DC-DC converter control method, electronic device, computer-readable medium, and computer program product

Abstract

The present invention relates to a control method, electronic device, computer-readable medium, and computer program product for a DC-DC converter, and belongs to the field of automatic control of DC-DC conversion. The specific control method includes: first establishing an affine nonlinear model containing parasitic parameters of circuit components; constructing a Brunovsky standard form of the model based on a target holographic feedback method, and converting the system into a linear form; establishing a linear prediction model based on the model predictive control principle, and constructing a dynamic semi-elliptical interval control curve. y H and y L The objective function is constrained. Finally, a sequential quadratic programming algorithm is used to solve the objective function and generate a sequence of control variables, U, to dynamically adjust the duty cycle of the switching transistors. This method is primarily used to suppress the cross-effects between branches caused by parasitic parameters and achieve high-precision coordinated control of multiple output voltages.

Description

Technical Field

[0001] The present invention relates to the field of DC-DC conversion automatic control. More specifically, the present invention relates to a control method for a DC-DC converter, an electronic device, a computer-readable medium, and a computer program product. Background Art

[0002] With the increasing popularity of portable electronic devices, multi-core processors, and smart home systems, performance requirements for power conversion devices are increasing in terms of cost-effectiveness, compactness, and operational stability. Against this backdrop, single-inductor, multiple-output (SIMO) DC-DC converters have gained widespread application in various electronic systems due to their technical advantages, such as their ability to significantly reduce system size and component count while supporting multiple independent outputs. However, these converters are limited by their single-inductor shared architecture, which inevitably results in crosstalk between channels when driving multiple loads. This degrades the dynamic response of the output voltage and severely limits overall system performance.

[0003] To suppress cross-effects, the industry has proposed a variety of control strategies. In existing technologies, closed-loop control schemes based on current ripple or average current can mitigate channel coupling effects by optimizing inductor current distribution, but such methods typically result in increased output voltage ripple. Other studies have used switching linear hybrid control laws or single-discharge timing control to improve transient response speed, but their hardware implementation complexity is high and their adaptability to circuit environments is limited. Furthermore, while adaptive control techniques based on precise feedback linearization exhibit good dynamic performance in specific topologies, they rely on strict system model conditions and require precise matching of the output function in practical applications. Especially for non-ideal converter models containing parasitic parameters, component parameter drift will significantly exacerbate cross-effects, making it difficult to meet the design conditions of traditional linearization methods.

[0004] In recent years, the target holographic feedback (OHF) linearization method has attracted attention due to its more flexible engineering applicability. This method transforms nonlinear systems into a standard linear form by constructing a set of target tracking equations with multiple state variables. Global linearization is achieved by simply satisfying the relative order of 1, effectively circumventing the stringent model accuracy requirements of traditional methods. Currently, the OHF method has been applied in DC-DC converters such as buck converters, boost converters, and Cuk converters. Model predictive control (MPC) is a robust and forward-looking control strategy. Leveraging its core advantages of model-driven, rolling optimization, and constraint handling, it has been widely used in the automotive, industrial, and energy sectors. Summary of the Invention

[0005] The object of the present invention is to provide a control method, electronic device, computer-readable medium and computer program product of a DC-DC converter, so as to suppress the cross-effect of non-ideal DC-DC converters.

[0006] In order to achieve the purpose and other advantages of the present invention, a control method of a DC-DC converter is provided, comprising:

[0007] S1. Establishing an affine nonlinear model of a non-ideal SIDO buck converter, wherein the affine nonlinear model includes parasitic parameters of circuit components;

[0008] S2. Based on the target holographic feedback method, constructing the Brunovsky standard form of the affine nonlinear model and transforming the nonlinear system into a linear system;

[0009] S3. Based on the model predictive control principle, a prediction model of the linear system is established, and an objective function is constructed with a dynamic semi-elliptical interval control curve as a constraint condition;

[0010] S4. Solve the objective function using a sequential quadratic programming algorithm to generate a control variable sequence U, and dynamically adjust the duty cycle of the switch tube according to U to achieve coordinated control of the branch output voltage;

[0011] Among them, the dynamic semi-elliptical interval control curve y H and y L ,satisfy:

[0012] ,

[0013] Where, k Indicates the current moment; i =0, 1, 2, …, N P , N p is the prediction step length; y H-H and y L-L are the upper and lower boundaries of the curve interval, y max and y min are the upper and lower boundaries of the tolerance interval, y H-H = (1 + η ) · y max , y L-L = (1- η )· y min , η is a constant and 0 < η < 1.

