CN103117657B - Control method of full-bridge DC-DC system based on on-chip model predictive control - Google Patents

Abstract

The invention discloses a control method of a full-bridge DC-DC system based on on-chip model predictive control. The control method includes following steps: (1) dividing an object state of the non-linear full-bridge DC-DC system into different polyhedral areas, and obtaining a best control law according to the control laws in each polyhedral area; (2) storing edge constraint conditions of each polyhedral area and the best control law in step (1) into a control chip of the on-chip model predictive control; (3) inquiring the corresponding area in the control chip according to the actual object state of the non-linear full-bridge DC-DC system, and then controlling a feedback control input of the full-bridge DC-DC system according to the best control law in the polyhedral areas. By the control method, rapid control of the full-bridge DC-DC system is realized and robustness is strong.

Description

Control method of full-bridge DC-DC system based on-chip model predictive control

Technical Field

The invention relates to the technical field of power supply control, in particular to a control method of a full-bridge DC-DC system based on-chip model predictive control.

Background

The full-bridge DC-DC system has wide application in the technical field of power control, the invention with application publication number CN102170234A discloses a digital control system of a DC/DC converter, which comprises a power conversion circuit, wherein the power conversion circuit comprises an input end, a full-bridge circuit consisting of four field effect tubes and an output end which are sequentially connected, the input end and the output end are both provided with sampling circuits, the key point is that the digital control system also comprises a control unit, the control unit comprises a control module, an analog-to-digital conversion module, a pulse width modulation module and a CAN gateway module, the analog-to-digital conversion module, the pulse width modulation module and the CAN gateway module are connected with the control end of the field effect tubes, and the output end of the pulse; the control module receives information of the input end and the output end through the analog-to-digital conversion module and the sampling circuit, and controls the output of the pulse width modulation module according to the information.

The multivariable constraint system can be directly subjected to predictive control processing by using a full-bridge DC-DC system model, so that the multivariable constraint system becomes one of the most widely applied optimal control strategies, however, the traditional full-bridge DC-DC system model predictive control needs to be subjected to online optimization calculation, a corresponding control law is calculated according to the actual object state of the full-bridge DC-DC system, and then optimal control parameters are found out, and due to the fact that the online calculation is large in calculation amount, the operation speed is slow, and the control hysteresis is caused, so that the method is only suitable for controlling slow process objects, such as petrochemical engineering and other processes.

Due to the fact that the sampling period of the converter of the full-bridge DC-DC system is very short, and due to the mixed nature and the multiple constraint conditions of the full-bridge DC-DC system, the predictive control design based on online repeated optimization calculation of the full-bridge DC-DC system model in the prior art becomes difficult.

The programmable DSP chip can conveniently use a signal processing algorithm with good performance in real-time signal processing, and meanwhile, the programmable DSP chip becomes one of software widely applied by algorithm researchers based on the strong numerical calculation and analysis capability of MATLAB, so that the algorithm can be conveniently stored and called by using the MATLAB and the DSP chip, and the programmable DSP chip becomes one of candidate ways for improving the control efficiency of the full-bridge DC-DC system.

Disclosure of Invention

The invention discloses a full-bridge DC-DC system control method based on-chip model predictive control, which is characterized in that a segmented affine model of the full-bridge DC-DC system is established, the segmented affine model is utilized to solve optimal control laws of different object state areas in an off-line manner, and then the corresponding optimal control laws are selected according to the actual object state of the full-bridge DC-DC system, so that the full-bridge DC-DC system is rapidly controlled.

A control method of a full-bridge DC-DC system based on-chip model predictive control comprises the following steps:

(1) dividing the object state of the nonlinear full-bridge DC-DC system into different polyhedral areas, and obtaining the optimal control law in each polyhedral area according to the control law of each polyhedral area;

(2) storing the boundary constraint conditions and the optimal control law of each polyhedral area in the step (1) into a control chip for on-chip model predictive control;

(3) and inquiring the polyhedral area to which the full-bridge DC-DC system belongs in the control chip according to the actual object state of the full-bridge DC-DC system, and controlling the feedback control input of the full-bridge DC-DC system according to the optimal control law in the polyhedral area.

