An Inverse Decoupling Adaptive Sliding Mode Control Method

Abstract

The present invention belongs to the category of robust control technology of multilevel DC-DC (MD) Buck converter with constant power load (CPL), and specifically relates to an inverse decoupling adaptive sliding mode control method, including the following steps: S1: establishing a nonlinear mathematical model with CPL MD Buck converter, using the inverse system method to verify the reversibility of the system, and deriving the inverse system expression; S2: decoupling the system linearization into multiple pseudo-linear subsystems; S3: introducing an adaptive mechanism to update the sliding mode switching gain in real time, and designing adaptive sliding mode controllers for the subsystems respectively; S4: verifying the stability and robustness of the closed-loop system based on Lyapunov theory; S5: performing simulation comparison with the existing control method to verify that the inverse decoupling adaptive sliding mode control method has stronger robustness and better superiority; S6: building a principle prototype. The present invention can solve the problem of balancing the flying capacitor voltage with CPL MD Buck converter and the problem of decoupling the flying capacitor voltage from the output voltage.

Description

Technical Field

[0001] The present invention belongs to the field of robust control technology of a multilevel DC-DC (MD) Buck converter with a constant power load (CPL), and specifically relates to an inverse decoupling adaptive sliding mode control method. Background Art

[0002] Multilevel DC-DC (MD) converters have been gradually applied and studied in the low-voltage and low-power fields because of their advantages such as lower voltage stress, more level output, and smaller filter volume compared to traditional two-level DC-DC converters. In order for MD converters to work properly and have good working performance, it is extremely critical to ensure that the flying capacitor (FC) voltage is stable to its own equilibrium value. However, MD converters belong to a type of multi-input and multi-output, strongly coupled nonlinear system. Not only are there a large number of switches and complex working modes, but there is also a strong coupling between the output voltage and the FC voltage, which brings challenges to the balance of the FC voltage and the decoupling control of the system. In order to solve the above problems, the processing methods in the existing literature can be roughly divided into three categories, namely, the method of adding external circuits, the linear decoupling method, and the exact feedback linearization (EFL) decoupling method.

[0003] Regarding the method of adding external circuits, the document "STALA R. The switch-mode flying-capacitor DC-DC converters with improved natural balancing [J]. IEEE Transactions on Industrial Electronics, 2010, 57 (4): 1369-1382" proposed an enhanced FC voltage balancing method, which achieves the goal of improving FC voltage balance by adding peripheral circuits to the original topology.

[0004] The document "DELMAS L, GATEAU G, MEYNARD T. Stacked multicell converter: control and natural balancing [C] / / IEEE 33rd Annual Power Electronics Specialists Conference, Australia, 2002: 689-694" proposed a parallel RLC resonant circuit control scheme, which speeds up the FC voltage regulation by connecting an RLC resonant circuit in parallel at the converter output. The method of adding an external circuit increases the volume and loss of the system, and it is difficult to take into account the coupling of the FC voltage and the output voltage, which increases the difficulty of controlling the system. For this reason, the linear decoupling method was proposed.

[0005] The document "Zhang Yuanyuan, Ruan Xinbo. A control strategy for flying capacitor voltage in multilevel DC converters [J]. Proceedings of the CSEE, 2004, 24(08): 37-41" proposed a balanced control strategy for the MD Buck converter. This strategy achieves the decoupling of the FC voltage and the output voltage through a linear decoupling matrix, and then designs a proportional-integral controller to improve the system stability.

[0006] The document "SUN RY, WU XH, LI WL, et al. Equivalent sliding mode control for multi-level buck converter [C] / / IEEE International Conference on Aircraft Utility Systems. Beijing, China, 2016: 460-465" proposed an equivalent sliding mode control method for the MD Buck converter. The linear decoupling matrix was first used to achieve system decoupling, and then the equivalent sliding mode controller was designed separately to improve the control effect of the system. However, the linear decoupling control method is based on a small signal model and is difficult to meet the requirements of existing applications for higher control performance and stronger robustness. In order to overcome the above defects, a decoupling control method based on EFL is proposed.

[0007] The document "WU Jiarong, LU Yimin. Decoupling and optimal control of multilevelbuck DC-DC converters with inverse system theory[J]. IEEE Transactions on Industrial Electronics, 2020, 67(9): 7861-7870" proposed an EFL control scheme for the MD Buck converter. On the basis of achieving system decoupling, the optimal control technology was used to improve the dynamic performance of the system.

