is implemented according to the following technical solutions:

[0005] In a first aspect, the present invention provides a two-level PWM rectifier fixed switching frequency model prediction direct power control method, the method comprising:

[0006] A mathematical model of PWM rectifier in discrete domain is constructed in two-phase stationary αβ coordinate system, and a new reactive power is defined to replace the traditional reactive power.

[0007] According to the constructed discrete domain PWM rectifier mathematical model, the dynamic equations of active power and new reactive power at time k are obtained;

[0008] Discretize the dynamic equations of active power and new reactive power to obtain the predicted values ​​of active power and new reactive power at time k+1;

[0009] A DC power mathematical model based on an adaptive super-helical sliding mode controller is established, and the active power given value at time k is obtained through the mathematical model.

[0010] The cost function is constructed by constraining the sum of squares of the errors between the predicted values ​​of active power and new reactive power and the given values;

[0011] The cost function is used to solve the partial derivatives of the voltage vectors in the two-phase stationary αβ coordinate system, and the expression of the optimal voltage vector is calculated online.

[0012] The control strategy of suppressing reactive power change rate is applied to decompose the optimal voltage vector online expression into the term of reducing reactive power change rate and the term of accurately tracking active power. The weight factor is determined by the operating state of the two-level PWM rectifier to optimize the optimal voltage vector expression.

[0013] Determine the value of the weight factor for reducing reactive power fluctuation and the weight factor for accurately tracking active power, and verify whether the optimal voltage vector based on suppressing the reactive power change rate can meet the preset active power setting and output the optimal voltage vector;

[0014] The optimal voltage vector generates a switching signal and applies it to the two-level PWM rectifier.

[0015] In some embodiments, the active power and the new reactive power at time k are calculated from the grid voltage at time k and its delay, and the grid-side current at time k. k With the new reactive power Q nov,k The calculation expression is:

[0016]

[0017] Among them, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β Delay value of 90° electrical angle, i α 、i β is the current on the αβ axis grid side.

[0018] In some embodiments, the predicted values ​​of active power and new reactive power at time k+1 are:

[0019]

[0020] Where, P and Qnov with superscript k+1 are single-step prediction values, and superscript k is the calculated value at the current moment; T s is the control period, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β The delay value of 90° electrical angle, w is the grid angular frequency, L is the grid side inductance, R is the line impedance, e is the grid voltage, i is the grid side current, v is the rectifier AC side voltage, v α 、v β It is the AC voltage vector of the rectifier under αβ axis.

[0021] In some embodiments, the DC side power mathematical model based on the adaptive super-helical sliding mode controller can be expressed as:

[0022] P ref =u dc i dc +Cv dc (α|σ| 1 / 2 sign(σ)+∫βsign(σ)dt)

[0023] In the formula, u dc is the DC bus voltage measurement value obtained by real-time sampling of the sensor, i dc is the DC side load current measurement value obtained by real-time sampling of the sensor, C is the DC side capacitance, σ is the sliding mode variable, α and β are the adaptive gains, and sign is the sign function.

[0024] In some embodiments, the cost function is:

[0025] g=(P ref -P k+1 ) 2 +(Q nov,ref -Q nov,k+1 ) 2

[0026] Among them, P ref With Q nov,ref are active power and new reactive power given values, P ref The output of the improved super-helical sliding mode controller is obtained, Q nov,ref To ensure unity power factor operation, it is generally set to 0var; P k+1 With Q nov,k+1 are the predicted values ​​of active power and new reactive power.

[0027] In some embodiments, the reactive power change rate suppression control strategy is:

[0028] The influence of the optimal voltage vector on dQ / dt is derived, and the influencing factors are summarized into two categories: one is the high-frequency pulsation part related to the power error, and the other is the steady-state component of the grid voltage and grid-side current; the steady-state component is relatively fixed, and its value does not change much under the operating state of the rectifier, and its coefficient includes T s The dynamic characteristics of reactive power are mainly determined by the high-frequency changes of power error.

[0029] In some embodiments, the optimal voltage vector online expression is decomposed into a term for reducing the reactive power change rate and a term for accurately tracking the active power, and its expression is:

[0030]

[0031] In the formula, and In order to suppress the reactive power fluctuation, and To accurately track the active power term, the corresponding expression is as follows:

[0032]

[0033] Where η is the weight factor for reducing the reactive power change rate, and μ is the weight factor for accurately tracking the active power term.

[0034] In some embodiments, the value selection principle of the weight factor η for reducing the reactive power change rate item and the weight factor μ for accurately tracking the active power item is as follows: η is a fixed value, which is a positive number less than 1; the smaller the value of η, the less the high-frequency harmonic component of the grid-side current; μ is one of 0 and 1, and in order to reduce the harmonic component of the grid-side current, μ is pre-set to 0; if the optimal voltage vector with μ of 0 cannot meet the given active power value, it is modified to 1.