[0014] Preferably, in S3,

[0015] The objective function J is: ,in:

[0016] ,

[0017] ,

[0018] Where, e y ( k+i )for k+i Time output function y The predicted value of 1 exceeds the size of the curve interval; y 1( k+i )and y 2( k+i ) are respectively k+i Output variables at any moment y 1 and y 2, y1 is the inductor current I L , y2 is the output voltage vb of branch b, y 2ref for y The reference value of 2, y 2ref = V bref , V bref is the branch output voltage v b The reference value, is a diagonal matrix, r w is the weighting coefficient, I Nc×Nc is the identity matrix with Nc rows and Nc columns, Nc is the control step size; and when y 1( k + i ) is outside the curve range, e y ( k + i )for y 1( k + i ) to the boundary of the curve interval; when y 1( k + i ) is within the curve interval, e y ( k + i ) is 0; T represents the transpose of the matrix; U is the control variable sequence.

[0019] Preferably, in S3, the tolerance interval boundary y max and y min Set to: , where the initial value of the tolerance interval boundary is y max (0) =(1+ ε )· y 1ref , y min (0) = (1- ε )· y 1ref , ε is the interval width coefficient, and 0 < ε < 0.1, y 1ref for y A reference value of 1, , V aref is the branch output voltage v a The reference value, R a and R b is the branch load resistance;

[0020] η ( x )satisfy: ,in, α is the compensation coefficient, and α > 0.

[0021] Preferably, in S1, the affine nonlinear model is:

[0022] ,

[0023] in, ,

[0024] ,

[0025] ,

[0026] .

[0027] Where, for x The first derivative of , y is the output function, V in is the input voltage, L is the filter inductor, C a and Cb is the branch output filter capacitor, R a and R b is the branch load resistance, v a and v b is the branch output voltage, V aref and V bref They are v a and v b The reference value, Q 1. Q a and Q b They are all switching tubes. D 1 is a freewheeling diode, i L is the inductor current, I Lref for i L The reference value, R L for L The equivalent series resistance, R 1on 、 R aon 、 R bon They are Q 1. Q a 、 Q b The on-resistance, V D1 、 R D1 They are D 1 forward voltage drop and forward resistance, R Ca 、 R Cb They are C a and C b The equivalent series resistance, d 1. d a and d b Switching tube Q 1. Q a and Q bThe duty cycle of d a + d b = 1, the state variable is x = [ x 1, x 2, x 3] T = [ i L , v Ca , v Cb ] T , T represents the transpose of the matrix, v Ca and v Cb They are C a and C b The voltage across the terminals, i a and i b Both are branch output currents.

[0028] Preferably, in S2, the Brunovsky standard form of the affine nonlinear model is ,

[0029] in, for The first derivative of ,Pick y m = [ y m1 , y m2 ] T = [ 1, 3] T , , , , ,

[0030] ,

[0031] Let duty cycle:

[0032] , , ,

[0033] but .

[0034] Preferably, the sampling time is T s , discretizing the Brunovsky standard form of the affine nonlinear model, we can obtain:

[0035] ,

[0036] in, ;

[0037] Its controllable matrix is , rank ( M ) = 3;

[0038] In S3, the prediction model is: , U is the control variable sequence, Φ is the state variable sequence, Y is the output variable sequence, E 1. E 2. F 1 and F 2 is a matrix; where

[0039] ,

[0040] ,

[0041] .

[0042] Preferably, in S4, the duty cycle d is:

[0043] .

[0044] The present invention also provides an electronic device comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor performs the above-mentioned control method.