The method separates the optimization process of the control parameters from real-time control, namely, the optimal control laws corresponding to all object states are calculated in advance, the corresponding relation between the object states and the optimal control laws is stored in a control chip of an on-chip model, and then the corresponding optimal control laws are selected according to the actually acquired object states of the full-bridge DC-DC system, so that the full-bridge DC-DC system is rapidly controlled.

The on-chip model comprises a full-bridge DC-DC system, a control chip, a power tube driving circuit and a detection filter circuit, wherein the full-bridge DC-DC system comprises a power supply, a full-bridge DC-DC circuit and a load, and the load can be any electric element, such as an electric lamp.

The control chip can adopt a DSP chip, and the MATLAB is utilized to compile the array into a C language file to be stored in the DSP chip.

The full-bridge DC-DC circuit can be a unidirectional full-bridge DC-DC circuit or a bidirectional full-bridge DC-DC circuit, and can be controlled by adopting the method provided by the invention.

And determining the dimension of the polyhedron region by using the number of the object states of the full-bridge DC-DC system and the auxiliary control quantity.

The object states of the full-bridge DC-DC system can be set voltage, set current, upper and lower limits of voltage, upper and lower limits of current and other hard constraint conditions, and the number of the object states can be changed according to needs.

The auxiliary control quantity comprises adjustable set voltage, adjustable set current, upper and lower limits of adjustable voltage, upper and lower limits of adjustable current and the like, and can be selected according to requirements.

Preferably, the boundaries of each polyhedron region and the control law inside each polyhedron satisfy a linear constraint condition. The boundary of each polyhedral area meets the linear constraint condition, the control law inside each polyhedral area also meets the linear constraint condition, the division of the polyhedral areas and the calculation of the optimal control law are facilitated, and the boundary constraint condition of each polyhedral area and the optimal control law inside each polyhedron are stored in a control chip of the on-chip model in an array form.

Preferably, in the step (1), a piecewise affine model of the full-bridge DC-DC system is first established by using a mechanism modeling manner, and then the object state of the full-bridge DC-DC system is divided into different polyhedral regions according to the piecewise affine model.

Preferably, in the step (2), the boundary constraint conditions and the optimal control law of each polyhedron region are compiled into C language by MATLAB and stored in the DSP file directory. Based on the powerful data operation and the analysis ability of MATLAB, the analysis and calculation of the model can be carried out rapidly, the model is used for developing and debugging the on-chip model rapidly, meanwhile, the MATLAB can be used for conveniently selecting the type and the number of the object states, the array stored in the DSP is updated in time, the control chip can conveniently inquire the optimal control law according to the object states acquired in real time, and then the feedback control input of the full-bridge DC-DC system is controlled.

The invention relates to a full-bridge DC-DC system control method based on-chip model predictive control, which separates optimization calculation and real-time control of control parameters in the control process, utilizes an established piecewise affine model of the full-bridge DC-DC system to solve an optimal control law in an off-line manner, divides a full-bridge DC-DC state object state into different polyhedral areas, each polyhedral area is related to the control law of the object state, then selects a corresponding optimal control law on line according to the actual full-bridge DC-DC object state to complete the rapid control of the full-bridge DC-DC system, and realizes the real-time communication between a computer and a control chip through the serial communication between the control chip and the computer, thereby achieving the purpose of real-time monitoring of the computer.

Drawings

FIG. 1 is a flow chart of a control method of the present invention;

FIG. 2 is a schematic diagram of a system for on-chip model predictive control in accordance with the present invention;

FIG. 3 is a flow chart of a real-time control process in the method of the present invention;

FIG. 4 is a graph showing the result of voltage control in the example;

fig. 5 is a graph showing the result of voltage control when the load fluctuates in the embodiment.

Detailed Description

The following describes the control method of the full-bridge DC-DC system based on-chip model predictive control in detail with reference to the accompanying drawings.