[0035] In some embodiments, the verification of whether the optimal voltage vector based on suppressing the reactive power change rate can meet the active power given value includes:

[0036] The optimal voltage vector v is calculated by using the control strategy of suppressing the reactive power change rate αβ_opt , then μ is equal to 0:

[0037]

[0038] After obtaining the optimal voltage vector, substitute it into the active power expression at time k+1:

[0039]

[0040] Establish an acceptable active power fluctuation range; the specific fluctuation range is modified in real time according to the actual application occasion, and the P ref With P k+1 The difference between the two determines the optimal voltage vector v under the control strategy of suppressing the reactive power change rate. αβ_opt Whether it meets the needs of users, if P k+1 If the voltage is within the acceptable power range, the optimal voltage vector is directly output; if it is outside the range, the optimal voltage vector needs to be recalculated. At this time, μ is equal to 1, and the calculated optimal voltage vector is directly output. t The expression changes to:

[0041]

[0042] In a second aspect, the present invention provides a two-level PWM rectifier fixed switching frequency model prediction direct power control device, the device comprising:

[0043] PWM rectifier mathematical model building module, used to build a discrete domain PWM rectifier mathematical model in a two-phase stationary αβ coordinate system, and define a new reactive power to replace the traditional reactive power;

[0044] A dynamic equation module is used to obtain the dynamic equation of active power and new reactive power at time k according to the constructed discrete domain PWM rectifier mathematical model;

[0045] The prediction value module is used to discretize the dynamic equations of active power and new reactive power to obtain the prediction values ​​of active power and new reactive power at time k+1;

[0046] Mathematical model module, used to establish a DC side power mathematical model based on an adaptive super-helical sliding mode controller, and obtain the active power given value at time k through the mathematical model;

[0047] A cost function construction module is used to construct a cost function by constraining the sum of squares of errors between the predicted values ​​of active power and new reactive power and the given values;

[0048] An online optimal voltage vector calculation module is used to solve partial derivatives of voltage vectors in two-phase stationary αβ coordinate systems using cost functions, and to calculate the expression of the optimal voltage vector online;

[0049] The reactive power change rate suppression control strategy module is used to apply the reactive power change rate suppression control strategy, decompose the optimal voltage vector online expression into a reactive power change rate reduction term and an active power accurate tracking term, and determine the weight factor according to the operating state of the two-level PWM rectifier to optimize the optimal voltage vector expression;

[0050] The module for checking and tracking the active power given value is used to determine the value of the weight factor for reducing the reactive power fluctuation term and accurately tracking the active power term weight factor, and to verify whether the optimal voltage vector based on suppressing the reactive power change rate can meet the preset active power given value and output the optimal voltage vector;

[0051] The action module is used for applying the optimal voltage vector generation switching signal to the two-level PWM rectifier.

[0052] In a third aspect, the present invention provides a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, implement the above-mentioned two-level PWM rectifier model-fixed switching frequency prediction direct power control method.

[0053] Beneficial effects of the present invention:

[0054] Based on the traditional finite set model predictive direct power control, the present invention uses an online method for solving the optimal voltage expression to replace the traditional traversal optimization, which effectively shortens the calculation time and does not need to consider the action time or selection rules of the optimal voltage vector. The adaptive super-spiral sliding mode controller is used instead of the PI controller to set the active power setting value, which solves the problems of slow dynamic response speed and poor robustness of the PI controller. In view of the fact that the control strategy under the unbalanced working condition of the power grid requires the extraction of positive and negative sequence components, the extended reactive power theory is used to simplify the calculation amount, and a reactive power suppression control strategy is proposed, which effectively reduces the THD of the grid-side current. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings are part of the present invention and are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but do not constitute an improper limitation of the present invention. Obviously, the drawings described below are only some embodiments. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0056] Figure 1 It is a two-level PWM rectifier topology diagram;

[0057] Figure 2 A flow chart of a two-level PWM rectifier fixed switching frequency model prediction direct power control method provided by an embodiment of the present invention;

[0058] Figure 3 A voltage space vector diagram of a two-level PWM rectifier used in an embodiment of the present invention;

[0059] Figure 4 A schematic diagram of η and actual power offset given values ​​under a reactive power change rate suppression control strategy provided by an embodiment of the present invention;

[0060] Figure 5 A flow chart of a control strategy for suppressing reactive power change rate provided by an embodiment of the present invention;

[0061] Figure 6 The following is a comparison diagram of experimental waveforms of the reactive power change rate suppression control method of the present invention and the traditional MPDPC control method provided in an embodiment, wherein Figure 6 a and Figure 6 b shows the DC bus voltage U when the traditional MPDPC control strategy is applied in steady state. dc , a-phase grid voltage e a and a phase grid-side current i a The waveform diagram and i a Spectrum analysis diagram of . Figure 6 c and Figure 6 d shows the DC bus voltage U when the reactive power change rate control strategy is applied in steady state. dc , a-phase grid voltage e a and a phase grid-side current i a The waveform diagram and i a Spectrum analysis diagram of

[0062] Figure 7 A dynamic load increase and decrease test diagram of obtaining a given active power value by using the improved adaptive super-helical sliding mode controller of the present invention is provided in one embodiment. Figure 7 a and Figure 7 b shows the DC bus voltage U of the loading and unloading experiment when the adaptive super-helical sliding mode controller is used instead of the traditional PI controller to obtain the active power given value. dc , a-phase grid voltage e a and a phase grid-side current i a ;

[0063] Figure 8 A schematic structural diagram of a two-level PWM rectifier fixed switching frequency model prediction direct power control device provided by an embodiment of the present invention.

[0064] It should be noted that these drawings and textual descriptions are not intended to limit the conceptual scope of the present invention in any way, but are intended to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The following embodiments are used to illustrate the present invention but are not used to limit the scope of the present invention.

[0066] In order to deepen the recognition and understanding of the present invention, the technical solution of the present invention is further introduced below in conjunction with the accompanying drawings and specific implementation methods.

[0067] Figure 1 The schematic diagram of the two-level PWM rectifier is given. Two power switching devices are connected in series in each phase bridge arm, and the DC link is stored by capacitor C; the grid-side input filter link is composed of a single L-type filter, and the resistor R is the equivalent resistance of the internal resistance of the power switching device and the internal resistance of the inductor. The two-level rectifier produces two valid switching states in each branch Sx∈{0,1},x∈{a,b,c}. Assuming that the voltage of the capacitor is equal to the DC link voltage, these two switching states will produce an output voltage (relative to Figure 1 The neutral point N) shown is u xN ∈{u dc ,0}.