[0008] The document "CAI Pei, WU Xiaohua, SUN Runyu. Exact feedback linearization of general four-level buck DC-DC converters [C] / / 29th Chinese Control and Decision Conference. Chongqing, China, 2017: 4638-4643" proposed an EFL control method for the four-level Buck converter, which achieved the decoupling of the FC voltage and the output voltage and reduced the difficulty of controller design. However, the EFL method has a strong dependence on the accurate mathematical model of the controlled object.

[0009] Sliding mode control (SMC) is highly robust to system uncertainty, external disturbances and unmodeled dynamics, and can effectively overcome the above-mentioned defects of the EFL method. However, the switching gain of SMC depends on the upper limit of uncertainty, which is difficult to estimate in the actual operation of the system, resulting in difficulty in maintaining the stability of the system. In addition, with the promotion of power electronics applications, a large number of electronic loads are connected to MD converters. When the output power of the electronic load is constant, this type of load can be regarded as a constant power load (CPL). The negative impedance characteristics of CPL will not only increase the nonlinearity of the system, but also seriously reduce the damping coefficient of the system, bringing difficulties to the stable control of the MD converter with CPL. Existing research on MD converters still stays in the case where only a pure resistive load is connected to the output end, which is difficult to meet the needs of powering multiple different types of loads at the same time.

[0010] Linear decoupling proportional integral control (LD-PIC) is a widely used existing technology. This strategy realizes the decoupling of FC voltage and output voltage through a linear decoupling matrix, and then designs a proportional-integral controller to improve system stability. The specific implementation process is in the document "Zhang Yuanyuan, Ruan Xinbo. A control strategy for flying capacitor voltage in multi-level DC converter [J]. Proceedings of the CSEE, 2004, 24(08): 37-41". However, the linear decoupling control method is based on a small signal model and is difficult to meet the requirements of existing applications for higher control performance and stronger robustness.

[0011] In order to solve the problem of balancing the flying capacitor voltage and decoupling the flying capacitor voltage and the output voltage of a CPL MD Buck converter, the present invention proposes an inverse decoupling adaptive sliding mode control method, which can offset the influence of the negative impedance characteristics of CPL, effectively improve the stability of the converter, and have stronger anti-interference and robustness as well as better control performance.

[0012] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0013] The object of the present invention is to provide an inverse decoupling adaptive sliding mode control method to solve the problem of balancing the flying capacitor voltage of a CPL MDBuck converter and the problem of decoupling the flying capacitor voltage from the output voltage.

[0014] In order to achieve the above object, the present invention provides the following technical solutions:

[0015] An inverse decoupling adaptive sliding mode control method comprises the following steps:

[0016] S1: Establish a nonlinear mathematical model of the CPL MD Buck converter, use the inverse system method to verify the reversibility of the system, and derive the inverse system expression;

[0017] S2: Decouple the system linearization into multiple pseudo-linear subsystems to offset the influence of negative impedance characteristics;

[0018] S3: Introduce an adaptive mechanism to update the sliding mode switching gain in real time and design adaptive sliding mode controllers for each subsystem;

[0019] S4: Verify the stability and robustness of the closed-loop system based on Lyapunov theory;

[0020] S5: Compare the simulation with the existing control method to verify that the inverse decoupling adaptive sliding mode control method has stronger robustness and better superiority;

[0021] S6: Build a prototype.

[0022] Preferably, in step S1, (v Ck ,i L , v o ) is the state variable, with d=(d1,…,d k , …, d p-1 , d p ) is the input variable, and the nonlinear mathematical model of the multi-level DC-DC Buck converter with constant power load established by the state space average modeling method is: Among them, v Ck C k Voltage across the terminals, i L is the inductor current, d k Q k The duty cycle of the switch tube Q is the basic unit. k 、Diode D k and the flying capacitor C k Composition, and Q k and D k Not conducting each other; for C k , k=1,2,…,p-1; and for Q k and D k , k=1,2,…,p; The MD Buck converter with CPL is composed of p basic units connected in series, including p switches, p diodes and p-1 flying capacitors; In an ideal state, the duty cycles of the p switches are equal, all d, and the phases are turned on in sequence with a difference of 2π / p. At this time, the FC voltage is balanced at kV in / p, output voltage v o =d×V in , where V in is the input voltage; in order to make the system have a good tracking effect, the FC voltage and the output voltage are taken as the output function respectively, that is, For y in which the FC voltage and the output voltage are used as the output function k and o Find the first and second derivatives respectively, As the input of the inverse system, the inverse system expression is obtained as Among them, K and F satisfy:

[0023]

[0024]

[0025] Preferably, in step S2, the inverse system is connected in series to the nonlinear mathematical model with CPL MD Buck converter established in step S1 to form a pseudo-linear subsystem Where s is the Laplace operator. After being processed by the inverse system method, the FC voltage and the output voltage are linearly decoupled into p-1 first-order linear subsystems and one second-order linear subsystem, respectively.

