Quasi-cascadeless model predictive control method and device for PWM rectifier

Abstract

The present invention discloses a quasi-cascadeless model predictive control method and device for PWM rectifiers. The present invention aims at the problem of insufficient dynamic response and anti-disturbance capability of DC bus voltage of PWM rectifier under traditional cascade control structure. According to the discrete mathematical model of AC and DC sides of PWM rectifier, DC bus voltage and grid current prediction values ​​under different switching states are calculated; an auxiliary controller is constructed to obtain grid current reference value, and DC bus voltage and grid current are selected as control items and constraint items of cost function respectively, and weight coefficient of constraint item is selected according to grid current prediction value; the switching state that minimizes cost function is found, and it is applied to PWM rectifier of next control cycle. The method of the present invention is mainly used for DC bus voltage control of PWM rectifier or similar topological structure, without traditional DC bus voltage-grid current double closed-loop cascade structure, and has better dynamic response capability and anti-disturbance capability.

Description

Technical Field

[0001] The present application relates to the research field of PWM rectifier model predictive control, specifically a quasi-cascadeless model predictive control method and device for PWM rectifier, which can improve the dynamic response capability and anti-disturbance capability of the DC bus voltage of the PWM rectifier. Background Art

[0002] Due to the advantages of bidirectional power flow, good current sinusoidality and controllable DC bus voltage, PWM rectifiers have received widespread attention and are used in many fields, such as high-voltage DC transmission, wind power generation, solar power generation and energy storage. In most applications, the main function of PWM rectifiers is to control the DC bus voltage to ensure excellent dynamic response and anti-disturbance capabilities.

[0003] In order to achieve this control goal, the cascade proportional-integral (PI) control structure is the most commonly used, that is, the inner loop controls the grid-connected current of the PWM rectifier, and the outer loop adjusts the DC bus voltage to provide a reference for the current inner loop. However, due to the limitations of the PI controller bandwidth and the integral link, the dynamic performance and anti-disturbance capability based on the cascade PI control need to be improved.

[0004] Model predictive control has fast dynamic response and strong anti-interference ability, and is a better choice for PWM rectifier control. J.Rodriguez, P.Falkowski, etc. introduced model predictive control into the current / power control link of PWM rectifier, improving the current / power response capability. Y.Zhang, etc. proposed model-free predictive control, which improved the parameter robustness of model predictive control. It is worth noting that in the above-mentioned model predictive control research, a DC bus voltage control loop is still required, and the control loop is based on a PI controller (PI-MPC). Therefore, even if the inner loop has a fast dynamic response, the overall control performance is still limited by the outer loop controller, and the inherent advantages of model predictive control cannot be fully utilized.

[0005] If the DC bus voltage is used as the target of model predictive control, the optimal switch state is directly selected according to the predicted value of the DC bus voltage, and a cascade-free predictive control structure can be formed without the outer loop based on the PI controller. However, due to the nonlinear characteristics of the PWM rectifier, when the DC bus voltage is controlled, the PWM rectifier may operate at an unacceptable balance point, such as extremely large AC current and extremely small modulation voltage. E. Quevedo et al. added active power as an additional control variable in the cost function to ensure that the PWM rectifier operates at a reasonable balance point, but the reference value of the active power needs to be obtained through complex calculations, and the DC bus voltage reference value is obtained by a step-by-step approach, which limits the dynamic response capability. F. Wang et al. used the load power in the cost function to ensure the correct steady-state balance point, but this method not only requires load-side information, but also requires the use of a corresponding observer to predict the load power of the next control cycle.

[0006] Therefore, how to reduce the complexity of the DC bus voltage model predictive control algorithm, reduce the dependence on additional information, and improve the dynamic response and anti-disturbance capabilities are the key to PWM rectifier control. Summary of the invention

[0007] In view of this, the embodiment of the present application provides a quasi-cascadeless model predictive control method and device for PWM rectifiers, which directly selects the optimal switch state according to the predicted value of the DC bus voltage, adds a grid-connected current constraint term to the cost function, and the reference value of the grid-connected current constraint term is obtained through an auxiliary controller. The method can directly and effectively control the DC bus voltage, has a simple control algorithm, does not require additional information, and improves the dynamic response and anti-disturbance capability of the DC bus voltage.

[0008] According to a first aspect of an embodiment of the present application, a quasi-cascadeless model predictive control method for a PWM rectifier is provided, comprising:

[0009] In one control cycle, the three-phase grid voltage sampling value, the three-phase grid current sampling value, the DC bus voltage sampling value and the load current sampling value are collected at the current moment;

[0010] According to the three-phase grid voltage sampling value and the three-phase grid-connected current sampling value, coordinate transformation is performed according to the grid angle obtained by the phase-locked loop to obtain the dq-axis components of the grid voltage and the grid-connected current in the rotating coordinate system;

[0011] Substituting the DC bus voltage sampling value, the load current sampling value, and the dq-axis components of the grid voltage and the grid-connected current into the PWM rectifier discrete mathematical model, the predicted value of the dq-axis component of the grid-connected current and the predicted value of the DC bus voltage under different switching states are calculated;

[0012] According to the predicted value of the dq-axis component of the grid-connected current, the sampled value of the DC bus voltage and the predicted value of the DC bus voltage under the different switching states, an auxiliary controller is designed to obtain a reference value of the dq-axis component of the grid-connected current, the square of the difference between the reference value and the predicted value of the DC bus voltage is selected as the DC bus voltage control item in the cost function, and the square of the difference between the dq-axis reference value and the predicted value of the grid-connected current is introduced as the grid-connected current constraint item in the cost function, and the weight coefficient of the grid-connected current constraint item is selected according to the predicted value of the dq-axis component of the grid-connected current, so as to construct a cost function of the quasi-cascadeless model predictive control;

[0013] According to the cost function, a switch state that minimizes the cost function is found and applied to the PWM rectifier in the next control cycle.