[0068] To establish the discrete model of the two-level PWM rectifier, it is assumed that the power devices are ideal switching devices and the nonlinear effects such as the on-state voltage drop and switching delay of the power devices are ignored. Table I summarizes the relationship between the output voltage, switch state and gate signal generated by the converter.

[0069] Table 1 Relationship between output voltage, switch state and gate signal, x∈{a,b,c}

[0070]

[0071] Therefore, the combination of different switch states of the three phases abc can get 2 3 =8 basic switching states, namely (0,0,0), (0,0,1), (1,1,1), (1,1,0), (1,0,0), (1,0,1), (0,1,1), (0,1,0).

[0072] Now, combined with the 8 switch state inputs of the PWM rectifier, the prediction of the active power and new reactive power of the PWM rectifier at the k+1 moment corresponding to the 8 switch states is completed. Figure 2 As shown, in one embodiment of the present invention, a two-level PWM rectifier fixed switching frequency model prediction direct power control method is proposed. The method can be used in an unbalanced power grid. The specific method steps include:

[0073] Step S100: constructing a discrete domain PWM rectifier mathematical model in a two-phase stationary αβ coordinate system, and defining a new reactive power to replace the traditional reactive power.

[0074] In an embodiment of the present application, the method steps for constructing a discrete domain PWM rectifier mathematical model include:

[0075] Collect the PWM rectifier status information at time k. The status information includes the grid-side current i a (k), i b (k), i c (k), grid-side voltage e a (k), e b (k), e c (k), DC bus voltage u dc (k) and DC side load current i dc (k). The grid-side current, grid-side voltage, DC bus voltage and DC load current are sampled in real time by the hardware sampling circuit and sent to the digital signal processor. Based on the collected PWM rectifier state information at time k, a discrete domain PWM rectifier mathematical model is constructed in the two-phase static αβ coordinate system.

[0076] Furthermore, the mathematical model of the PWM rectifier in the discrete domain is constructed in the two-phase stationary αβ coordinate system as follows:

[0077]

[0078] Where, L and R are the inductance value and equivalent resistance of the filter inductor respectively; αβ (k) is the grid-side current of αβ axis at time k; e αβ (k) is the grid voltage of αβ axis at time k; v αβ (k) is the output voltage of the αβ axis rectifier on the AC side at time k. The variables in the αβ coordinate system can be transformed using Clarke transformation T αβ get.

[0079] i αβ (k) = T αβ i abc (k),e αβ (k) = T αβ e abc (k),v αβ (k) = T αβ v abc (k) i abc (k)=[i a (k),i b (k),i c (k)] T ,e abc (k) = [e a (k),e b (k),e c (k)] T ,v abc (k)=[v ao (k),v bo (k),v co (k)] T .

[0080]

[0081] In order to reduce the control complexity under unbalanced grid voltage, a new reactive power Q is defined nov,k Replace the traditional reactive power Q k , the new reactive power is the expanded reactive power.

[0082] Step S200: According to the constructed discrete domain PWM rectifier mathematical model, a dynamic equation of active power and new reactive power at time k is obtained.

[0083] The active power and new reactive power at time k are calculated from the grid voltage and its delay at time k and the grid-side current at time k. k With the new reactive power Q nov,k The calculation expression is:

[0084]

[0085] Among them, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β Delay value of 90° electrical angle, i α 、i β is the current on the αβ axis grid side.

[0086] Solve the derivatives of active power and new reactive power with respect to time t to obtain the dynamic equations of active power and new reactive power at time k:

[0087]

[0088] Under the condition of unbalanced grid voltage, the grid voltage e α With e β The derivative information of its delay can be expressed as:

[0089]

[0090] Where ω is the angular frequency of the fundamental component of the voltage.

[0091] Step S300: discretize the dynamic equations of active power and new reactive power to obtain predicted values ​​of active power and new reactive power at time k+1.

[0092] The derivative expressions of active power and new reactive power are discretized using the Euler forward method to obtain the predicted values ​​of active power and new reactive power at time k+1:

[0093]

[0094] Where, P and Qnov with superscript k+1 are single-step prediction values, and superscript k is the calculated value at the current moment; T s is the control period, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β The delay value of 90° electrical angle, w is the grid angular frequency, L is the grid side inductance, R is the line impedance, e is the grid voltage, i is the grid side current, v is the rectifier AC side voltage, v α 、v β It is the AC voltage vector of the rectifier under αβ axis.

[0095] Step S400: establishing a DC side power mathematical model based on an adaptive super-helical sliding mode controller, and obtaining an active power set value at time k through the mathematical model.

[0096] The numerical value and sign function of the sliding mode quantity are added to the adaptive law to achieve the goal of optimizing performance. A DC side power mathematical model based on an adaptive super-helical sliding mode controller is established. The active power given value at time k can be directly obtained through the mathematical model. The active power setpoint for direct power control is predicted by the model to meet the variable DC bus voltage demand.

[0097] In the embodiment of the present application, the basic form of the control rate of the super-helical sliding mode controller is:

[0098]

[0099] In the formula, σ is the sliding mode variable, α and β are adaptive gains, and sign is the sign function. The sign function can be expressed as:

[0100]

[0101] The adaptive law of α and β can be expressed as:

[0102]

[0103] In the formula, K p , δ0 is an arbitrary normal number.