[0045] The present invention also provides a computer-readable medium having a computer program stored thereon, which implements the above-mentioned control method when executed by a processor.

[0046] The present invention also provides a computer program product, comprising a computer program / instruction, which implements the above control method when executed by a processor.

[0047] The present invention has at least the following beneficial effects:

[0048] This invention proposes a control method for a DC-DC converter. This method first considers the parasitic parameters of circuit components and establishes an affine nonlinear model of the converter. Then, based on the target holographic feedback method, a Brunovsky standard form that conforms to the converter model is constructed. Furthermore, based on model predictive control theory, a prediction equation for the system is established, and an appropriate objective function is selected. Finally, a semi-elliptical interval curve with a variable tolerance interval is designed in conjunction with an interval control method to control the output, effectively suppressing output errors caused by parasitic parameters. Simulation results show that compared with common-mode-differential-mode control methods, this control method has less crosstalk and better dynamic performance.

[0049] Other advantages, objectives and features of the present invention will be reflected in part from the following description and will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a non-ideal SIDO Buck converter circuit topology;

[0051] Figure 2 This is a conventional interval control diagram;

[0052] Figure 3 It is a schematic diagram of the dynamic semi-elliptical interval control of the present invention;

[0053] Figure 4 This is a control block diagram of the improved interval model predictive control based on target holographic feedback (OHF-IZMPC) proposed in the present invention;

[0054] Figure 5 yes R a Simulation results when changing;

[0055] Figure 6 yes R b Simulation results when changing;

[0056] Figure 7 yes V in Simulation results when . DETAILED DESCRIPTION

[0057] The present invention will be further described in detail below with reference to the embodiments and drawings so that those skilled in the art can implement the invention with reference to the description.

[0058] It should be understood that terms such as “having”, “including” and “comprising” used herein do not preclude the existence or addition of one or more other elements or combinations thereof.

[0059] It should be noted that the experimental methods described in the following embodiments are conventional methods unless otherwise specified, and the reagents and materials can be obtained from commercial channels unless otherwise specified.

[0060] Example 1:

[0061] 1. Converter topology and modeling

[0062] Figure 1 The circuit topology of the non-ideal SIDO Buck converter is given. V in is the input voltage, L is the filter inductor, C a and C b They are all branch output filter capacitors. R a and R b are all branch load resistances, v a and v b are all branch output voltages, i a and i b are all branch output currents, Q 1. Q a and Q b All are MOSFET tubes. D 1 is a freewheeling diode, i L is the inductor current, R L for L The equivalent series resistance, R 1on 、 R aon 、 R bon They are Q 1. Q a 、 Q b The on-resistance, V D1 、 R D1 They are D 1 forward voltage drop and forward resistance, R Ca 、 R Cb They are C a andC b The equivalent series resistance, v Ca and v Cb for C a and C b Voltage across both ends.

[0063] Take the state variable as x = [ x 1, x 2, x 3] T = [ i L , v Ca , v Cb ] T , T represents the transpose of the matrix, d 1. d a and d b Switching tube Q 1. Q a and Q b The duty cycle of d a + d b = 1. According to the circuit topology, the affine nonlinear model of the continuous conduction mode single-inductor dual-output Buck converter (CCM SIDO Buck converter) can be established as:

[0064] (1)

[0065] in, f ( x ), g 1( x ), g 2( x ), h ( x ) and are:

[0066] (2)

[0067] (3)

[0068] (4)

[0069] (5)

[0070] in, I Lref 、 V aref and V bref They are x 1. x 2 and x The reference value is 3.

[0071] It is not difficult to see that the system is a strongly coupled nonlinear system.

[0072] 2. Target holographic feedback design

[0073] Based on the target holographic feedback method, the following Brunovsky standard form of the affine nonlinear model can be constructed:

[0074] (6)

[0075] in, for The first derivative of ,Pick y m = [ y m1 , y m2 ] T = [ 1, 3] T , , , , ,

[0076] (7)

[0077] Let duty cycle:

[0078] , , ,

[0079] Then formula (7) can be expressed as: (8)

[0080] Through formula (8), the nonlinear system can be transformed into a linear system.

