Noise suppression and anti-disturbance control method for DC-DC Buck control system based on cascaded linear anti-disturbance control

Abstract

The present invention belongs to the field of power electronic automatic control, and discloses a noise suppression and anti-interference control method of a cascade LADRC based on a filter for a DC-DC-buck. The content is that, based on the ADRC idea, a new cascade LADRC noise suppression and anti-interference control method based on a filter is proposed for the output voltage amplitude and phase loss caused by the output filtering of a traditional buck converter. By comparing the simulation analysis of traditional double closed-loop PI, LADRC and LADRC containing a filter, the method of the present invention compensates for the output amplitude and phase loss while suppressing noise, effectively utilizes the real-time estimation and compensation capability of the cascade LADRC for the buck system, improves the robustness of the system, ensures the tracking accuracy of the system, and at the same time, effectively improves the anti-interference capability of the system in the case of sudden load change.

Description

Technical Field

[0001] The invention relates to a Buck converter voltage control technology. Background Art

[0002] DC-DC step-down converters play an important role in the processing and conversion of electric energy. Their processing and conversion methods have become a hot topic of research. At present, converters are developing towards high efficiency, high frequency, light weight, green, and integration at an unprecedented speed. DC-DC switching converters mainly realize the transmission of energy at the input and output ends and the conversion of electric energy forms by performing high-frequency on and off operations on power switches. Nowadays, with the increasing frequency of system control and load mutation regulation, it is becoming more and more important to find an effective switch controller control algorithm to achieve stable and efficient power conversion.

[0003] In power electronic control, the classic PID has been widely used for nearly a century due to its natural anti-disturbance ability and independence from the model. However, due to its inherent defects, it will bring about problems such as poor system control robustness, time lag and easy system oscillation. In the 1990s, Researcher Han Jingqing proposed a newer nonlinear control theory, namely active disturbance rejection control (ADRC). Active disturbance rejection control based on extended state observer (LESO) is a nonlinear robust control technology. Its most prominent feature is that all uncertain factors acting on the controlled object are attributed to "unknown disturbances", and their errors are estimated and compensated through the extended observer. However, due to the nonlinear non-smooth feedback structure adopted by the original ADRC, its theoretical analysis is very difficult, and there are many controller parameters that need to be adjusted. Based on a deep understanding of Han Jingqing's active disturbance rejection idea, Dr. Gao Zhiqiang proposed a universal linear active disturbance rejection (LADRC) controller, which enabled ADRC technology to break through the bottleneck of complex parameter setting and simplified the analysis and implementation of ADRC. However, the above research did not consider other constraints except for the uncertain factors of the system. In the input and output links of the actual system, there will be output noise. Although the use of traditional filters can suppress noise, it will change the amplitude and phase of the system output while eliminating the noise, causing a difference between the filter output and the original output of the system without adding noise. This difference will be amplified in LESO, resulting in the state of the original system cannot be well estimated, which will cause system instability and lead to a decrease in control quality. Summary of the invention

[0004] The present invention provides a noise suppression and anti-disturbance control method for a Buck converter voltage control system, which combines a cascaded linear anti-disturbance control and a low-pass filter. Through the real-time estimation and compensation effect of the cascaded linear anti-disturbance control, after the filter suppresses the output noise, the output voltage amplitude and phase loss caused by filtering are compensated, so that the robustness of the control system is effectively improved while suppressing the output noise, and the control accuracy of the closed-loop system is ensured; for sudden load changes, the designed cascaded linear anti-disturbance control can well eliminate the disturbance caused by the sudden load change through the estimation and compensation effect, thereby meeting the performance requirements of the noise suppression and anti-disturbance control of the Buck converter voltage control system.