[0104] The derivation process of the DC side power mathematical model based on the adaptive super-helical sliding mode controller is as follows, and the sliding mode variable σ is defined as:

[0105] σ=u dcref -u dc

[0106] In the formula, u dcref is the given value of DC bus voltage; u dc It is the DC bus voltage measurement value obtained by real-time sampling of the sensor.

[0107] Ignoring the turn-on loss and turn-off loss of the power switch device, the power conservation approximation on the AC and DC sides yields:

[0108]

[0109] Where P cap is the active power stored in the DC side capacitor; P ref is the active power given by the output of the adaptive super-helical sliding mode controller, P load is the active power consumed by the DC side load. Considering the general situation, assuming the load is a pure resistive load, then P load It can be expressed as:

[0110] P load =udc i dc

[0111] In the formula, i dc It is the DC side load current measurement value obtained by real-time sampling of the sensor.

[0112] Furthermore, the DC side power mathematical model based on the adaptive super-helical sliding mode controller can be expressed as:

[0113] P ref =u dc i dc +Cv dc (α|σ| 1 / 2 sign(σ)+∫βsign(σ)dt)

[0114] In the formula, u dc is the DC bus voltage measurement value obtained by real-time sampling of the sensor, i dc is the DC side load current measurement value obtained by real-time sampling of the sensor, C is the DC side capacitance, σ is the sliding mode variable, α and β are the adaptive gains, and sign is the sign function.

[0115] By improving the adaptive rate of the adaptive super-helical sliding mode controller, the dynamic response speed of the system is accelerated and the fluctuation of the active power set value on the DC side is reduced, thereby improving the dynamic performance and steady-state performance of the PWM rectifier.

[0116] Furthermore, the active power given value of the DC side at time k based on the improved adaptive super-helical sliding mode controller is:

[0117]

[0118] in, is the given value of active power at time k, u dc (k) is the DC bus voltage i at time k dc (k) is the DC bus current at time k, and C is the DC energy storage capacitor.

[0119] Step S500: construct a cost function g by constraining the sum of squares of errors between the predicted values ​​of active power and new reactive power and the given values.

[0120] In the embodiment of the present application, the cost function g is defined as the sum of squares of the errors between the single-step predicted power value and the given value. The cost function g constrains two performance indicators, active power error and new reactive power error:

[0121] g=(P ref -P k+1 ) 2 +(Q nov,ref -Q nov,k+1 )2

[0122] Among them, P ref With Q nov,ref are active power and new reactive power given values, P ref The output of the improved super-helical sliding mode controller is obtained, Q nov,ref To ensure unity power factor operation, it is generally set to 0var. k+1 With Q nov,k+1 are the predicted values ​​of active power and new reactive power.

[0123] Step S600: using the cost function to solve the partial derivatives of the voltage vectors in the two-phase stationary αβ coordinate system respectively, and calculating the expression of the optimal voltage vector online.

[0124] The theoretical basis for obtaining the optimal voltage vector vαβ_opt is: Since the cost function g explicitly contains the voltage vector v αβ , so the cost function g can be considered to be about the voltage vector v αβ The core of model predictive control is to solve the optimal voltage vector to minimize the cost function g. From the basic knowledge of advanced mathematics, we know that when the function solves the partial derivative of a variable and sets the partial derivative to zero, the variable at that moment is the extreme point. Similarly, in the two-phase stationary coordinate system, the cost function g is respectively the optimal voltage vector v α With v β Solve the partial derivative and find the voltage vector v that satisfies the zero partial derivative. αβ This is the optimal voltage vector.

[0125] The cost function g is used to calculate the voltage vector v α The partial derivative expression can be expressed as:

[0126] k[v α W1+v β W2]=A α +B α +C α

[0127] Among them, A α , B α , C α represents the non-optimal voltage vector term of the α-axis component, and W1 and W2 represent the optimal voltage vector term coefficients:

[0128]

[0129] Where k = 2L / 3T s , ΔP k =P ref -P k , ΔQ nov,k =-Qnov,k .

[0130] The cost function g is used to calculate the voltage vector v β The partial derivative expression can be expressed as:

[0131] k[v α W2+v β W3]=A β +B β +C β

[0132] Among them, A β , B β , C β It represents the non-optimal voltage vector term of the β-axis component, and W3 is the optimal voltage vector term coefficient.

[0133]

[0134] Using Cramer's law to solve the linear equations, the optimal voltage vector v can be obtained α_opt With v β_opt Online calculation expression:

[0135]

[0136] Among them, e α 、e β is the grid voltage in the two-phase stationary αβ coordinate system, A α , B α , C α represents the non-optimal voltage vector term of the α-axis component, W1 and W2 represent the optimal voltage vector term coefficients, and A β , B β , C β represents the non-optimal voltage vector term of the β-axis component, W3 is the optimal voltage vector term coefficient, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β Delay value for 90° electrical angle.

[0137] Step S700: Apply a reactive power change rate reduction control strategy, decompose the optimal voltage vector online expression into a reactive power change rate reduction term and an active power accurate tracking term, and determine the weight factor according to the operating state of the two-level PWM rectifier to optimize the optimal voltage vector expression.

[0138] This step effectively reduces the high-frequency harmonics and harmonic distortion rate of the grid-side current. In order to reduce the control complexity under unbalanced grid conditions, the extended reactive power is defined as a new reactive power to replace the traditional reactive power.

[0139] However, when the PWM rectifier is actually running, the actual reactive power on the grid side cannot be guaranteed to be 0var, and its quantitative relationship is constrained by the following formula:

[0140]

[0141] Where Q is the actual reactive power; A is the grid voltage imbalance, defined as for and The angle between the actual reactive power and the active power. The actual reactive power cannot be limited to 0var like the new reactive power. On the contrary, there is a double frequency fluctuation, and the peak value of the fluctuation is determined by the grid voltage imbalance and the active power reference value. Especially when When it is ±90°, the actual reactive power fluctuation reaches its peak value.