As shown in fig. 1, a method for controlling a full-bridge DC-DC system based on-chip model predictive control includes the following steps:

(1) establishing a piecewise affine model of the full-bridge DC-DC system by using a mechanism modeling mode, dividing the object state of the full-bridge DC-DC system into different polyhedral areas according to the piecewise affine model, and calculating the optimal control law in each polyhedral area;

performing piecewise affine on the power tube switching cycles of the nonlinear full-bridge DC-DC system, dividing each power tube switching cycle T into sigma sections (namely sigma sub-cycles), approximating a controlled object (the full-bridge DC-DC system) to a linear object in each section, and establishing a piecewise affine model;

x(k+1)=A_σx(k)+B_σu(k)+f_σ

y(k)=C_σx(k)+D_σu(k)+g_σ

subj.to

y_min≤y(k)≤y_max

u_min≤u(k)≤u_max

wherein k is the running time of the full-bridge DC-DC system;

u is a control input, and u is ∈ [0,1 ];

x is the object state;

y is the output of the controlled object;

A_σ、B_σ、C_σ、D_σas coefficients of a discrete state space model of the controlled object, f_σ、g_σFor the input and output disturbance coefficients, the sampling period is T/sigma.

The piecewise affine model is based on the periodic change di/dt of the output current of the nonlinear full-bridge DC-DC system and can be divided into two states of output current increase and output current decrease, the sampling period of the continuous model is discretized by T/sigma according to the two states of the output current, and a space model of the two states is obtained, and the following formula is shown in the specification:

$x(k+1)=\left\{\begin{array}{cc}\mathrm{Fx}\left(k\right)+\mathrm{gu}\left(k\right)& \mathrm{when}(\mathrm{di}/\mathrm{dt})>0\\ \mathrm{Gx}\left(k\right)+\mathrm{gu}\left(k\right)& \mathrm{when}(\mathrm{di}/\mathrm{dt})<0\end{array}\right.$

wherein k is the running time of the full-bridge DC-DC system;

x is the object state;

f, G and G are all coefficients of a discrete state space model of a controlled object (a full-bridge DC-DC system), and the sampling period is T/sigma.

The sampling period of the output current being the switching period of the power tubeIn any sampling period of the output current, the space model always satisfies the equation of the above formula, but when the output current is changed from di/dt>0 jump to di/dt<When 0, mean value modeling is needed, and if jump is in the mu sub-period of the switching period of the power tube, a space model can be obtained as

<math> <mrow> <mi>x</mi> <mrow> <mo>(</mo> <mi>k</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>=</mo> <mfenced open='{' close=''> <mtable> <mtr> <mtd> <mi>Fx</mi> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> <mo>+</mo> <mi>gu</mi> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> </mtd> <mtd> <mi>when</mi> <mrow> <mo>(</mo> <mi>di</mi> <mo>/</mo> <mi>dt</mi> <mo>)</mo> </mrow> <mo>></mo> <mn>0</mn> </mtd> </mtr> <mtr> <mtd> <mi>Gx</mi> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> <mo>+</mo> <mi>gu</mi> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> </mtd> <mtd> <mi>when</mi> <mrow> <mo>(</mo> <mi>di</mi> <mo>/</mo> <mi>dt</mi> <mo>)</mo> </mrow> <mo>&lt;</mo> <mn>0</mn> </mtd> </mtr> <mtr> <mtd> <mo>[</mo> <mrow> <mo>(</mo> <mi>&sigma;u</mi> <mo>-</mo> <mi>u</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mi>F</mi> <mo>+</mo> <mrow> <mo>(</mo> <mi>&mu;</mi> <mo>-</mo> <mi>&sigma;u</mi> <mo>)</mo> </mrow> <mi>G</mi> <mo>]</mo> <mi>x</mi> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> <mo>+</mo> <mi>gu</mi> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> </mtd> <mtd> <mi>when</mi> <mrow> <mo>(</mo> <mi>di</mi> <mo>/</mo> <mi>dt</mi> <mo>)</mo> </mrow> <mi>switches</mi> </mtd> </mtr> </mtable> </mfenced> </mrow> </math>

Wherein k is the running time of the full-bridge DC-DC system;

x is the object state;

f, G and G are all coefficients of a discrete state space model, and the sampling period is T/sigma;

u is a control input, and u is ∈ [0,1 ];

f is equivalent to A in the piecewise affine model_σ；

G corresponds to A in the piecewise affine model_σ；

(u-. mu. +1) F + (u-. mu. + u) G corresponds to A in the piecewise affine model_σ；

g corresponds to B in the piecewise affine model_σ。

The space model is more accurate than a traditional average period model, because the modeling method only approximately models in the sub-period of the current change jump, and the traditional average period model models the whole switching period of the power tube, and the error is larger. The control law of the object state of the full-bridge DC-DC system is the optimal control which satisfies the finite time constraint and satisfies the following formula:

<math> <mrow> <msub> <mi>J</mi> <mi>N</mi> </msub> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>k</mi> </msub> <mo>,</mo> <msub> <mi>u</mi> <mi>k</mi> </msub> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mi>min</mi> <mrow> <msub> <mi>u</mi> <mn>0</mn> </msub> <mo>,</mo> <mo>.</mo> <mo>.</mo> <mo>.</mo> <msub> <mi>u</mi> <mrow> <mi>N</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </mrow> </munder> <msub> <mrow> <mo>|</mo> <mo>|</mo> <msub> <mi>Q</mi> <mi>f</mi> </msub> <msub> <mi>x</mi> <mi>N</mi> </msub> <mo>|</mo> <mo>|</mo> </mrow> <mn>1</mn> </msub> <mo>+</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>0</mn> </mrow> <mrow> <mi>N</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <mrow> <mo>(</mo> <msub> <mrow> <mo>|</mo> <mo>|</mo> <msub> <mi>Ru</mi> <mi>k</mi> </msub> <mo>|</mo> <mo>|</mo> </mrow> <mn>1</mn> </msub> <mo>+</mo> <msub> <mrow> <mo>|</mo> <mo>|</mo> <msub> <mi>Qx</mi> <mi>k</mi> </msub> <mo>|</mo> <mo>|</mo> </mrow> <mn>1</mn> </msub> <mo>)</mo> </mrow> </mrow> </math>

<math> <mrow> <mi>subj</mi> <mo>.</mo> <mi>to</mi> <msub> <mi>x</mi> <mi>k</mi> </msub> <mo>&Element;</mo> <mi>&Phi;</mi> <mo>,</mo> <mo>&ForAll;</mo> <mi>k</mi> <mo>&Element;</mo> <mo>{</mo> <mn>1</mn> <mo>,</mo> <mo>.</mo> <mo>.</mo> <mo>.</mo> <mo>,</mo> <mi>N</mi> <mo>}</mo> <mo>,</mo> </mrow> </math>

<math> <mrow> <msub> <mi>u</mi> <mi>k</mi> </msub> <mo>&Element;</mo> <mi>&Psi;</mi> <mo>,</mo> <mo>&ForAll;</mo> <mi>k</mi> <mo>&Element;</mo> <mo>{</mo> <mn>1</mn> <mo>,</mo> <mo>.</mo> <mo>.</mo> <mo>.</mo> <mo>,</mo> <mi>N</mi> <mo>-</mo> <mn>1</mn> <mo>}</mo> <mo>,</mo> </mrow> </math>

x₀=x(0).

wherein x is_kThe state of the controlled object is the state at the moment k;

u_kis the control input at time k;

x₀is the initial state of the full bridge DC-DC system;

phi is x_kIn a range of values of [ psi ], [ u ]_kThe value range of (a).

Is a weight column vector, and N is a prediction time domain;

m is the number of control inputs; n is the state number of the controlled object;

at the same time, the user can select the desired position,wherein A and B are coefficients of a discrete state space model of the controlled object, and A generally refers to the former A_σ(i.e. each A)_σAll satisfy the formula), B generally refers to the front B_σ。

All object states can be expressed by a linear combination of the initial state x (0) and the control input u. Therefore, the objective function J_N(x (0)) can be simplified to

$J_N\left(x\left(0\right)\right)=\mathrm{Hx}\left(0\right)+\underset{U_N}{\min }{\mathrm{VU}}_N$

s.t.GU_N≤W+Ex(0)

Wherein,

wherein u is^T ₀Is u₀Transposing;

n is a prediction time domain;

a and B are coefficients of a discrete state space model of a controlled object;

m is the number of control inputs;

n is the state number of the controlled object.

G, W, E are objective functions J_NThe value range of the constraint condition coefficient in (x (0)) can be determined by the value range of the controlled object x, the range of the input u and the constraint of the output y, and can be specifically determined according to the design requirementAnd determining.