[0081] The sampling time is T s , discretizing the Brunovsky standard form of the affine nonlinear model, we can obtain:

[0082] ,

[0083] in, ;

[0084] Its controllable matrix is , rank ( M ) = 3. Therefore, the discretized system is completely controllable.

[0085] 3. Controller design

[0086] 3.1 Design of Model Predictive Controller

[0087] Define the future control variable sequence separately U , the future state variable sequence Φ and future output variable sequences Y for:

[0088] (9)

[0089] Based on the model predictive control principle, the prediction model of the system can be established as follows:

[0090] (10)

[0091] Among them, the matrix E 1. E 2. F 1 and F 2 are:

[0092] (11)

[0093] (12)

[0094] Next, we select the objective function J as:

[0095] (13)

[0096] in, R s = [ y 1ref y 2ref y 1ref y 2ref … y 1ref y 2ref ] T is 2 N P elements of the reference output sequence, y 1refand y 2ref They are y 1 and y 2 reference value, and y 1ref = I Lref , y 2ref = V bref , is a diagonal matrix, r w is the weighting coefficient.

[0097] 3.2 Design of Improved Interval Model Predictive Controller

[0098] The present invention outputs variables y 2 Still using set value control, introducing interval control to output variables y 1 to control and allow y 1 changes within a certain range in order to achieve precise control of the output voltage and improve the robustness of the control system.

[0099] The conventional interval control diagram is as follows Figure 2 As shown, y max and y min are the upper and lower boundaries of the tolerance interval respectively. The larger the interval range, the stronger the robustness and the lower the control accuracy. Such intervals cannot take into account the performance indicators such as robustness and control accuracy of the control system. To solve this problem, the present invention proposes a semi-elliptical interval. The interval diagram is as follows: Figure 3 As shown. Among them, y H-H and y L-L are the upper and lower boundaries of the curve interval, y H and y L They are the newly added dynamic semi-elliptical interval control curves, y max and y min are the upper and lower boundaries of the tolerance interval, e y ( k )for k The predicted value at the moment exceeds the size of the curve interval, and y H-H =(1 + η )· y max , y L-L= (1 - η )· y min , η is a constant and 0 < η < 1.

[0100] When the controlled quantity is disturbed, it is first controlled to be within the curve range, and then further controlled to be within the tolerance range. N P The adjustment is completed in one moment. y H and y L The design concept is as follows: when the control system is disturbed and the predicted output is higher than the tolerance interval, it is expected that the predicted output will initially approach the tolerance interval at a faster speed to reduce actual energy loss; and when the control system is disturbed and the predicted output is lower than the tolerance interval, the predicted output is allowed to initially approach the tolerance interval at a slower speed to reduce waveform jitter. Based on this, combined with practical considerations, the curve y H and y L Set them as the lower left and lower right curves of the ellipse respectively.

[0101] y H and y L The expressions are:

[0102] (14)

[0103] in i = 1, 2, …, N P From this we can get the prediction error e y ( k + i ) expression:

[0104] (15)

[0105] when y 1( k + i ) is outside the curve range, e y ( k + i )for y 1( k + i ) to the interval boundary; when y 1( k +i ) is within the curve interval, e y ( k + i ) is 0.

[0106] Now assume Figure 2 and Figure 3 The corresponding systems are system (a) and (b), and the systems (a) and (b) are y 1ref 、 y max and y min are the same, that is, the control accuracy is the same. Assume k The system is affected by disturbances at the moment. y 1( k ) can be roughly divided into three situations depending on the location: (1). y min < y 1( k )< y max ,at this time y 1( k ) is still within the tolerance range, and the system does not make any additional adjustments; (2). y L-L < y 1( k )< y min or y max < y 1( k )< y H-H , at this time y 1( k ) exceeds the tolerance range, the system (a) needs to be adjusted significantly, which may cause jitter. y 1( k ) does not exceed Figure 3 The control interval boundary of system (b) only requires N P The output is adjusted to within the tolerance range at each moment, with a small adjustment range, which can effectively suppress the jitter phenomenon; (3). y 1( k )< y L-L or y H-H < y 1( k ),use ea y ( k ) represents the system (a)y 1( k ) is beyond the tolerance range, e y ( k ) represents the system (b) y 1( k ) exceeds the control interval due to ea y ( k ) > e y ( k ), which means that the adjustment action of system (a) will be greater, and the system may produce greater jitter. Therefore, compared with the general interval setting, the curve interval designed by the present invention has less jitter and better robustness of the system while maintaining the same control accuracy.