[0014] Optionally, according to the DC bus voltage sampling value, the load current sampling value, and the dq-axis components of the grid voltage and the grid-connected current, substituting them into a PWM rectifier discrete mathematical model, calculating the predicted value of the dq-axis component of the grid-connected current and the predicted value of the DC bus voltage under different switching states, including:

[0015] According to the dq-axis components of the grid voltage and the grid-connected current, the mathematical model of the AC side of the PWM rectifier in the rotating coordinate system is discretized using the forward Euler method, and the dq-axis components of the grid voltage and the grid-connected current are substituted into the mathematical model of the AC side of the PWM rectifier to obtain the predicted value of the dq-axis component of the grid-connected current of the PWM rectifier;

[0016] According to the DC bus voltage sampling value, the predicted value of the grid voltage dq axis component and the grid current dq axis component, based on the AC and DC side power balance relationship of the PWM rectifier, the DC side current of the PWM rectifier at the current moment is calculated;

[0017] According to the DC bus voltage sampling value and the DC side current, the mathematical model of the DC bus capacitor of the PWM rectifier is discretized by using the forward Euler method to calculate the predicted value of the DC bus voltage of the PWM rectifier.

[0018] Optionally, according to the predicted value of the grid-connected current dq-axis component under the different switching states, the DC bus voltage sampling value and the predicted value of the DC bus voltage, an auxiliary controller is designed to obtain a reference value of the grid-connected current dq-axis component, the square of the difference between the DC bus voltage reference value and the predicted value is selected as the DC bus voltage control item in the cost function, and the square of the difference between the grid-connected current dq-axis reference value and the predicted value is introduced as the grid-connected current constraint item in the cost function, and the weight coefficient of the grid-connected current constraint item is selected according to the predicted value of the grid-connected current dq-axis component, and the cost function of the quasi-cascadeless model predictive control is constructed, including:

[0019] According to the DC bus voltage sampling value, proportional integral control is selected as the auxiliary controller, the difference between the DC bus voltage reference value and the sampling value is sent to the auxiliary controller, and the output value of the auxiliary controller is taken as the reference value of the d-axis component of the grid-connected current; considering the high power factor operation of the rectifier, 0 is taken as the reference value of the q-axis component of the grid-connected current;

[0020] According to the DC bus voltage prediction value, the square of the difference between the DC bus voltage reference value and the prediction value is selected as the DC bus voltage control item in the cost function; according to the grid current dq axis component reference value and the prediction value, the square of the difference between the grid current dq axis component reference value and the prediction value is selected as the grid current constraint item in the cost function;

[0021] According to the predicted value of the grid-connected current dq-axis component, the square sum of the predicted value of the grid-connected current d-axis component and the predicted value of the grid-connected current q-axis component is calculated; the weight coefficient of the DC bus voltage control item is taken as one, and if the square sum is less than or equal to the square of the current threshold, the weight coefficient of the grid-connected current constraint item in the cost function is taken to be two to three orders of magnitude smaller than the weight coefficient of the DC bus voltage control item; if the square sum is greater than the square of the current threshold, the weight coefficient of the grid-connected current constraint item in the cost function is taken to be two to three orders of magnitude larger than the weight coefficient of the DC bus voltage control item;

[0022] According to the DC bus voltage control item, the grid current constraint item and the grid current constraint item weight coefficient, the product of the grid current constraint item and the grid current constraint item weight coefficient is calculated, and the sum of the DC bus voltage control item and the product is set as the cost function of the quasi-cascadeless model predictive control.

[0023] According to a second aspect of an embodiment of the present application, there is provided a quasi-cascadeless model predictive control device for a PWM rectifier, comprising:

[0024] The sampling module is used to collect the three-phase grid voltage sampling value, the three-phase grid current sampling value, the DC bus voltage sampling value and the load current sampling value at the current moment within a control cycle;

[0025] A coordinate conversion module is used to perform coordinate conversion according to the three-phase grid voltage sampling value and the three-phase grid-connected current sampling value and the grid angle obtained by the phase-locked loop to obtain the dq-axis components of the grid voltage and the grid-connected current in the rotating coordinate system;

[0026] A calculation module, used to substitute the DC bus voltage sampling value, the load current sampling value, and the dq-axis components of the grid voltage and the grid-connected current into a discrete mathematical model of the PWM rectifier, and calculate the predicted value of the dq-axis component of the grid-connected current and the predicted value of the DC bus voltage under different switching states;