[0005] According to one aspect of an embodiment of the present invention, a Buck converter voltage control system noise suppression and anti-interference control system is provided, comprising:

[0006] LESO1, treats all the uncertain internal and external partial disturbances of the system as total disturbances and expands them into a new state variable x3, and uses a special feedback mechanism to establish a state observer that can observe the expanded state variable, observe the new state variable together with the state variables x1 and x2, and compensate it in real time through the virtual control variable q;

[0007] LESO2, expands the total disturbance of the filter part system into a new state variable A special feedback mechanism is used to establish a state observer that can observe the expanded state, and the new state variable is compared with the state variable Observe and control the virtual quantity q together f Perform real-time compensation;

[0008] The controller defines the system state error ε(e f ), combined with the disturbance estimation of LESO output Solve the virtual control quantity q(q f ) and then compensate the total disturbance in real time;

[0009] The comparator compares the virtual control quantity q output by the actuator with the sawtooth wave to obtain the modulated PWM wave, which is then used as a switching signal to adjust the duty cycle D through the drive circuit, thereby controlling the output voltage V0.

[0010] Beneficial effects of the present invention:

[0011] (1) The control method of the present invention uses the real-time estimation compensation function of the cascaded linear anti-disturbance control to compensate for the output voltage amplitude and phase loss caused by filtering after the traditional buck converter suppresses the output noise. While suppressing the noise, it effectively improves the robustness of the control system and ensures the control accuracy of the closed-loop system.

[0012] (2) For sudden load changes, the designed cascaded linear anti-disturbance control can eliminate the interference caused by it through disturbance compensation, improve the system's anti-interference ability, and greatly improve the noise suppression and anti-disturbance control performance of the Buck converter voltage control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the drawings of the embodiment are briefly introduced below.

[0014] Figure 1 A topological diagram of a Buck converter voltage control system using a cascaded linear auto-disturbance rejection and noise suppression control method provided in one embodiment of the present invention.

[0015] Figure 2 A voltage output comparison diagram of a Buck converter voltage control system under four control methods, namely, dual closed-loop PI, linear auto-disturbance rejection, linear auto-disturbance rejection with filter, and cascaded linear auto-disturbance rejection with filter, when output noise exists, provided in one embodiment of the present invention.

[0016] Figure 3 A voltage output comparison diagram of a Buck converter voltage control system provided by an embodiment of the present invention when a load mutation occurs under four control methods: dual closed-loop PI, linear auto-disturbance rejection, linear auto-disturbance rejection with filter, and cascaded linear auto-disturbance rejection with filter.

[0017] Figure 2 , 3 We can see that the electric noise suppression and interference rejection control method of the cascaded linear auto-interference rejection system based on output filtering and error compensation of the present invention is more robust, more accurate and has stronger anti-interference ability than double closed-loop PI, linear auto-interference rejection and linear auto-interference rejection with filter. DETAILED DESCRIPTION

[0018] The present invention provides a noise suppression and anti-disturbance control method for a Buck converter voltage control system. Firstly, a state equation of the Buck converter is established. Then, based on a second-order system of a control object, a cascaded linear auto-disturbance rejection control method is designed to achieve real-time estimation and compensation of the total disturbance of the system. Thus, the output amplitude and phase loss caused by a traditional buck converter are compensated through the real-time estimation and compensation effect of the cascaded linear auto-disturbance rejection control. Furthermore, the disturbance caused by a sudden load change is eliminated through the disturbance compensation capability of the cascaded linear auto-disturbance rejection control, thereby improving the anti-disturbance capability of the system.

[0019] Figure 1 The topology diagram of the Buck converter voltage control system using the cascaded linear active disturbance rejection control method. Among them, LESO is the linear extended state observer; the large dashed box is the cascaded linear active disturbance rejection control block diagram; the small dashed box is the filter block diagram; P fis the output noise; y is the system output voltage with high-frequency output noise; v r is the set voltage; q, q f is the virtual control quantity output by the actuator; the low-pass filter is

[0020] The system voltage is made to operate according to the set state through the cascaded linear active disturbance rejection control method based on output filtering. The Buck converter voltage control system is a second-order control object. The disturbance effect affecting the output of the controlled object is expanded into a new state variable using the linear extended state observer (LESO), and the new state variable is observed together with other state variables. After the filter Q(s) suppresses the influence of the output noise, the output amplitude and phase loss caused by the output filtering are compensated through the real-time estimation compensation of the cascaded linear active disturbance rejection control (CLADRC).