[0107] Next, you need to select the appropriate y 1ref 、 y max and y min , in order to achieve precise control of the output voltage.

[0108] Select y 1ref as follows:

[0109] (16)

[0110] For non-ideal SIDO Buck converter, if y 1 tracking y 1ref and y 2Tracking y 2ref , v a and V aref There will be some errors.

[0111] To eliminate v a and V aref The error between y max and y min as follows:

[0112] (17)

[0113] in, y max (0) = (1+ ε)· y 1ref , y min (0) = (1- ε )· y 1ref , ε is the interval width coefficient, and 0 < ε < 0.1, η(x) Satisfy the following formula:

[0114] (18)

[0115] in, α is the compensation coefficient, and α > 0.

[0116] From equations (17) and (18), we can see that the tolerance interval boundary designed by the present invention has an adaptive adjustment mechanism, and the initial boundary of the tolerance interval is y 1ref is the axis of symmetry, where | e y | + | y 2 – y 2ref | = 0 means the control system is in a stable state, | e y | + | y 2 – y 2ref | ≠ 0 means that the control system is in the adjustment state and has not yet reached a stable state.

[0117] The mechanism of action of formula (17) is as follows:

[0118] (1) When |e y | + | y 2 – y 2ref | ≠ 0, the final value cannot be determined because the system adjustment is not yet completed. v a and V aref Is there an error between , and the tolerance interval does not change, that is y max ( k +1) = y max ( k ), y min ( k +1) = y min ( k );

[0119] (2) When | e y | + | y 2 – y 2ref When | = 0, the system is in a stable state. v a = V aref , y max ( k +1) = y max ( k ), y min ( k +1) = y min ( k ), the tolerance interval does not change; if v a > V aref , then y max ( k +1) < y max ( k ), y min ( k +1) < y min ( k ), y 1Tolerance interval is adjusted downward, expectation y 1 will decrease in the future. From formula (16), we can see that y 2 unchanged and y When 1 decreases, v a will be reduced, achieving v a track V aref similarly, if v a < V aref , then y max ( k +1) > y max ( k ), y min ( k +1) > y min (k ),expect y 1Increase and realize in the future v a track V aref goal.

[0120] It can be seen that the mechanism of formula (17) can be realized v a The goal is to track its reference value. So far, the designed controller has achieved stable control of the two output voltages of the converter.

[0121] Based on the above analysis, we can conclude that:

[0122] (19)

[0123] The objective function (19) is solved using the Sequential Quadratic Programming (SQP) algorithm to obtain the system’s required U .

[0124] At this point, the duty cycle of the system can be obtained d as follows:

[0125] (20)

[0126] In summary, the OHF-IZMPC control block diagram proposed by the present invention is as follows: Figure 4 shown.

[0127] Example 2:

[0128] To verify the feasibility and superiority of the control method proposed in this paper, a simulation model of a non-ideal SIDO Buck converter based on OHF-IZMPC control is established in MATLAB / Simulink, and a comparative analysis is performed with the common-mode voltage-differential-mode voltage (CMV-DMV) control method. The specific parameter values ​​are shown in Table 1, where V aref and V bref The output voltage v a and v b Reference value, switching frequency f s = 1 / T s .