[0027] A construction module, used to design an auxiliary controller to obtain a reference value of a dq-axis component of a grid-connected current according to the predicted value of a dq-axis component of a grid-connected current, a sampled value of a DC bus voltage, and a predicted value of a DC bus voltage under the different switching states, select the square of the difference between the DC bus voltage reference value and the predicted value as a DC bus voltage control item in a cost function, introduce the square of the difference between the dq-axis reference value of the grid-connected current and the predicted value as a grid-connected current constraint item in the cost function, select a weight coefficient of the grid-connected current constraint item according to the predicted value of the dq-axis component of the grid-connected current, and construct a cost function of a quasi-cascadeless model predictive control;

[0028] The control module is used to find out the switch state that minimizes the cost function according to the cost function, and apply the switch state to the PWM rectifier of the next control cycle.

[0029] According to a third aspect of an embodiment of the present application, there is provided an electronic device, including:

[0030] one or more processors;

[0031] A memory for storing one or more programs;

[0032] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in the first aspect.

[0033] According to a third aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which computer instructions are stored. When the instructions are executed by a processor, the steps of the method described in the first aspect are implemented.

[0034] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0035] It can be seen from the above embodiments that the method of the present invention does not require a traditional cascade control structure, and directly selects the optimal switching state based on the DC bus voltage, so that the dynamic response and anti-disturbance capability of the DC bus voltage are more excellent; the present invention adds a grid-connected current constraint term to the cost function, and flexibly adjusts its weight coefficient to ensure that the PWM rectifier operates at a suitable steady-state equilibrium point; the present invention calculates the current reference value through an auxiliary controller without the need for additional information, thereby simplifying the algorithm complexity; the present invention provides ideas for improving the dynamic response and anti-disturbance capability of similar cascade control structures.

[0036] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0038] Figure 1 The present invention is a flow chart showing a quasi-cascadeless model predictive control method for a PWM rectifier according to an exemplary embodiment.

[0039] Figure 2 This is the main circuit structure diagram of the PWM rectifier;

[0040] Figure 3 The control structure diagram of the PWM rectifier based on quasi-cascadeless model predictive control (QNC-MPC);

[0041] FIG4( a ) is a diagram showing the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and QNC-MPC when the DC bus voltage reference value increases from 200 V to 220 V;

[0042] FIG4( b ) is a diagram showing the simulation results of the three-phase grid-connected current of the double PI cascade when the DC bus voltage reference value increases from 200 V to 220 V;

[0043] Figure 4(c) is a diagram showing the simulation results of the three-phase grid-connected current of PI-MPC when the DC bus voltage reference value increases from 200V to 220V;

[0044] Figure 4(d) is a diagram showing the simulation results of the three-phase grid-connected current of QNC-MPC when the DC bus voltage reference value increases from 200V to 220V;

[0045] FIG5(a) is a diagram showing the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and QNC-MPC when the DC bus voltage reference value drops from 220 V to 200 V;

[0046] FIG5( b ) is a diagram showing the simulation results of the three-phase grid-connected current of the double PI cascade when the DC bus voltage reference value drops from 220 V to 200 V;

[0047] Figure 5(c) shows the simulation result of the three-phase grid-connected current of PI-MPC when the DC bus voltage reference value drops from 220V to 200V;

[0048] Figure 5(d) is a diagram showing the simulation results of the three-phase grid-connected current of QNC-MPC when the DC bus voltage reference value drops from 220V to 200V;

[0049] Figure 6(a) shows the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and QNC-MPC when a 200Ω load is suddenly added;

[0050] Figure 6(b) is a simulation result of the three-phase grid-connected current of the dual PI cascade when a 200Ω load is suddenly added;

[0051] Figure 6(c) shows the simulation result of the three-phase grid-connected current of PI-MPC when a 200Ω load is suddenly added;

[0052] Figure 6(d) shows the simulation result of the three-phase grid-connected current of QNC-MPC when a 200Ω load is suddenly added;

[0053] FIG7( a ) is a diagram showing the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and QNC-MPC when the load suddenly drops by 200Ω;

[0054] FIG7( b ) is a diagram showing the simulation results of the three-phase grid-connected current of the dual PI cascade when the load suddenly drops by 200Ω;

[0055] Figure 7(c) shows the simulation result of the three-phase grid-connected current of PI-MPC when the load suddenly drops by 200Ω;

[0056] Figure 7(d) shows the simulation result of the three-phase grid-connected current of QNC-MPC when the load suddenly drops by 200Ω;

[0057] Figure 8 The figure is a block diagram of a quasi-cascadeless model predictive control device for a PWM rectifier according to an exemplary embodiment. DETAILED DESCRIPTION

[0058] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0059] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0060] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0061] Figure 1 is a flow chart of a quasi-cascadeless model predictive control method for a PWM rectifier according to an exemplary embodiment. Figure 1 As shown, the method may include the following steps:

[0062] S1: In one control cycle, the three-phase grid voltage sampling value, the three-phase grid current sampling value, the DC bus voltage sampling value and the load current sampling value are collected at the current moment;

[0063] Specifically, taking a 2.0kW PWM rectifier as an example, the current direction adopts the generator convention, the grid line voltage is 110V, and the PWM rectifier filter inductor L s =2mH, resistor R s =0.127Ω, DC bus capacitance C = 4000μF;

[0064] Figure 2 The main circuit structure diagram of the PWM rectifier is used to collect the three-phase grid voltage U at the current time (time k). ga (k) U gb (k) U gc (k), the three-phase grid-connected current I at the current moment sa (k) I sb (k) I sc (k), the DC bus voltage V at the current moment dc (k), current load current I L (k), providing basic information for quasi-cascadeless model predictive control of PWM rectifier.