For example, the initial state x (0) must be greater than or equal to 0.5, i.e., x (0) ≧ 0.5, and the number of controlled states N =1, and since the control input must satisfy u (k) ≧ 0, the number of control inputs m =1, and the prediction time domain N =2, then it is possible to obtain

<math> <mrow> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <mn>0</mn> </mtd> <mtd> <mn>0</mn> </mtd> </mtr> <mtr> <mtd> <mo>-</mo> <mn>1</mn> </mtd> <mtd> <mn>0</mn> </mtd> </mtr> <mtr> <mtd> <mn>0</mn> </mtd> <mtd> <mo>-</mo> <mn>1</mn> </mtd> </mtr> </mtable> </mfenced> <msub> <mi>U</mi> <mi>N</mi> </msub> <mo>&le;</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <mo>-</mo> <mn>0.5</mn> </mtd> </mtr> <mtr> <mtd> <mn>0</mn> </mtd> </mtr> <mtr> <mtd> <mn>0</mn> </mtd> </mtr> </mtable> </mfenced> <mo>+</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <mn>1</mn> </mtd> </mtr> <mtr> <mtd> <mn>0</mn> </mtd> </mtr> <mtr> <mtd> <mn>0</mn> </mtd> </mtr> </mtable> </mfenced> <mi>x</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </mrow> </math>

Wherein U is_N=U₂=[u₀,u₁]^T，

$G=\left[\begin{array}{cc}0& 0\\ -1& 0\\ 0& -1\end{array}\right],$ $W=\left[\begin{array}{c}-0.5\\ 0\\ 0\end{array}\right],$ $E=\left[\begin{array}{c}1\\ 0\\ 0\end{array}\right].$

x (0) is an optimization parameter, and an optimal decision vector U_N(i.e. optimal control law U_N) Depending on x (0).

Taking an object state x (0) of a full-bridge DC-DC system as a parameter variable, and obtaining an optimal decision vector U when the x (0) is changed in a certain polyhedron region_NThe change rule of (2) to obtain U_NAnd x (0), so thatAnd (4) repeated online optimization calculation of the space-free model predictive control algorithm.

The spatially feasible domain of object states for a full-bridge DC-DC system isWherein And p is the number of constraints, n is the dimension of the object state x, and i is the number of polyhedral regions into which it is divided.

U_NIs a piecewise linear continuous function of x (0), i.e.: u shape_N=M_ix(0)+P_iWherein M is_iAnd P_iIs U_NAnd a matrix of x (0) linear relationships, x(0)∈χ_iwhereinIs a spatial polyhedron of object states.

Using MATLAB to mixAnd U_N=M_ix(0)+P_iThe formed array is compiled by using C language, stored in a DSP file directory, and during actual control, the region chi is found out according to the actual full-bridge DC-DC object state x_iThen the corresponding optimal control law U can be obtained_N。

(2) And storing the boundary constraint conditions and the optimal control law of the polyhedral region into a control chip of the on-chip model.

And uploading the operating data of the full-bridge DC-DC system to the MATLAB by using serial port communication, and displaying the operating condition of the full-bridge DC-DC system in real time through a visual interface such as a GUI (graphical user interface).

As shown in FIG. 2, the on-chip model of the invention comprises a control chip (comprising a DSP and a MATLAB), a full-bridge DC-DC system (comprising a power supply, a DC-DC, a load), a detection filter circuit and a power tube driving circuit.

The object states of the full-bridge DC-DC system are divided into output voltage, filter inductance current, previous control input and voltage set value.

And pouring the mathematical model of the controlled object state into the MATLAB, performing full-bridge DC-DC object state region division and control law calculation in the corresponding region offline, compiling the object state region and the optimal state law in the corresponding region into a C language file, and storing the C language file in the DSP.

(3) And in the real-time control, the DSP controller calls a corresponding C language file, and according to actual running states such as detected object states (such as current values and voltage values) and the like, the DSP searches the C language file for an area where the object states fall through a quick search algorithm, so that the optimal control law in the corresponding area is obtained.