[0142] It is worth noting that the selection of the optimal voltage vector will directly affect the speed of change of Q, which is ignored in the construction of the cost function. The current cost function design is limited to minimizing the instantaneous active power deviation ΔP and the instantaneous newly defined reactive power deviation ΔQ nov The optimal voltage vector solved based on the cost function may cause a sharp fluctuation in the rate of change of Q, and even affect the increase or decrease trend of Q. When this happens, the grid-side current will be affected by the high-frequency oscillation of reactive power and produce a large current ripple, which will in turn affect the THD of the grid-side current and cause the operating temperature of the grid-side inductor to rise, thereby affecting the stability and efficiency of the system.

[0143] First, the influence of the optimal voltage vector on the change rate of reactive power Q is analyzed. The reactive power derivative expression is obtained based on the definition of reactive power and the mathematical model of the PWM rectifier in the αβ coordinate system:

[0144]

[0145] The optimal voltage vector v αβ_opt Substituting the reactive power Q change rate, we can get:

[0146]

[0147] In the formula,

[0148]

[0149] In deriving v αβ_opt After analyzing the influence of dQ / dt, the influencing factors can be summarized into two categories: one is the high-frequency pulsation part related to the power error, and the other is the steady-state component of the grid voltage and grid-side current. The steady-state component is relatively fixed, and its value does not change much under the operating state of the rectifier, and its coefficient includes T sTherefore, the dynamic characteristics of reactive power are mainly determined by the high-frequency changes of power error.

[0150] In summary, consider the optimal voltage vector v αβ_opt The online expression is decomposed into two parts: reducing the reactive power fluctuation term and accurately tracking the active power term. Constraints are imposed on the two terms respectively, and two weight factors are set to meet the control objectives with different requirements.

[0151] Specifically, the optimal voltage vector v αβ_opt The online expression is rewritten as:

[0152]

[0153] In the formula, and In order to suppress the reactive power fluctuation, and To accurately track the active power term. The corresponding expression is as follows:

[0154]

[0155] Among them, η is the weight factor for reducing the reactive power change rate, and μ is the weight factor for accurately tracking the active power term.

[0156] Furthermore, the principles for selecting the values ​​of η and μ are as follows: η is a fixed value, generally a positive number less than 1; the smaller the value of η, the less the high-frequency harmonic component of the grid-side current; μ is either 0 or 1, and in order to reduce the harmonic component of the grid-side current, μ is pre-set to 0; if the optimal voltage vector with μ of 0 cannot meet the given active power value, it is modified to 1.

[0157] When μ is equal to 1, the rewritten v αβ_opt The expression is exactly the same as the traditional expression. Under this condition, the system can achieve stable control of instantaneous active and new reactive power, but does not control the change rate of reactive power, resulting in more obvious fluctuations at the peak of the grid-side current. When μ is equal to 0, the instantaneous active and new reactive power of the system will fluctuate around the power given value, and the introduction of η effectively slows down the change rate of reactive power, thereby effectively reducing the THD of the grid-side current.

[0158] Step S800: Determine the value of the weight factor for reducing the reactive power fluctuation term and the weight factor for accurately tracking the active power term, and verify the optimal voltage vector v based on suppressing the reactive power change rate. αβ_opt Can the preset active power be met and the optimal voltage vector v be output? αβ_opt .

[0159] By artificially defining η as a weight factor for reducing the reactive power change rate, the influence of high-frequency power error on the reactive power change rate is effectively suppressed, and the increase and decrease trend of reactive power is effectively smoothed, thereby eliminating the high-frequency harmonic components of the grid-side current, effectively improving the grid-side current THD, and enhancing the steady-state operation performance of the PWM rectifier. However, the existence of η will make the PWM rectifier unable to accurately track the active power. The specific underlying mechanism is Figure 4 We can get:

[0160] Figure 4 Taking active power as an example, kT s At this moment, the processor enters an interrupt and performs ADC sampling and calculation of the optimal voltage vector. At this time, the actual power deviation of the system is P ref -P k However, the power deviation involved in the optimal voltage vector calculation in the interrupt program is adjusted to η(P ref -P k ). Since the value of η is less than 1, the power deviation during the processor operation is lower than the actual deviation. This setting causes the system to no longer track the set power P ref , but P affected by η real .

[0161] However, the expression of the optimal voltage vector shows that the existence of power deviation is a necessary condition for stable operation. Only when power deviation occurs can the system dynamically adjust the optimal voltage vector according to the deviation to ensure that the set working conditions are met. Therefore, it is impossible to achieve zero-static-error control of active power and reactive power without using the control strategy of suppressing the reactive power change rate. Figure 4 As shown, if η is set properly, P real In P ref Small fluctuations around P are acceptable. On the contrary, if η is set too low, ref When a sudden load change occurs, the power deviation received by the processor is too small, making it impossible for the rectifier to maintain stable operation. Therefore, a reasonable choice of η is crucial to ensure system stability.

[0162] The present invention indirectly controls η by adjusting the value of μ to ensure the operating stability of the rectifier. In certain stable operating scenarios, directly adjusting η to limit the rate of change of reactive power may encounter disadvantages. If it is assumed that the system is in a steady state, η is also relatively stable. When P refA sudden increase will inevitably lead to an increase in ΔP. To ensure the stable operation of the system, η needs to be adjusted in real time according to its preset change rules. The design of the change rules is directly related to the operating efficiency of the system. If the η change rate is set too slow, the dynamic response capability of the system will be affected; conversely, if the η change rate is too fast, it may lead to a sharp adjustment of η, causing a new round of fluctuations in the reactive power change rate. Based on this, setting η to a more conservative value and using μ to fine-tune η becomes a feasible strategy.