[0065] S2: performing coordinate transformation according to the three-phase grid voltage sampling values ​​and the three-phase grid-connected current sampling values ​​and the grid angle obtained by the phase-locked loop to obtain the dq-axis components of the grid voltage and the grid-connected current in the rotating coordinate system;

[0066] Specifically, the current three-phase grid voltage sampling value U ga (k) U gb (k) U gc (k) is sent to the phase-locked loop to adjust the grid voltage phase θ g Tracking; the sampled three-phase grid voltage U ga (k) U gb (k) U gc (k) and grid-connected current I sa (k) I sb (k) I sc (k) transforming into the dq-axis components of the synchronous rotating coordinate system through Clark and Park transformation, thereby providing the required information for calculating the grid-connected current and DC bus voltage prediction values;

[0067] The current grid voltage dq-axis component U gd (k) U gq (k) is:

[0068]

[0069] Among them, T Clark is the Clark transformation matrix, T Park is the Park transformation matrix;

[0070] The current grid-connected current dq-axis component I sd (k) I sq (k) is:

[0071]

[0072] S3: Substituting the DC bus voltage sampling value, the load current sampling value, and the dq-axis components of the grid voltage and the grid-connected current into the PWM rectifier discrete mathematical model, and calculating the predicted values ​​of the dq-axis components of the grid-connected current and the predicted values ​​of the DC bus voltage under different switching states; this step includes the following sub-steps:

[0073] S31: discretizing the mathematical model of the AC side of the PWM rectifier in the rotating coordinate system by using the forward Euler method according to the dq-axis components of the grid voltage and the grid-connected current, substituting the dq-axis components of the grid voltage and the grid-connected current into the mathematical model of the AC side of the PWM rectifier, and obtaining the predicted value of the dq-axis components of the grid-connected current of the PWM rectifier;

[0074] Specifically, according to the PWM AC side topology, its mathematical model is:

[0075]

[0076] Among them, U gd and U gq is the dq axis component of the grid voltage, I sd and I q is the dq axis component of the PWM rectifier grid-connected current, U sd and U sq is the dq axis component of the three-phase modulation voltage of the PWM rectifier, ω is the synchronous rotation angular frequency of the power grid, and p is the differential operator;

[0077] Using the forward Euler method to discretize equation (3), we can get the predicted value I of the dq-axis component of the grid-connected current at time k+1: sd (k+1), I sq (k+1), is:

[0078]

[0079] Among them, U sd (k) and U sq (k) is the dq axis component of the three-phase modulation voltage of the PWM rectifier at the current moment, T s To control the cycle duration;

[0080] Through this step, the known information can be used to predict the grid-connected current value under different switch states (i.e. different PWM modulation voltages) at time k+1, providing accurate information for the cost function calculation and the DC bus voltage prediction value calculation at time k+1.

[0081] S32: Calculate the DC side current of the PWM rectifier at the current moment according to the DC bus voltage sampling value, the grid voltage dq axis component and the predicted value of the grid current dq axis component and based on the AC and DC side power balance relationship of the PWM rectifier;

[0082] Specifically, considering the power balance of the AC and DC sides of the PWM rectifier, the DC side current I of the PWM rectifier at the current moment is dc (k) is:

[0083]

[0084] Through this step, the DC side current of the PWM rectifier at the current moment can be calculated using known information, providing accurate information for calculating the predicted value of the DC bus voltage at the k+1 moment.

[0085] S33: According to the DC bus voltage sampling value and the DC side current, a mathematical model of the DC bus capacitor of the PWM rectifier is discretized using the forward Euler method to calculate a predicted value of the DC bus voltage of the PWM rectifier.

[0086] Specifically, according to the PWM DC side topology, its mathematical model is:

[0087]

[0088] Among them, V dc is the DC bus voltage, I dc is the DC side current, I L is the load current;

[0089] Using the forward Euler method to discretize equation (6), we can get the predicted value V of the DC bus voltage at time k+1: dc (k+1), is:

[0090]

[0091] Substituting equation (5) into equation (7) yields the predicted value of the DC bus voltage of the PWM rectifier:

[0092]

[0093] Through this step, the predicted value of the DC bus voltage at time k+1 can be calculated using known information, providing accurate information for cost function construction and calculation.