[0026] Preferably, when designing an adaptive sliding mode controller for the subsystem in step S3, the errors of p-1 first-order linear subsystems are defined as Among them, V Ck for y k The reference value of p-1 first-order linear subsystems is obtained by taking the first-order derivative of the error The sliding surface and exponential reaching law are designed as s1=e1,…,s k =e k ,…,s p-1 =e p-1 , Combining the above equations, we can get the control laws of p-1 FC voltages: Define the error of the second-order linear subsystem as e o =y o -V oref , and the derivative is Where V oref for y o The reference value of the second-order sliding surface and sliding mode control law are Where β is a constant greater than zero; Combining the above equations, the control law of the output voltage is: Introduce an adaptive law to update c k and c o , and Respectively represent Q k and the estimated value of M, the adaptive law is designed as Among them, ρ k >0 and γ>0.

[0027] Preferably, in step S4, the Lyapunov function is but Therefore, the control law d of the CPL MD Buck converter is designed as

[0028] Preferably, in step S5, the existing control method linear decoupling proportional integral control (LD-PIC) is used for simulation comparison with the inverse decoupling adaptive sliding mode control method.

[0029] Preferably, in step S6, a three-level DC-DC Buck converter prototype with a constant power load is built and compared with the LD-PIC.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] (1) The inverse decoupling adaptive sliding mode control method of the present invention adopts the inverse system theory to realize the linearization and decoupling of the system, offset the influence of the negative impedance characteristics of CPL, and facilitate the design of the controller.

[0032] (2) The inverse decoupling adaptive sliding mode control method of the present invention uses SMC technology to reduce the dependence of the inverse system method on the precise mathematical model, which is beneficial to enhancing the robustness of the system.

[0033] (3) The inverse decoupling adaptive sliding mode control method of the present invention introduces adaptive technology to estimate the switching gain of SMC, which is beneficial to improving the stability performance of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is the circuit topology diagram of the CPL MD Buck converter;

[0035] Figure 2 This is the control block diagram of the CPL MD Buck converter;

[0036] Figure 3 The simulation waveforms of the two control methods are shown when R jumps at 0.1s (27.5Ω→70Ω), 0.15s (70Ω→27.5Ω), 0.2s (27.5Ω→10000Ω) and 0.25s (10000Ω→27.5Ω).

[0037] Figure 4 The simulation waveforms of the two control methods when P changes at 0.35s (25W→256W) and 0.45s (256W→25W) respectively;

[0038] Figure 5 V in Simulation waveforms of the two control methods at 0.6s (60V→48V) and 0.7s (48V→78V) respectively;

[0039] Figure 6 V oref Simulation waveforms of the two control methods when the voltage changes at 0.85s (30V→40V) and 0.95s (40V→25V) respectively;

[0040] Figure 7 The experimental results are as follows when R periodically jumps between 50Ω and 100Ω;

[0041] Figure 8 These are the experimental results when P periodically jumps between 7W and 3W;

[0042] Fig. 9 V in The experimental results when the voltage changes between 28V→26V→30V;

[0043] Fig.10 V oref The experimental results when jumping between 11V→8V→11V respectively.