[0129] Table 1 Main parameters

[0130]

[0131] Figure 5 Given R a The simulation results when the Figure 5 (a) is the CMV-DMV control method, Figure 5 (b) is the OHF-IZMPC control method of the present invention. R a When the resistance changes from 20 Ω to 15 Ω in 30 ms, under CMV-DMV control, v a and v b The maximum fluctuation is 0.31 V, and the adjustment time is 8 ms, that is, the cross-effect of output branch a on b is 0.31 V; under OHF-IZMPC control, v a The maximum fluctuation is 3 mV. v b The maximum fluctuation is 15 mV, that is, the cross-influence of output branch a on b is only 15 mV. R a When the resistance changes from 15Ω to 20Ω in 50ms, under CMV-DMV control, v a and v b The maximum fluctuations are 0.36 V and 0.29 V, respectively, and the adjustment time is 8 ms and 7 ms, respectively. That is, the cross-effect of output branch a on b is 0.29 V. Under OHF-IZMPC control, v a The maximum fluctuation is 4 mV. v b The maximum fluctuation is 15 mV, that is, the cross-effect of output branch a on branch b is only 15 mV. Therefore, OHF-IZMPC effectively reduces the cross-effect of branch a on branch b.

[0132] Figure 6 Given R b The simulation results when the Figure 6 (a) is the CMV-DMV control method, Figure 6 (b) is the OHF-IZMPC control method of the present invention. R b When the resistance changes from 20 Ω to 15 Ω in 70 ms, under CMV-DMV control, v a and v bThe maximum fluctuations are 0.38 V and 0.32 V respectively, and the adjustment time is 7 ms. That is, the cross-effect of output branch a on b is 0.38 V. Under OHF-IZMPC control, v a The maximum fluctuation is 14 mV. v b The maximum fluctuation is 20 mV, that is, the cross-influence of output branch a on b is only 14 mV. R b When the resistance changes from 15Ω to 20Ω in 90ms, under CMV-DMV control, v a and v b The maximum fluctuations of are 0.26 V and 0.28 V respectively, and the adjustment time is 8 ms. That is, the cross-effect of output branch a on b is 0.26 V. Under OHF-IZMPC control, v a The maximum fluctuation is 15 mV. v b The maximum fluctuation of is 20 mV, that is, the cross-influence of output branch a on b is only 15 mV. Therefore, OHF-IZMPC effectively suppresses the output voltage fluctuation and cross-influence of the system.

[0133] Figure 7 Given V in The simulation results when the Figure 7 (a) is the CMV-DMV control method, Figure 7 (b) is the OHF-IZMPC control method of the present invention. V in When the voltage jumps from 20 V to 18 V in 110 ms, under CMV-DMV control, v a and v b The maximum fluctuations of are 0.26 V and 0.24 V, respectively, and the adjustment time is 5 ms. Under OHF-IZMPC control, v a The maximum fluctuation is 8mV. v b The maximum fluctuation is 6 mV. V in When the voltage jumps from 18 V to 20 V in 130 ms, v a and v b The maximum fluctuations of are 0.36 V and 0.32 V, respectively, and the adjustment time is 4 ms. Under OHF-IZMPC control,v a The maximum fluctuation is 7 mV. v b The maximum fluctuation is 5 mV. Therefore, OHF-IZMPC can effectively suppress input voltage disturbances.

[0134] The number of devices and processing scales described herein are intended to simplify the description of the present invention. Applications, modifications, and variations of the control method, electronic device, computer-readable medium, and computer program product of the present invention will be readily apparent to those skilled in the art.