[0094] S4: According to the predicted value of the grid-connected current dq-axis component, the DC bus voltage sampling value and the predicted value of the DC bus voltage under the different switching states, an auxiliary controller is designed to obtain a reference value of the grid-connected current dq-axis component, and the square of the difference between the DC bus voltage reference value and the predicted value is selected as the DC bus voltage control item in the cost function, and the square of the difference between the grid-connected current dq-axis reference value and the predicted value is introduced as the grid-connected current constraint item in the cost function, and the weight coefficient of the grid-connected current constraint item is selected according to the predicted value of the grid-connected current dq-axis component, and the cost function of the quasi-cascadeless model predictive control is constructed; this step includes the following sub-steps:

[0095] S41: According to the DC bus voltage sampling value, select proportional integral control as the auxiliary controller, send the difference between the DC bus voltage reference value and the sampling value to the auxiliary controller, and take the output value of the auxiliary controller as the reference value of the d-axis component of the grid-connected current; considering the high power factor operation of the rectifier, take 0 as the reference value of the q-axis component of the grid-connected current;

[0096] Specifically, given a DC bus voltage reference value V dc_ref , calculate its difference with the DC bus voltage sampling value V dc The difference between the two values ​​is sent to the auxiliary controller based on proportional integral control, and the output of the auxiliary controller is used as the reference value I of the d-axis component of the grid-connected current. sd_ref ,Right now:

[0097]

[0098] Among them, k p , k i are the proportional and integral coefficients of the auxiliary controller;

[0099] In addition, the PWM rectifier is generally set to operate at a high power factor, so the reference value of the q-axis component of the grid-connected current is I sq_ref for:

[0100] I sq_ref =0 (10)

[0101] Through this step, the reference value of the dq axis component of the grid-connected current is obtained by using known information, providing appropriate reference information for the current constraint term in the cost function, thereby ensuring that the PWM rectifier operates at a suitable steady-state balance point, and there is no need to collect additional DC side load information, thereby reducing the amount of calculation;

[0102] S42: According to the DC bus voltage prediction value, select the square of the difference between the DC bus voltage reference value and the prediction value as the DC bus voltage control item in the cost function; according to the grid current dq axis component reference value and the prediction value, select the square of the difference between the grid current dq axis component reference value and the prediction value as the grid current constraint item in the cost function;

[0103] Specifically, select the DC bus voltage reference value V dc_ref and the predicted value V dc The square of the (k+1) difference is used as the DC bus voltage control term J in the cost function dc ,Right now:

[0104] J dc =|V dc_ref -V dc (k+1)| 2 (11)

[0105] Select the reference value I of the d-axis and q-axis components of the grid-connected current sd_ref ,I sq_ref and the predicted value I sd (k+1), I sq The square of the (k+1) difference is used as the grid-connected current constraint term J in the cost function Is ,Right now:

[0106] J Is =|I sd_ref -I sd (k+1)| 2 +|I sq_ref -I sq (k+1)| 2 (12)

[0107] Through this step, the DC bus voltage control term and the grid-connected current constraint term in the cost function are constructed to provide relevant calculation basis for the selection of the switch state;

[0108] S43: Calculate the square sum of the grid-connected current d-axis component prediction value and the grid-connected current q-axis component prediction value according to the grid-connected current dq-axis component prediction value; take the weight coefficient of the DC bus voltage control item as one, if the square sum is less than or equal to the square of the current threshold, then take the weight coefficient of the grid-connected current constraint item in the cost function to be two to three orders of magnitude smaller than the weight coefficient of the DC bus voltage control item; if the square sum is greater than the square of the current threshold, then take the weight coefficient of the grid-connected current constraint item in the cost function to be two to three orders of magnitude larger than the weight coefficient of the DC bus voltage control item;

[0109] Specifically, the weight coefficient of the DC bus voltage control term is taken as 1, and the predicted value of the d-axis component of the grid-connected current I is calculated. sd(k+1) and the predicted value of the q-axis component of the grid-connected current I sq The square sum of (k+1) and the current threshold I max The square of is compared, and the weight coefficient w of the grid-connected current constraint term in the cost function is selected according to the size relationship between the two. If the sum of the squares is less than or equal to the square of the current threshold, the weight coefficient of the grid-connected current constraint term in the cost function is taken as 0.001; if the sum of the squares is greater than the square of the current threshold, the weight coefficient of the grid-connected current constraint term in the cost function is taken as 100, that is:

[0110]

[0111] Through this step, the size of the grid-connected current constraint item is selected according to the size of the grid-connected current prediction value, so as to ensure that when the PWM rectifier operates at a normal steady-state equilibrium point, the DC bus voltage control item is dominant and responds to the dynamics of the DC bus voltage. When the grid-connected current prediction value is greater than the threshold, the grid-connected current constraint item is dominant and controls the grid-connected current to maintain a normal value. The weight coefficient of the grid-connected current constraint item can be flexibly adjusted to maximize the control capability of the DC bus voltage.

[0112] S44: According to the DC bus voltage control item, the grid current constraint item and the grid current constraint item weight coefficient, calculate the product of the grid current constraint item and the grid current constraint item weight coefficient, and set the sum of the DC bus voltage control item and the product as the cost function of the quasi-cascadeless model predictive control.