[0044] It should be noted that Figure 3-10 In the figure, the left figure (a) shows the simulation result of the linear decoupled proportional integral control proposed in the document "Zhang Yuanyuan, Ruan Xinbo. A control strategy for flying capacitor voltage in multi-level DC converter [J]. Proceedings of the CSEE, 2004, 24(08): 37-41", and the right figure (b) shows the simulation result of the inverse decoupled adaptive sliding mode control proposed in the present invention. DETAILED DESCRIPTION

[0045] The following is a clear and complete description of the technical solution of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Referring to the accompanying drawings, first, a nonlinear mathematical model with a CPL MD Buck converter is established, the inverse system method is used to prove the reversibility of the system, the inverse system expression is derived, and the system linearization is decoupled into multiple pseudo-linear subsystems to offset the influence of the negative impedance characteristics. Then, an adaptive mechanism is introduced to update the sliding mode switching gain in real time, and adaptive sliding mode controllers are designed for the subsystems respectively. Then, the stability and robustness of the closed-loop system are verified based on the Lyapunov theory. Simulation comparison with the existing control method shows that the inverse decoupling adaptive sliding mode control method proposed in the present invention has stronger robustness and better superiority. Finally, a principle prototype is built for experiment, and the experimental comparison results verify the correctness and effectiveness of the inverse decoupling adaptive sliding mode control method of the present invention.

[0047] Figure 1 The circuit topology of the CPL MD Buck converter is given. Among them, P is the power of CPL, and the shaded dotted box is its basic unit, which consists of the switch tube Q k 、Diode D k and the flying capacitor C k Composition, and Q k and D k There is no conduction between them.k , k=1,2,…,p-1; and for Q k and D k , k=1,2,…,p. The MD Buck converter with CPL is composed of p basic units connected in series, including p switches, p diodes and p-1 flying capacitors. In an ideal state, the duty cycles of the p switches are equal, all d, and the phases are turned on in sequence with a difference of 2π / p. At this time, the FC voltage is balanced at kV in / p, output voltage v o =d×V in , where V in is the input voltage.

[0048] Assume that the system operates in the inductor current continuous mode, with (v Ck ,i L , v o ) is the state variable, where v Ck C k Voltage across the terminals, i L is the inductor current, d=(d1,…,d k , …, d p-1 , d p ) is the input variable, where d k Q k The duty cycle of the MD Buck converter with CPL can be established by using the state space average modeling method:

[0049]

[0050] In the formula, R, L, and C are resistance, inductance, and capacitance, respectively. From formula (1), we know that there is a strong coupling between the FC voltage and the output voltage. In order to improve the system performance, it is necessary to decouple the system.

[0051] For equation (1), the system has p input variables and p+1 state equations, that is, the system dimension n=p+1. In order to make the system have a good tracking effect, the FC voltage and output voltage are taken as the output function, that is,

[0052]

[0053] For y in (2), k and o Find the first-order derivative and the second-order derivative respectively, and we get

[0054]

[0055] From formula (3), we know that each term contains d. in represents the first-order derivative of y1 with respect to time t, then the Jacobian matrix It can be calculated as

[0056]

[0057] From formula (4), we know that the determinant According to the inverse system theory, system (4) has a vector relative order α = (α1, α2, …, α p-1 ,α o ) T =(1,1,…,1,2) T , the sum of which is α1+α2+…+α p-1 +α o =p+1, system (1) can be completely decoupled. As the input of the inverse system, the inverse system expression can be obtained from formula (3):

[0058]

[0059] In the formula, K and F satisfy:

[0060]

[0061]

[0062] Connecting equation (5) in series before system (1) can form a pseudo-linear subsystem:

[0063]

[0064]

[0065] Where s is the Laplace operator. After the inverse system method, the FC voltage and the output voltage are linearly decoupled into p-1 first-order linear subsystems and one second-order linear subsystem, respectively.

[0066] When designing an adaptive sliding mode controller for a subsystem, the error of the p-1 first-order linear subsystem is defined as

[0067]

[0068] Where V Ck for y k The reference value of . Taking the first-order derivative of equation (8) we get

[0069]

[0070] According to the sliding mode control theory, the sliding surface and exponential reaching law are designed as follows:

[0071] s1=e1,…,s k =e k ,…,sp-1 = e p-1 (10)

[0072]

[0073] Combining equations (9), (10), and (11), the control laws for p - 1 FC voltages can be obtained as

[0074]

[0075] Define the error of the second - order linear subsystem as

[0076] e o = y o - V oref (13)

[0077] where V oref is the reference value of y o . Taking the second - derivative of equation (13), we get

[0078]

[0079] Take the second - order sliding surface and the sliding - mode control law as

[0080]

[0081]

[0082] where β is a constant greater than zero. Combining equations (13) - (16), the control law of the output voltage is obtained as

[0083]

[0084] Next, the stability and robustness of the system are verified.