[0163] The control strategy process of suppressing reactive power change rate is as follows: Figure 5 As shown, when the processor enters the interrupt, it first completes the sampling of the variables required by S100. After that, the optimal voltage vector is calculated by using the control strategy of suppressing the reactive power change rate. The specific expression is as follows:

[0164]

[0165] At this time, μ is equal to 0. After obtaining the optimal voltage vector, it is substituted into the active power expression at time k+1:

[0166]

[0167] v αβ_opt Substitute v into the above formula αβ Get the theoretical active power value P(k+1) of the system at time k+1. Set an acceptable active power fluctuation range, which is set to 0.2P in this paper. ref The specific fluctuation range can be modified in real time according to the actual application scenario. If you are concerned about the harmonic interference of electrical equipment on the power grid, the range can be appropriately increased; if you are concerned about whether the DC side voltage of the PWM rectifier can be output stably, the range can be appropriately reduced. Finally, by comparing P ref The difference between P(k+1) and P(k+1) is used to determine the control strategy of suppressing the reactive power change rate v αβ_opt Whether it meets the user's needs. If P(k+1) is within the acceptable power range, the optimal voltage vector is directly output.

[0168] If it is outside the interval, the optimal voltage vector needs to be recalculated. At this time, μ is equal to 1, and the calculated optimal voltage vector is directly output. αβ_opt The expression changes to:

[0169]

[0170] Among them, e α 、e β is the grid voltage in the two-phase stationary αβ coordinate system, A α , B α , C αrepresents the non-optimal voltage vector term of the α-axis component, W1 and W2 represent the optimal voltage vector term coefficients, and A β , B β , C β represents the non-optimal voltage vector term of the β-axis component, W 3 is the optimal voltage vector term coefficient, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β Delay value for a delay of 90° electrical angle.

[0171] Step S900: The optimal voltage vector v αβ_opt The generated switching signal is applied to the two-level PWM rectifier.

[0172] Specifically, the optimal voltage vector v determined in step S900 is αβ_opt The corresponding gate level signal s is generated by the SVPWM module abc (k), and apply the corresponding gate signal to the two-level PWM rectifier. The voltage space vector diagram of the two-level PWM rectifier is shown in Figure 2. Figure 3 shown.

[0173] In order to verify the feasibility of the switching frequency limiting algorithm under the limited control set described in the present invention, an experimental platform is built for verification. a / E b / E c The grid voltage frequency is f = 50Hz, the filter inductor is L = 4.3mH, the DC side capacitor is C = 3200μF, and the DC load is R L =75Ω / 50Ω. The control parameter of the improved super-helical sliding mode controller is K p =0.12, δ0=0.05, λ=2, ε=1. The weight factor of the reactive power conversion rate control strategy is η=0.8, μ=0 or 1.

[0174] Set the DC bus voltage set value U dc * =160V, the fixed frequency model predicts that the direct power control switching frequency is 10kHz, Figure 6 a and Figure 6 b shows the DC bus voltage U when the traditional MPDPC control strategy is applied in steady state. dc , a-phase grid voltage e a and a phase grid-side current i a The waveform diagram and i a Spectrum analysis diagram of . Figure 6c and Figure 6 d shows the DC bus voltage U when the reactive power change rate control strategy is applied in steady state. dc , a-phase grid voltage e a and a phase grid-side current i a The waveform diagram and i a Spectrum analysis diagram of . Figure 6 It can be seen that both the traditional algorithm and the algorithm provided by the present invention can operate stably, the bus voltage reaches a given value, and the current distortion rate of the grid-side current is lower than the national standard. When the reactive power change rate suppression control strategy is adopted, the reactive power change is relatively smooth due to the effective constraint of the reactive power change rate, so the grid-side current has fewer high-frequency harmonic components and the steady-state performance can be further improved. Experiments show that the proposed reactive power change rate suppression control strategy can effectively reduce the THD of the grid-side current.

[0175] Figure 7 a and Figure 7 b shows the DC bus voltage U of the loading and unloading experiment when the adaptive super-helical sliding mode controller is used instead of the traditional PI controller to obtain the active power given value. dc , a-phase grid voltage e a and a phase grid-side current i a .Depend on Figure 7 It can be seen that under the load addition and reduction test, the DC side voltage hardly drops or rises, and remains stable at the given value. Therefore, the experiment shows that the adaptive super-helical sliding mode controller has stronger robustness and dynamic response than the PI controller.

[0176] The present invention introduces the extended reactive power theory, which effectively simplifies the calculation complexity. On the basis of the finite set model predictive direct power control (FCS-MPDPC), an online calculation module for the optimal voltage vector is added to shorten the program running time and realize the fixed switching frequency. In addition, in order to solve the problem of high-frequency oscillation of reactive power caused by applying the optimal voltage vector, a FSF-MPDPC that reduces the reactive power change rate is proposed, which effectively reduces the grid-side current ripple and improves the current harmonic distortion rate (THD). Finally, an adaptive super-helical sliding mode (ASTW) controller is used to replace the traditional PI controller to obtain the power setting. The ASTW controller has a faster dynamic response speed and stronger anti-disturbance performance, and effectively reduces the power fluctuation in steady-state operation, and improves the steady-state and dynamic performance of the PWM rectifier.