[0113] Specifically, the grid-connected current constraint term J is calculated as Is The product of the grid-connected current constraint term weight coefficient w sets the DC bus voltage control term J dc The sum of the products is the cost function J of the quasi-cascade-free model predictive control, that is:

[0114] J=|V dc_ref -V dc (k+1) 2 +w(I sd_ref -I sd (k+1) 2 +|I sq_ref -I sq (k+1) 2 ) (14)

[0115] Through this step, a complete cost function of quasi-cascadeless model predictive control is constructed to directly control the DC bus voltage, improve the DC bus voltage response capability, and constrain the grid-connected current to avoid the PWM rectifier running at an abnormal steady-state equilibrium point;

[0116] S5: According to the cost function, find the switch state that minimizes the cost function, and apply it to the PWM rectifier of the next control cycle.

[0117] Specifically, for the three-phase two-level PWM rectifier of the example, there are a total of 8 switching states; in each control cycle, the predicted values ​​of the dq-axis components of the grid-connected current and the predicted values ​​of the DC bus voltage under the 8 switching states are calculated, and then the cost functions of the 8 switching states are calculated; the switching state that minimizes the cost function is selected, and the opening or closing of each switch tube of the PWM rectifier at the next moment (k+1 moment) is determined according to the selected switching state, so as to realize the control of the PWM rectifier.

[0118] According to the example, the PWM rectifier control structure based on quasi-cascadeless model predictive control (QNC-MPC) is as follows: Figure 3 shown.

[0119] Figure 4(a) shows the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and quasi-cascadeless model predictive control (QNC-MPC) when the DC bus voltage reference value increases from 200V to 220V. Figure 4(b) shows the simulation results of the three-phase grid-connected current of the dual PI cascade when the DC bus voltage reference value increases from 200V to 220V. Figure 4(c) shows the simulation results of the three-phase grid-connected current of PI-MPC when the DC bus voltage reference value increases from 200V to 220V. Figure 4(d) shows the simulation results of the three-phase grid-connected current of QNC-MPC when the DC bus voltage reference value increases from 200V to 220V. Figure 4(a)-Figure 4(d) It can be seen that the three control methods have little difference in the DC bus voltage rising stage due to the limitation of the PWM rectifier current threshold. However, the DC bus voltage outer loop based on the PI controller of the dual PI cascade and PI-MPC leads to a longer adjustment time and a larger overshoot. Under the same PI control parameters, when the current exits saturation, QNC-MPC can quickly control the DC bus voltage to 220V, with better dynamic response capability.

[0120] Figure 5(a) shows the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and quasi-cascadeless model predictive control (QNC-MPC) when the DC bus voltage reference value drops from 220V to 200V. Figure 5(b) shows the simulation results of the three-phase grid-connected current of the dual PI cascade when the DC bus voltage reference value drops from 220V to 200V. Figure 5(c) shows the simulation results of the three-phase grid-connected current of PI-MPC when the DC bus voltage reference value drops from 220V to 200V. Figure 5(d) shows the simulation results of the three-phase grid-connected current of QNC-MPC when the DC bus voltage reference value drops from 220V to 200V. Figure 5(a)-Figure 5(d)It can be seen that both the dual PI cascade and PI-MPC control methods have long adjustment time and large overshoot, while QNC-MPC can greatly shorten the adjustment time and enhance the dynamic response capability of the PWM rectifier.

[0121] Figure 6(a) shows the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and quasi-cascadeless model predictive control (QNC-MPC) when a 200Ω load is suddenly added. Figure 6(b) shows the simulation results of the three-phase grid-connected current of the dual PI cascade when a 200Ω load is suddenly added. Figure 6(c) shows the simulation results of the three-phase grid-connected current of PI-MPC when a 200Ω load is suddenly added. Figure 6(d) shows the simulation results of the three-phase grid-connected current of QNC-MPC when a 200Ω load is suddenly added. Figure 6(a)-Figure 6(d) It can be seen that in the face of a sudden increase in load, the DC bus voltage under the dual PI cascade and PI-MPC control methods has large fluctuations, while the DC bus voltage under QNC-MPC has almost no fluctuations and has a strong anti-disturbance ability.

[0122] Figure 7(a) shows the simulation results of the DC bus voltage of the dual PI cascade, PI-MPC and quasi-cascadeless model predictive control (QNC-MPC) when the load suddenly drops by 200Ω. Figure 7(b) shows the simulation results of the three-phase grid-connected current of the dual PI cascade when the load suddenly drops by 200Ω. Figure 7(c) shows the simulation results of the three-phase grid-connected current of PI-MPC when the load suddenly drops by 200Ω. Figure 7(d) shows the simulation results of the three-phase grid-connected current of QNC-MPC when the load suddenly drops by 200Ω. Figure 7(a)-Figure 7(d) It can be seen that when the load suddenly decreases, the DC bus voltage of the dual PI cascade and PI-MPC will increase by about 1 V, and then slowly decrease to 200 V. Under QNC-MPC, the DC bus voltage has almost no change, and the anti-disturbance ability of the PWM rectifier is enhanced.