[0021] Figure 2 The voltage output comparison diagram of the Buck converter voltage control system under four control methods when output noise exists, namely, dual closed-loop PI, linear anti-disturbance control, linear anti-disturbance control with filter, and cascaded linear anti-disturbance control with filter. The set voltage is 20V; the double dashed line of CLADRC (with filter) is the output curve of the present invention; the dotted line of dual closed-loop PI is the output curve of voltage control based on traditional PID; the dotted line of LADRC and the solid line of LADRC (with filter) are the output curves of voltage control under single linear anti-disturbance control with or without filter.

[0022] According to the working principle of Buck converter, the state space average model of Buck converter voltage control system is established. In the model, when the switch is turned on, Kirchhoff's voltage law is used to obtain From Kirchhoff's current law we get When the switch is open, Kirchhoff's voltage law gives From Kirchhoff's current law we get From this, the state space equation of the system can be obtained as:

[0023]

[0024] Where V in [V] is the input voltage, D∈[0,1] is the duty cycle, R[Ω] is the load resistance of the circuit, C[F] is the filter capacitor, L[H] is the filter inductor, V0[V] is the average capacitor voltage, I L [A] is the average inductor current, so it can be regarded as an uncertain nonlinear system:

[0025]

[0026] In the formula, a 11 =-1 / RC, a12 =-1 / C,a 21 =-1 / L, b = V in / L, c = 1, y is the output voltage and sensor noise p f The control goal of the present invention is to control the duty cycle D by applying an algorithm to achieve the goal of quickly and accurately making the output voltage V0 follow the set voltage v r purpose.

[0027] The linear ADRC method consists of three parts: 1) Linear tracking differentiator (LTD): used to track the set voltage v r , 2) Linear extended state observer (LESO): used to estimate the total disturbance of the system in real time, 3) Linear state error feedback (LSEF): used to compensate for the total disturbance of the system in real time.

[0028] LTD is relatively independent in LADRC. In the present invention, LTD can realize that when the system starts, the output voltage of the Buck circuit can quickly track the system setting voltage v without overshoot by arranging a reasonable transition process. r , and get the input differential signals of various orders That is, v1→v r , i = 1, 2. Let v = [v1 v2 v3] T , LTD can be expressed as:

[0029]

[0030] In the formula, b TD =[0 0 -a1R 3 ] T ∈R 3 , R is the adjustment gain that determines the tracking speed, coefficient a j (j=1,2,3) must satisfy the matrix It's Hurwitz's.

[0031] LESO is the core technology of LADRC. Its model independence and robustness are based on LESO's real-time estimation of the total disturbance of the system. In order to estimate the system state in real time, the total disturbance of the system is expressed as a new state variable, and the state equation is also transformed into an expanded state space equation.

[0032]

[0033] In the formula, a 12 b can be regarded as a time-varying uncertain control gain, satisfying b1<a 12 b<b2 (where b1 and b2 are both constants), take the constant b0∈(b1, b2)≠0, and then (a 12 b-b0)u as the unknown control disturbance. All the internal and external partial disturbances of the system uncertainty are regarded as the total disturbance It can be expressed as:

[0034]

[0035] Therefore, the dynamic equation of the output voltage V0(t) can be expressed as:

[0036]

[0037] Take the state variable x1 = V0, Then the total disturbance of the system is expanded into a new state variable x3 = g(·), recorded as Let x = [x1 x2 x3] T , the expanded state space equation of the system is:

[0038]

[0039] In the formula, C = [1 0 0], On this basis, let z = [z1z2z3] T , where z (i) →x (i) , i = 1, 2, 3, z is the estimated vector of the system state x, using the real-time input q, output y, the corresponding LESO can be constructed as:

[0040]

[0041] Where K = [k1 k2 k3] T is the observer error feedback gain matrix to be designed, with coefficient k j (j=1,2,3) must satisfy the matrix It's Hurwitz's. Let e i =x i -z i , i=1,2,3, let e=[e1 e2 e3] T ,have

[0042]

[0043] In the formula, when (A-KC) is Hurwitz and h is bounded, the estimation error converges asymptotically and the LESO design is effective.

[0044] The system control object belongs to the second-order system. LTD generates tracking signals. The observations of the third-order LESO are the output observations and disturbance observations. The state errors of the system are defined as ε1 = z1-v1, ε2 = z2-v2, and the real-time compensation of the total disturbance x3 is considered. Let ε = [ε1 ε2] T , the following LSEF can be designed:

[0045]

[0046] In the formula, l = [l1 l2], the coefficient l i (i=1,2) must satisfy the matrix It's Hurwitz's.