[0177] In one embodiment, a two-level PWM rectifier fixed switching frequency model prediction direct power control device is shown, and the device includes:

[0178] PWM rectifier mathematical model building module, used to build a discrete domain PWM rectifier mathematical model in a two-phase static αβ coordinate system, and define a new reactive power to replace the traditional reactive power;

[0179] A dynamic equation module is used to obtain the dynamic equation of active power and new reactive power at time k according to the constructed discrete domain PWM rectifier mathematical model;

[0180] The prediction value module is used to discretize the dynamic equations of active power and new reactive power to obtain the prediction values ​​of active power and new reactive power at time k+1;

[0181] Mathematical model module, used to establish a DC side power mathematical model based on an adaptive super-helical sliding mode controller, and obtain the active power given value at time k through the mathematical model;

[0182] A cost function construction module is used to construct a cost function by constraining the sum of squares of errors between the predicted values ​​of active power and new reactive power and the given values;

[0183] An online optimal voltage vector calculation module is used to solve partial derivatives of the voltage vectors in the two-phase stationary αβ coordinate system using a cost function, and calculate the expression of the optimal voltage vector online;

[0184] The reactive power change rate suppression control strategy module is used to apply the reactive power change rate suppression control strategy, decompose the optimal voltage vector online expression into a reactive power change rate reduction term and an active power accurate tracking term, and determine the weight factor according to the operating state of the two-level PWM rectifier to optimize the optimal voltage vector expression;

[0185] The module for checking and tracking the active power given value is used to determine the value of the weight factor for reducing the reactive power fluctuation term and accurately tracking the active power term weight factor, and to verify whether the optimal voltage vector based on suppressing the reactive power change rate can meet the preset active power given value and output the optimal voltage vector;

[0186] The action module is used for applying the optimal voltage vector generation switching signal to the two-level PWM rectifier.

[0187] It should be noted that the two-level PWM rectifier model-determined switching frequency prediction direct power control device provided in the above embodiment only uses the division of the above functional modules as an example when executing a two-level PWM rectifier model-determined switching frequency prediction direct power control method. In actual applications, the above functional allocation can be completed by different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the two-level PWM rectifier model-determined switching frequency prediction direct power control device provided in the above embodiment and the two-level PWM rectifier model-determined switching frequency prediction direct power control method embodiment belong to the same concept, and the embodiment and implementation process thereof are detailed in the two-level PWM rectifier model-determined switching frequency prediction direct power control method embodiment, which will not be repeated here.

[0188] In one embodiment, a computer-readable storage medium is provided, on which computer program instructions are stored. When the instructions are executed by a processor, the above-mentioned two-level PWM rectifier model-determined switching frequency prediction direct power control method is implemented.

[0189] The above is only a preferred embodiment of the present invention, and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this patent can make some changes or modify the technical contents suggested above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the invention without departing from the content of the technical solution of the present invention still fall within the scope of the solution of the present invention.

Claims

1. A two-level PWM rectifier fixed switching frequency model prediction direct power control method, characterized in that: The method comprises: A mathematical model of PWM rectifier in discrete domain is constructed in two-phase stationary αβ coordinate system, and a new reactive power is defined to replace the traditional reactive power. According to the constructed discrete domain PWM rectifier mathematical model, the dynamic equations of active power and new reactive power at time k are obtained; Discretize the dynamic equations of active power and new reactive power to obtain the predicted values ​​of active power and new reactive power at time k+1; A DC power mathematical model based on an adaptive super-helical sliding mode controller is established, and the active power given value at time k is obtained through the mathematical model. The cost function is constructed by constraining the sum of squares of the errors between the predicted values ​​of active power and new reactive power and the given values; The cost function is used to solve the partial derivatives of the voltage vectors in the two-phase stationary αβ coordinate system, and the expression of the optimal voltage vector is calculated online. The control strategy of suppressing reactive power change rate is applied to decompose the optimal voltage vector online expression into the term of reducing reactive power change rate and the term of accurately tracking active power. The weight factor is determined by the operating state of the two-level PWM rectifier to optimize the optimal voltage vector expression. Determine the value of the weight factor for reducing reactive power fluctuation and the weight factor for accurately tracking active power, and verify whether the optimal voltage vector based on suppressing the reactive power change rate can meet the preset active power setting and output the optimal voltage vector; The optimal voltage vector generates a switching signal and applies it to the two-level PWM rectifier.

2. The two-level PWM rectifier fixed switching frequency model prediction direct power control method according to claim 1, characterized in that: The predicted values ​​of active power and new reactive power at time k+1 are: Where, P and Qnov with superscript k+1 are single-step prediction values, and superscript k is the calculated value at the current moment; T s is the control period, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β The delay value of 90° electrical angle, w is the grid angular frequency, L is the grid-side inductance, R is the line impedance, e is the grid voltage, i is the grid-side current, v is the rectifier AC side voltage, v α 、v β It is the AC voltage vector of the rectifier under αβ axis.

3. The two-level PWM rectifier fixed switching frequency model prediction direct power control method according to claim 1 or 2, characterized in that: The active power given value P at time k is obtained by the mathematical model ref for: P ref =u dc I dc +Cv dc (a|s| 1 / 2 sign(σ)+∫βsign(σ)dt) In the formula, u dc is the DC bus voltage measurement value obtained by real-time sampling of the sensor, i dc is the DC side load current measurement value obtained by real-time sampling of the sensor, C is the DC side capacitance, σ is the sliding mode variable, α and β are the adaptive gains, and sign is the sign function.