[0123] In summary, the present invention proposes a quasi-cascadeless model predictive control method for PWM rectifiers. The innovation of the method is that, within a control cycle, the three-phase grid voltage, three-phase grid-connected current and DC bus capacitor voltage at the current moment are collected; according to the discrete mathematical model of the DC bus voltage and the switch state output by the PWM rectifier, the DC bus voltage prediction value under different switch states is calculated; the cost function is selected as the square of the difference between the DC bus voltage reference value and the sampled value, and the grid current constraint term is introduced to ensure the correct steady-state operation point, and the grid current constraint term weight coefficient is adjusted according to the grid current prediction value; the active current reference value is obtained by the auxiliary PI controller after the DC bus voltage reference value and the DC bus voltage are made different, and the reactive current reference value can be set to 0; the cost function is optimized to find the switch state that minimizes the cost function, and it is applied to the PWM rectifier of the next control cycle. Based on the method of the present invention, the PWM rectifier DC bus voltage can be directly and effectively controlled without the traditional DC bus voltage-grid current double closed-loop cascade structure, and has better dynamic response capability and anti-disturbance capability. The present invention provides a method for improving the dynamic response and anti-disturbance capability of similar cascade control structures.

[0124] Corresponding to the aforementioned embodiment of the quasi-cascadeless model predictive control method for a PWM rectifier, the present application also provides an embodiment of a quasi-cascadeless model predictive control device for a PWM rectifier.

[0125] Figure 8 1 is a block diagram of a quasi-cascadeless model predictive control device for a PWM rectifier according to an exemplary embodiment. Figure 8 , the device comprises:

[0126] Sampling module 1 is used to collect the three-phase grid voltage sampling value, the three-phase grid current sampling value, the DC bus voltage sampling value and the load current sampling value at the current moment within a control cycle;

[0127] A coordinate conversion module 2 is used to perform coordinate conversion according to the three-phase grid voltage sampling value and the three-phase grid current sampling value and the grid angle obtained by the phase-locked loop to obtain the dq-axis components of the grid voltage and the grid current in the rotating coordinate system;

[0128] A calculation module 3 is used to substitute the DC bus voltage sampling value, the load current sampling value, and the dq axis components of the grid voltage and the grid current into the discrete mathematical model of the PWM rectifier to calculate the predicted value of the dq axis component of the grid current and the predicted value of the DC bus voltage under different switching states;

[0129] Construction module 4, used for designing an auxiliary controller to obtain a reference value of the dq-axis component of the grid-connected current according to the predicted value of the dq-axis component of the grid-connected current, the sampled value of the DC bus voltage and the predicted value of the DC bus voltage under the different switching states, selecting the square of the difference between the reference value of the DC bus voltage and the predicted value as the DC bus voltage control term in the cost function, and introducing the square of the difference between the dq-axis reference value of the grid-connected current and the predicted value as the grid-connected current constraint term in the cost function, selecting the weight coefficient of the grid-connected current constraint term according to the predicted value of the dq-axis component of the grid-connected current, and constructing the cost function of the quasi-cascadeless model predictive control;

[0130] The control module 5 is used to find out the switch state that minimizes the cost function according to the cost function, and apply the switch state to the PWM rectifier in the next control cycle.

[0131] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0132] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0133] Correspondingly, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the quasi-cascadeless model predictive control method for PWM rectifier as described above.

[0134] Accordingly, the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by a processor, the quasi-cascadeless model predictive control method for a PWM rectifier as described above is implemented.

[0135] Those skilled in the art will readily appreciate other embodiments of the present application after considering the description and practicing the contents disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The description and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the claims.

[0136] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A quasi-cascadeless model predictive control method for a PWM rectifier, characterized in that: include: In one control cycle, the three-phase grid voltage sampling value, the three-phase grid current sampling value, the DC bus voltage sampling value and the load current sampling value are collected at the current moment; According to the three-phase grid voltage sampling value and the three-phase grid-connected current sampling value, coordinate transformation is performed according to the grid angle obtained by the phase-locked loop to obtain the dq-axis components of the grid voltage and the grid-connected current in the rotating coordinate system; Substituting the DC bus voltage sampling value, the load current sampling value, and the dq-axis components of the grid voltage and the grid-connected current into the PWM rectifier discrete mathematical model, the predicted value of the dq-axis component of the grid-connected current and the predicted value of the DC bus voltage under different switching states are calculated; According to the predicted value of the dq-axis component of the grid-connected current, the sampled value of the DC bus voltage and the predicted value of the DC bus voltage under the different switching states, an auxiliary controller is designed to obtain a reference value of the dq-axis component of the grid-connected current, the square of the difference between the reference value and the predicted value of the DC bus voltage is selected as the DC bus voltage control item in the cost function, and the square of the difference between the reference value and the predicted value of the dq-axis component of the grid-connected current is introduced as the grid-connected current constraint item in the cost function, and the weight coefficient of the grid-connected current constraint item is selected according to the predicted value of the dq-axis component of the grid-connected current, so as to construct a cost function of the quasi-cascadeless model predictive control; According to the cost function, a switch state that minimizes the cost function is found and applied to the PWM rectifier in the next control cycle.