[0047] In the present invention, LESO is the core of LADRC. Increasing the observer gain can improve its tracking performance, but it will also amplify high-frequency noise, thereby causing large-scale high-frequency vibration of the system control quantity, which is unbearable for the actuator. In order to eliminate the influence of high-frequency measurement noise, the present invention introduces a low-pass filter to filter the output voltage, and then sends the filtered signal to the controller. The filter usually exists in the form of RC, and performs filtering work based on the principle that capacitors pass high frequencies and block low frequencies, and inductors pass low frequencies and block high frequencies. It is generally a first-order inertia link, and its form is:

[0048]

[0049] In the formula, y q is the filtered signal, a i , b i 、c i is the filter coefficient, and the LESO based on the filter output becomes:

[0050]

[0051] Although the filter can effectively suppress the output noise, it will also lose the output amplitude and phase, making the filter output y q There is a difference between the original output y of the system without adding it, and then this difference is input into LESO, the control inertia is large, resulting in the inability to accurately estimate the state of the original system, which ultimately affects the control tracking performance of LADRC. In view of this, the present invention expands the output of the filter into a new state variable x of the system f =y q , that is, the filter and the system are combined into a composite system, and the output amplitude and phase loss are eliminated by using the real-time estimation and compensation of LADRC. The system can be expressed as:

[0052]

[0053] The control level in the system is that the control quantity q controls the duty cycle u to drive V0, and then V0 is used as a virtual control quantity to drive x f , and then achieve the purpose of control. Since there are uncertainties in the control process of both parts of the system, that is, unknown external and internal disturbances, a new filter-based cascade LADRC noise suppression and anti-disturbance control method can be designed based on the LADRC idea.

[0054] For equation 11, combined with equation 13, it can be transformed into

[0055]

[0056] Where b i It can be regarded as a time-varying uncertain control gain, satisfying b3<b i <b4 (where b3 and b4 are both constants), take the constant Will As the unknown control disturbance in the filter system. Then, the total disturbance of the filter system can be set as g f (y q ,V0,p f ):

[0057]

[0058] Then the dynamic equation based on the filter output can be

[0059]

[0060] make remember The expanded state space equation of the filter-based Buck converter voltage control composite system is obtained as follows:

[0061]

[0062] In the formula, C f =[1 0], On this basis, where z f →x′ f is x′ f The estimated vector of the virtual control quantity q f , output y, the corresponding LESO can be constructed as:

[0063]

[0064] In the formula, Kf is a matrix and vector of the same form as K, except that the coefficients are changed from k to k f (subscript omitted), the order is changed from 3 to 2.

[0065] In a new cascade LADRC controller based on filter design, The filter-based cascade LADRC controller is:

[0066]

[0067] In the formula, l f A TD , b TD ,l are matrices or vectors of the same form, except that the coefficients are changed from a,l (subscript omitted) to a f ,l f (subscript omitted), the gain changes from R,r to R f ,r f , the order is changed from 3 to 2.

[0068] The stability judgment method of the filter-based cascade LADRC noise suppression and anti-disturbance control method is: Let the control error matrix of LADRC e = [e1 e2 e3] T =[x1-z1 x2-z2 x3-z3] T , Expands to:

[0069]

[0070] In the formula, γ=[γ1 γ2 γ3], coefficient γ i (i=1,2,3) must satisfy the matrix It's Hurwitz's.

[0071] According to the Lyapunov stability criterion, when the total disturbance g(·) of the system is bounded, (A-KC)<0 is negative definite, and the Hurwitz criterion is satisfied, the observer tracking error e→0, then the filter-based cascade LADRC control method is effective.

[0072] When a sudden load change occurs in the Buck converter voltage control system, the cascaded linear anti-disturbance control method designed in the present invention can eliminate the interference caused by the disturbance through a disturbance compensation strategy, thereby improving the anti-disturbance capability of the system.