4. The two-level PWM rectifier constant switching frequency model prediction direct power control method according to claim 3, characterized in that: The cost function constructed is: g=(P ref -P k+1 ) 2 +(Q nov,ref -Q nov,k+1 ) 2 Among them, P ref With Q nov,ref are the active power and new reactive power given values, Q nov,ref To ensure unity power factor operation, it is generally set to 0var; P k+1 With Q nov,k+1 They are the predicted values ​​of active power and new reactive power respectively.

5. The two-level PWM rectifier constant switching frequency model prediction direct power control method according to claim 1, characterized in that: The control strategy for suppressing reactive power change rate is: The influence of the optimal voltage vector on dQ / dt is derived, and the influencing factors are summarized into two categories: one is the high-frequency pulsation part related to the power error, and the other is the steady-state component of the grid voltage and grid-side current; the steady-state component is relatively fixed, and its value does not change much under the operating state of the rectifier, and its coefficient includes T s The dynamic characteristics of reactive power are mainly determined by the high-frequency changes of power error.

6. The two-level PWM rectifier constant switching frequency model prediction direct power control method according to claim 5, characterized in that: The optimal voltage vector online expression is decomposed into a term for reducing the reactive power change rate and a term for accurately tracking the active power, and its expression is: In the formula, and In order to suppress the reactive power fluctuation, and To accurately track the active power term, the corresponding expression is as follows: Where η is the weight factor for reducing the reactive power change rate, and μ is the weight factor for accurately tracking the active power term.

7. The method for direct power control of a two-level PWM rectifier using a fixed switching frequency model prediction method according to claim 6, characterized in that: The principles for selecting the weight factor η for reducing the reactive power change rate and the weight factor μ for accurately tracking the active power are as follows: η is a fixed value, which is a positive number less than 1; the smaller the value of η is, the fewer the high-frequency harmonic components of the grid-side current are; μ is either 0 or 1, and in order to reduce the harmonic components of the grid-side current, μ is pre-set to 0; if the optimal voltage vector with μ of 0 cannot meet the given active power value, it is modified to 1.

8. The two-level PWM rectifier constant switching frequency model prediction direct power control method according to claim 7, characterized in that: The verification is based on whether the optimal voltage vector for suppressing the reactive power change rate can meet the active power given value, including: The optimal voltage vector v is calculated by using the control strategy of suppressing the reactive power change rate αβ_opt , then μ is equal to 0: After obtaining the optimal voltage vector, substitute it into the active power expression at time k+1: Establish an acceptable active power fluctuation range; the specific fluctuation range is modified in real time according to the actual application occasion, and the P ref With P k+1 The difference between the two determines the optimal voltage vector v under the control strategy of suppressing the reactive power change rate. αβ_opt Whether it meets the needs of users; if P k+1 If the voltage is within the acceptable power range, the optimal voltage vector is directly output; if it is outside the range, the optimal voltage vector needs to be recalculated, and the corresponding μ is equal to 1, and the calculated optimal voltage vector is directly output. The optimal voltage vector v αβ_opt The expression changes to: Among them, e α 、e β is the grid voltage in the two-phase stationary αβ coordinate system, A α , B α , C α represents the non-optimal voltage vector term of the α-axis component, W1 and W2 represent the optimal voltage vector term coefficients, A β , B β , C β represents the non-optimal voltage vector term of the β-axis component, W3 is the optimal voltage vector term coefficient, e′ α and e′ β They represent the grid voltage e in the two-phase stationary αβ coordinate system. α With e β Delay value for 90° electrical angle.

9. A two-level PWM rectifier fixed switching frequency model prediction direct power control device, characterized in that: The device comprises: PWM rectifier mathematical model building module, used to build a discrete domain PWM rectifier mathematical model in a two-phase stationary αβ coordinate system, and define a new reactive power to replace the traditional reactive power; A dynamic equation module is used to obtain the dynamic equation of active power and new reactive power at time k according to the constructed discrete domain PWM rectifier mathematical model; The prediction value module is used to discretize the dynamic equations of active power and new reactive power to obtain the prediction values ​​of active power and new reactive power at time k+1; Mathematical model module, used to establish a DC side power mathematical model based on an adaptive super-helical sliding mode controller, and obtain the active power given value at time k through the mathematical model; A cost function construction module is used to construct a cost function by constraining the sum of squares of errors between the predicted values ​​of active power and new reactive power and the given values; An online optimal voltage vector calculation module is used to solve partial derivatives of voltage vectors in two-phase stationary αβ coordinate systems using cost functions, and to calculate the expression of the optimal voltage vector online; The reactive power change rate suppression control strategy module is used to apply the reactive power change rate suppression control strategy, decompose the optimal voltage vector online expression into a reactive power change rate reduction term and an active power accurate tracking term, and determine the weight factor according to the operating state of the two-level PWM rectifier to optimize the optimal voltage vector expression; The module for checking and tracking the active power given value is used to determine the value of the weight factor for reducing the reactive power fluctuation term and accurately tracking the active power term weight factor, and to verify whether the optimal voltage vector based on suppressing the reactive power change rate can meet the preset active power given value and output the optimal voltage vector; The action module is used for applying the optimal voltage vector generation switching signal to the two-level PWM rectifier.

10. A computer-readable storage medium having computer program instructions stored thereon, wherein when the instructions are executed by a processor, the two-level PWM rectifier model-determined switching frequency prediction direct power control method according to any one of claims 1 to 8 is implemented.

Citation Information

Cited By

  • Power prediction active damping control method and device for L-Traps-CL filtering type rectifying device

    CN120855909A

  • Position-sensing-free multi-vector predictive control method and system for permanent magnet synchronous motor

    CN120896497A

  • Active front-end rectifier and predictive control method thereof

    CN121307993A