[0135] Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the description and implementation methods. They can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. A control method for a DC-DC converter, characterized in that: include: S1. Establishing an affine nonlinear model of a non-ideal SIDO buck converter, wherein the affine nonlinear model includes parasitic parameters of circuit components; S2. Based on the target holographic feedback method, constructing the Brunovsky standard form of the affine nonlinear model and transforming the nonlinear system into a linear system; S3. Based on the model predictive control principle, a prediction model of the linear system is established, and an objective function is constructed with a dynamic semi-elliptical interval control curve as a constraint condition; S4. Solve the objective function using a sequential quadratic programming algorithm to generate a control variable sequence U, and dynamically adjust the duty cycle of the switch tube according to U to achieve coordinated control of the branch output voltage; Among them, the dynamic semi-elliptical interval control curve y H and y L ,satisfy: Where k represents the current time; i = 0, 1, 2, ..., N P , N p is the prediction step length; y H-H and y L-L are the upper and lower boundaries of the curve interval, y max and y min are the upper and lower boundaries of the tolerance interval, y H-H =(1+η)·y max ,y L-L =(1-η)·y min , η is a constant and 0<η<1; In S3, the objective function J is: in, Where, e y (k+i) is the extent to which the predicted value of the output function y1 at time k+i exceeds the curve interval; y1(k+i) and y2(k+i) are the magnitudes of the output variables y1 and y2 at time k+i, respectively, and y1 is the inductor current I L , y2 is the output voltage vb of branch b, y 2ref is the reference value of y2, y 2ref =V bref , V bref is the branch output voltage v b The reference value, is a diagonal matrix, r w is the weighting coefficient, I Nc×Nc is the identity matrix with Nc rows and Nc columns, Nc is the control step size; and when y1(k+i) is outside the curve interval, e y (k+i) is the distance from y1(k+i) to the boundary of the curve interval; when y1(k+i) is within the curve interval, e y (k+i) is 0; T represents the transpose of the matrix; U is the control variable sequence; Tolerance interval boundary y max and y min Set to: Among them, the initial value y of the tolerance interval boundary max (0) = (1 + ε)·y 1ref ,y min (0) = (1 - ε) · y 1ref , ε is the interval width coefficient, and 0<ε<0.1, y 1ref is the reference value of y1, y 1ref =V aref / R a +V bref / R b , V aref is the branch output voltage v a Reference value, R a and R b is the branch load resistance; η(x) satisfies: Wherein, α is the compensation coefficient, and α>

0.

2. The control method of the DC-DC converter according to claim 1, wherein: In S1, the affine nonlinear model is: in, h(x)=[h1(x) h2(x) h3(x)] T =[x1-I Lref x2-V aref x3-V bref ] T , Where, is the first-order derivative of x, y is the output function, V in is the input voltage, L is the filter inductor, C a and C b is the branch output filter capacitor, R a and R b is the branch load resistance, v a and v b is the branch output voltage, V aref and V bref v a and v b Reference values ​​of Q1, Q a and Q b They are all switch tubes, D1 is a freewheeling diode, i L is the inductor current, I Lref for i L Reference value, R L is the equivalent series resistance of L, R 1on 、R aon 、R bon Q1, Q a , Q b On-resistance, V D1 、R D1 are the forward voltage drop and forward resistance of D1, R Ca 、R Cb C a and C b The equivalent series resistance, d1, d a and d b They are switch tubes Q1 and Q a and Q b The duty cycle, and d a +d b =1, the state variable is x=[x1, x2, x3] T =[i L , v Ca , v Cb ] T , T represents the transpose of the matrix, v Ca and v Cb C a and C b Voltage across both ends, i a and i b Both are branch output currents.

3. The control method of the DC-DC converter according to claim 2, wherein: In S2, the Brunovsky standard form of the affine nonlinear model is in, for The first derivative of Pick Duty cycle Then d = β -1 (x)[u - α(x)].

4. The control method of the DC-DC converter according to claim 3, wherein: Take the sampling time as T s , discretizing the Brunovsky standard form of the affine nonlinear model, we can obtain: in, Its controllable matrix is rank(M)=3; In S3, the prediction model is: U is the control variable sequence, Φ is the state variable sequence, Y is the output variable sequence, E1, E2, F1 and F2 are matrices; where 5. The control method of the DC-DC converter according to claim 4, wherein: In S4, the duty cycle d is:

6. An electronic device, characterized in that include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to cause the at least one processor to perform the control method according to any one of claims 1 to 5.

7. A computer-readable medium having a computer program stored thereon, characterized in that When the program is executed by a processor, the control method according to any one of claims 1 to 5 is implemented.

8. Computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the control method according to any one of claims 1 to 5 is implemented.

Images