[0073] like Figure 3As shown, it is a voltage output comparison diagram of the Buck converter voltage control system under four control methods: dual closed-loop PI, linear auto-disturbance rejection, linear auto-disturbance rejection with filter, and cascaded linear auto-disturbance rejection with filter when there is a sudden change in load. Among them, the set voltage is 20V; the double-dash line of CLADRC (with filter) is the output curve of the present invention; the double-dash line of the dual closed-loop PI is the voltage control output curve based on the traditional PID; the dotted line of LADRC and the solid line of LADRC (with filter) are the output curves of voltage control under single linear auto-disturbance rejection with or without filter.

[0074] It can be clearly seen from the curve in the figure that the filter-based cascade linear anti-disturbance control method designed in the present invention makes the voltage output curve adjustment time shorter and the oscillation frequency smaller, and can track the set voltage faster and more stably. When a sudden load change occurs at 0.12s, the output voltage overshoot is smaller, the adjustment time is shorter, the fluctuation is smaller, and the state can be restored faster and more stably, and the system has stronger anti-interference ability and robustness.

Claims

1. A Buck converter voltage control system noise suppression and anti-interference control method, characterized in that: The real-time estimation compensation of the cascaded linear anti-disturbance control can compensate the output voltage amplitude and phase loss caused by filtering after the filter suppresses the output noise. At the same time, the cascaded linear anti-disturbance control can eliminate the disturbance caused by the sudden change of load through the estimation compensation. According to the working principle of the Buck converter, the state space equation of the Buck converter voltage control system is obtained by Kirchhoff's voltage and current laws: Where V in [V] is the input voltage, D∈[0,1] is the duty cycle, R[Ω] is the load resistance of the circuit, C[F] is the filter capacitor, L[H] is the filter inductor, V0[V] is the average capacitor voltage, I L [A] is the average inductor current; Based on the real-time estimation of the total disturbance of the system by the extended state observer, the state variable x1 = V0 is taken. To estimate the system state in real time, the total disturbance of the system is expressed as a new state variable x3 = g(·), denoted by Let x = [x1 x2x3] T , we get the expanded state space equation of the system: In the formula, C = [1 0 0], On this basis, let z = [z1 z2 z3] T , where z (i) →x (i) , i = 1, 2, 3, z is the estimated vector of the system state x, and the extended state observer is established using real-time input q and output y: Where K = [k1 k2 k3] T is the observer error feedback gain matrix to be designed, with coefficient k j Need to satisfy the matrix is Hurwitz's, j = 1, 2, 3, Expand the output of the filter into the new state variable x of the system f =y q , the system is expressed as: In the formula, y q is the filtered signal, a i 、b i 、c i is the filter coefficient; The control level in the system is that the control quantity q controls the duty cycle u to drive V0, and then V0 is used as a virtual control quantity to drive x f , and then achieve the purpose of control; suppose the total disturbance of the filter part system is g f (y q ,V0,p f ),make remember The expanded state space equation of the filter-based Buck converter voltage control composite system is obtained: In the formula, C f =[1 0], On this basis, where z f →x′ f is x′ f The estimated vector of the virtual control quantity q f , output y, the corresponding extended state observer is constructed as: In the formula, K f is a matrix and vector of the same form as K, except that the coefficients are changed from k to k f , the order is changed from 3 to 2; Through LTD, when the system starts, the Buck circuit output voltage tracks the system set voltage v r , and get the input differential signals of various orders That is, v1→v r , i=1,2;Let v=[v1 v2 v3] T , LTD is expressed as: In the formula, R f To determine the tracking speed, the coefficient a j (j=1,2) must satisfy the matrix It's Hurwitz's. v f is a matrix of the same form as v; Through LTD, a virtual control quantity q generated by a linear active disturbance rejection f As the actuator input of another linear ADRC, two linear ADRCs are connected in series to form a cascaded linear ADRC, realizing cascade control of the control system. The expression is: In the formula, b TD =[00-a1R 3 ] T ∈R 3 ,q f is the virtual control quantity output by the linear anti-disturbance actuator, and the coefficient a j Need to satisfy the matrix is Hurwitz's, j = 1, 2, 3, The virtual control quantity q output by the other linear anti-disturbance actuator is compared with the sawtooth wave to obtain a modulated PWM wave, which is then used as a switching signal to adjust the duty cycle D through a driving circuit, thereby controlling the output voltage V0.

Images