2. A quasi-cascadeless model predictive control method for a PWM rectifier according to claim 1, characterized in that: Substituting the DC bus voltage sampling value, the load current sampling value, and the dq axis components of the grid voltage and the grid-connected current into the PWM rectifier discrete mathematical model, the predicted value of the dq axis component of the grid-connected current and the predicted value of the DC bus voltage under different switching states are calculated, including: According to the dq-axis components of the grid voltage and the grid-connected current, the mathematical model of the AC side of the PWM rectifier in the rotating coordinate system is discretized using the forward Euler method, and the dq-axis components of the grid voltage and the grid-connected current are substituted into the mathematical model of the AC side of the PWM rectifier to obtain the predicted value of the dq-axis component of the grid-connected current of the PWM rectifier; According to the DC bus voltage sampling value, the predicted value of the grid voltage dq axis component and the grid current dq axis component, based on the AC and DC side power balance relationship of the PWM rectifier, the DC side current of the PWM rectifier at the current moment is calculated; According to the DC bus voltage sampling value and the DC side current, the mathematical model of the DC bus capacitor of the PWM rectifier is discretized by using the forward Euler method to calculate the predicted value of the DC bus voltage of the PWM rectifier.

3. The quasi-cascadeless model predictive control method for a PWM rectifier according to claim 1, characterized in that: According to the predicted values ​​of the grid-connected current dq-axis components, the DC bus voltage sampling values, and the predicted values ​​of the DC bus voltage under the different switching states, an auxiliary controller is designed to obtain a reference value of the grid-connected current dq-axis components, the square of the difference between the DC bus voltage reference value and the predicted value is selected as the DC bus voltage control item in the cost function, and the square of the difference between the grid-connected current dq-axis components and the predicted value is introduced as the grid-connected current constraint item in the cost function, and the weight coefficient of the grid-connected current constraint item is selected according to the predicted value of the grid-connected current dq-axis components, and the cost function of the quasi-cascade-free model predictive control is constructed, including: According to the DC bus voltage sampling value, proportional integral control is selected as the auxiliary controller, the difference between the DC bus voltage reference value and the sampling value is sent to the auxiliary controller, and the output value of the auxiliary controller is taken as the reference value of the d-axis component of the grid-connected current; considering the high power factor operation of the rectifier, 0 is taken as the reference value of the q-axis component of the grid-connected current; According to the DC bus voltage prediction value, the square of the difference between the DC bus voltage reference value and the prediction value is selected as the DC bus voltage control item in the cost function; according to the grid current dq axis component reference value and the prediction value, the square of the difference between the grid current dq axis component reference value and the prediction value is selected as the grid current constraint item in the cost function; According to the predicted value of the grid-connected current dq-axis component, the square sum of the predicted value of the grid-connected current d-axis component and the predicted value of the grid-connected current q-axis component is calculated; the weight coefficient of the DC bus voltage control item is taken as one, and if the square sum is less than or equal to the square of the current threshold, the weight coefficient of the grid-connected current constraint item in the cost function is taken to be two to three orders of magnitude smaller than the weight coefficient of the DC bus voltage control item; if the square sum is greater than the square of the current threshold, the weight coefficient of the grid-connected current constraint item in the cost function is taken to be two to three orders of magnitude larger than the weight coefficient of the DC bus voltage control item; According to the DC bus voltage control item, the grid current constraint item and the grid current constraint item weight coefficient, the product of the grid current constraint item and the grid current constraint item weight coefficient is calculated, and the sum of the DC bus voltage control item and the product is set as the cost function of the quasi-cascadeless model predictive control.

4. A quasi-cascadeless model predictive control device for a PWM rectifier, characterized in that: include: The sampling module is used to collect the three-phase grid voltage sampling value, the three-phase grid current sampling value, the DC bus voltage sampling value and the load current sampling value at the current moment within a control cycle; A coordinate conversion module is used to perform coordinate conversion according to the three-phase grid voltage sampling value and the three-phase grid-connected current sampling value and the grid angle obtained by the phase-locked loop to obtain the dq-axis components of the grid voltage and the grid-connected current in the rotating coordinate system; A calculation module, used to substitute the DC bus voltage sampling value, the load current sampling value, and the dq-axis components of the grid voltage and the grid-connected current into a discrete mathematical model of the PWM rectifier, and calculate the predicted value of the dq-axis component of the grid-connected current and the predicted value of the DC bus voltage under different switching states; A construction module, used to design an auxiliary controller to obtain a reference value of a dq-axis component of a grid-connected current according to the predicted value of a dq-axis component of a grid-connected current, a sampled value of a DC bus voltage, and a predicted value of a DC bus voltage under the different switching states, select the square of the difference between the reference value of the DC bus voltage and the predicted value as a DC bus voltage control item in a cost function, introduce the square of the difference between the reference value of the dq-axis component of the grid-connected current and the predicted value as a grid-connected current constraint item in the cost function, select a weight coefficient of the grid-connected current constraint item according to the predicted value of the dq-axis component of the grid-connected current, and construct a cost function of a quasi-cascadeless model predictive control; The control module is used to find out the switch state that minimizes the cost function according to the cost function, and apply the switch state to the PWM rectifier of the next control cycle.

5. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 3.

6. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instruction is executed by a processor, the steps of the method according to any one of claims 1 to 3 are implemented.

Images