For example, the actual voltage x of a full bridge DC-DC system₁∈[0,0.5]Actual current x₂∈[0,1]Last control input x₃∈[0,1]And a voltage set value x₄∈[0.2,1]Composing object states, setting the upper and lower limits of each object state to compose a multidimensional space model, where x is the pair₁,x₂,x₃Performing per unit respectively, and determining the values according to the control input d ∈ [0,1]]Dividing the switching period of the power tube for controlling the input into three affine areas [0,1/3 ] according to the input amplitude]，(1/3,2/3]And (2/3, 1)]And establishing a piecewise affine model according to the three affine areas.

Affine model 1:

A₁=[0.9579 -0.2169 0 0;0.3756 0.9391 0 0;0 0 0 0;0 0 0 1] B₁=[0.2158;0.06986;1;0]

f₁=[0;0;0;0] C₁=[0.1895 0.9757 0 -1;0 0 -1 0] D₁=[0.03043;1] g₁=[0;0]

0≤u(1)≤0.33333 -3≤x(1)≤3 -1≤x(2)≤1 0≤x(3)≤1 0.2≤x(4)≤1

0.973477*x(1)-0.072084*x(2)+0.217132*u(1)≥-2.9342

0.973477*x(1)-0.072084*x(2)+0.217132*u(1)≤2.9342

0.965634*x(1)-0.143913*x(2)+0.216425*u(1)≥-2.9524

0.965634*x(1)-0.143913*x(2)+0.216425*u(1)≤2.9524

0.952591*x(1)-0.215715*x(2)+0.214562*u(1)≥-2.9832

0.952591*x(1)-0.215715*x(2)+0.214562*u(1)≤2.9832

affine model 2:

A₂=[0.9579 -0.2169 0 0;0.3756 0.9391 0 0;0 0 0 0;0 0 0 1] B₂=[0.2199;0.04227;1;0]

f₂=[-0.001383;0.009197;0;0] C₂=[0.1895 0.9757 0 -1;0 0 -1 0] D₂=[0.01175;1] g₂=[0.00623;0]

0.33333≤u(1)≤0.66667 -3≤x(1)≤3 -1≤x(2)≤1 0≤x(3)≤1 0.2≤x(4)≤1

0.997270*x(1)-0.073846*x(2)<=2.9318

0.965204*x(1)-0.143849*x(2)+0.218377*u(1)≥-2.9504

0.965204*x(1)-0.143849*x(2)+0.218377*u(1)≤2.9517

0.951741*x(1)-0.215523*x(2)+0.218493*u(1)≥-2.9792

0.951741*x(1)-0.215523*x(2)+0.218493*u(1)≤2.9819

affine model 3:

A₁=[0.9579 -0.2169 0 0;0.3756 0.9391 0 0;0 0 0 0;0 0 0 1] B₁=[0.222;0.0141;1;0]

f₁=[-0.002771;0.02798;0;0]C₁=[0.1895 0.9757 0 -1;0 0 -1 0] D₁=[0.00235;1] g₁=[0.01249;0]

0.66667≤u(1)≤1 -3≤x(1)≤3 -1≤x(2)≤1 0≤x(3)≤1 0.2≤x(4)≤1

0.997270*x(1)-0.073846*x(2)≤2.9318

0.989076*x(1)-0.147407*x(2)≤2.8755

0.951309*x(1)-0.215425*x(2+0.220462*u(1)≤2.982

as shown in fig. 3, the real-time control step in the control method of the present invention includes:

a. detecting to obtain the actual object stateWherein x₄The voltage setting value is set according to actual requirements.

b. Subtracting a set value X from the object state to obtain an object state error value delta X = X-X;

c. and judging whether the object state error value is smaller than the allowable error range tau or not. If so, the control input u is held_k+1For inputting u at a previous moment_kOtherwise, performing search calculation on the object state error value delta X = X-X;

d. finding out the area of the object state according to the object state value, and then finding out the area of the object state according to the control law U_N=M_ix(0)+P_iGet control input U_NAnd will U_NCorresponding u_k（U_NFor a vector, u in the vector is calculated₀As u_k) Assign u to_k+1If the object state is not in the feasible region, selecting the object state value closest to the actual state nearby;

e. applying a control input to the full-bridge DC-DC system through the power tube driving circuit;

f. object states and derived u of full-bridge DC-DC system_k+1Forming a new object state, and returning to the step a;

the operation of the object can be finally obtained according to the above steps as shown in table 1.

TABLE 1

Input voltage	200V
		Set voltage	50V
Weight Qf	[1 2 1 1]
		Weight R	1
Weight Q	[1 2 1 1]
		Number of sub-periods σ	3
SamplingFrequency of	10kHz

As shown in FIG. 4, the control voltage output of the full-bridge DC-DC system can be 50V by controlling according to the method of the invention, and the set voltage is reached in 4ms, as shown in FIG. 5, the load changes in 0.05s, but the full-bridge DC-DC system can obtain the correction tracking in 0.02s, which shows that the control method of the invention has strong robustness.

Claims (5)

1. A control method of a full-bridge DC-DC system based on-chip model predictive control is characterized by comprising the following steps:

(1) dividing the object state of a nonlinear full-bridge DC-DC system into different polyhedron regions, and obtaining the optimal control law in each polyhedron region according to the control law of each polyhedron region, which specifically comprises the following steps:

firstly, establishing a piecewise affine model of a full-bridge DC-DC system by utilizing a mechanism modeling mode, performing piecewise affine on the switching cycles of power tubes of a nonlinear full-bridge DC-DC system, dividing each switching cycle T of the power tubes into sigma sections, approximating the full-bridge DC-DC system to be a linear object in each section, and establishing the piecewise affine model;

x(k+1)＝A_σx(k)+B_σu(k)+f_σ

y(k)＝C_σx(k)+D_σu(k)+g_σ

subj.to

y_min≤y(k)≤y_max

u_min≤u(k)≤u_max

wherein k is the running time of the full-bridge DC-DC system;

u is a control input, and u is ∈ [0,1 ];

x is the object state;

y is the output of the controlled object;

A_σ、B_σ、C_σ、D_σas coefficients of a discrete state space model of the controlled object, f_σ、g_σThe sampling period is T/sigma for the input and output disturbance coefficients;

then, dividing the object state of the full-bridge DC-DC into different polyhedral areas according to a piecewise affine model to satisfy an objective function J_N(x (0)) is

$J_N\left(x\left(0\right)\right)=\mathrm{Hx}\left(0\right)+\underset{U_N}{\min }{\mathrm{VU}}_N$

s.t.GU_N≤W+Ex(0)

Wherein x (0) is an initial state;

wherein u is^T ₀Is u₀Transposing;

n is a prediction time domain;

a and B are coefficients of a discrete state space model of a controlled object;

m is the number of control inputs;

n is the dimension of the object state x;

g, W, E are objective functions J_NThe value range of the constraint condition coefficient in (x (0)) is determined by the value range of the controlled object x, the range of the input u and the constraint of the output y;

U_Nis a piecewise linear continuous function of x (0), i.e.: u shape_N＝M_ix(0)+P_iWherein M is_iAnd P_iIs U_NAnd a matrix of x (0) linear relationships,x(0)∈χ_iwhereinIs a spatial polyhedron of the state of the object, and p is the number of constraints, n is the dimension of the object state x, and i is the polyhedron region into which it is dividedThe number of the first and second groups is,is a weight column vector;

(2) storing the boundary constraint conditions and the optimal control law of each polyhedral area in the step (1) into a control chip for on-chip model predictive control;

(3) and inquiring the polyhedral area to which the full-bridge DC-DC system belongs in the control chip according to the actual object state of the full-bridge DC-DC system, and controlling the feedback control input of the full-bridge DC-DC system according to the optimal control law in the polyhedral area.

2. The method of controlling a full-bridge DC-DC system based on-chip model predictive control of claim 1, wherein the boundaries of each polyhedron region and the control law inside each polyhedron satisfy linear constraints.

3. The method of controlling the full-bridge DC-DC system based on the on-chip model predictive control of claim 1, wherein the full-bridge DC-DC system comprises a power supply, a full-bridge DC-DC circuit, and a load.

4. The method of controlling a full-bridge DC-DC system based on-chip model predictive control of claim 1, wherein the dimension of the polyhedral region is determined using the number of object states of the full-bridge DC-DC system and an auxiliary control amount.

5. The method for controlling the full-bridge DC-DC system based on the on-chip model predictive control as claimed in claim 1, wherein the boundary constraints and the optimal control law of each polyhedral region are compiled into C language by MATLAB in the step (2) and stored in a DSP file directory.