[0085] Assume that there are disturbances G k (t) and O(t) in systems (9) and (14) respectively. Then equations (9) and (14) can be modified as

[0086]

[0087] where G k (t)>0 and O(t)>0 are both bounded, that is, |G k (t)| < Q k , |O(t)| < M, where Q k and M are the upper bounds of G k (t) and O(t) respectively.

[0088] Take the Lyapunov function as

[0089]

[0090] Combining equations (11), (16) and (18), taking the first-order derivative of equation (19) yields

[0091]

[0092] To ensure Must meet

[0093]

[0094] From formula (21), we know that the control gain c k and c o The choice depends on G k The upper bound Q of (t) and O(t) k and M. If c k and c o Choosing too large to ensure the stability of the system may cause strong jitter in the system, seriously affecting the performance of the system. Therefore, the present invention introduces an adaptive law to update c k and c o Assumptions and Respectively represent Q k and the estimated value of M, the adaptive law can be designed as

[0095]

[0096] In the formula, ρ k >0 and γ>0.

[0097] To verify systems (6) and (7), the closed-loop system is asymptotically stable using the control laws in (12) and (16) and the adaptation law in (22).

[0098] Take the Lyapunov function as

[0099]

[0100] Substitute equations (19) and (22) into equation (23) and take the first-order derivative of equation (23):

[0101]

[0102] Therefore, systems (6) and (7) are k , e o )=(0,0) and is asymptotically stable. Combining (5), (12), (17) and (22), the control law d of the CPL MD Buck converter can be designed as

[0103]

[0104] Thus, the design Figure 2 The control block diagram of the CPL MD Buck converter is shown in Figure 1. In the figure, the linear decoupling of the system is completed by sampling the system variables, and then the adaptive sliding mode control algorithm is used to control the linear subsystem to achieve the purpose of balancing the FC voltage and the output voltage.

[0105] In order to verify the effectiveness and superiority of the inverse decoupling adaptive sliding mode control method proposed in this invention, a simulation model of a seven-level DC-DC Buck converter with CPL was built on the Matlab / Simulink platform for research, where p=6, and a closed-loop control Buck converter was used to equalize CPL. The system parameters are: V in =60V, V oref =30V, C1=C2=C3=C4=C5=200μF, L=1mH, C=330μF, R=27.5Ω, f s =50kHz, P=25W. To ensure the stability and robustness of the system, the parameters of the inverse decoupling adaptive sliding mode control method of the present invention are selected as follows: 11 =c 22 =c 33 =c 44 =c 55 =1×10 5 , ρ1=ρ2=ρ3=ρ4=ρ5=200, β=900, γ=800, c oo =1×10 5 The proposed inverse decoupling adaptive sliding mode control (ID-ASMC) is compared with the linear decoupling proportional integral control (LD-PIC) proposed in the literature “Zhang Yuanyuan, Ruan Xinbo. A control strategy for flying capacitor voltage in multilevel DC converters [J]. Proceedings of the CSEE, 2004, 24(08): 37-41”.

[0106] Figure 3 The simulation waveforms of the two control methods are given when R jumps at 0.1s (27.5Ω→70Ω), 0.15s (70Ω→27.5Ω), 0.2s (27.5Ω→10000Ω) and 0.25s (10000Ω→27.5Ω). It can be seen from the figure that under the two control methods, v C1 , v C2 , v C3 , v C4 , v C5Balanced at 10V, 20V, 30V, 40V, 50V, namely V in / 6,2V in / 6,3V in / 6,4V in / 6,5V in / 6,v o It is stable at 30V and can provide 25W power to CPL to maintain the stability of the entire system. However, after comparison, it is found that at the moment of R change, under LD-PIC, the FC voltage and v o There are certain fluctuations, but under the proposed ID-ASMC, the system shows good anti-interference performance.

[0107] Figure 4 The simulation waveforms of two control methods are given when P changes at 0.35s (25W→256W) and 0.45s (256W→25W). Under LD-PIC, v o All of them have serious oscillations and are difficult to stabilize at 30V. In addition, v C1 , v C2 , v C3 , v C4 , v C5 are affected to varying degrees and fluctuate around their respective equilibrium values. Under the proposed ID-ASMC, the FC voltage and v o They can all be stabilized at the reference value and are minimally affected by changes in P, showing stronger robustness.

[0108] Figure 5 Given V in The simulation waveforms of the two control methods when the voltage changes at 0.6s (60V→48V) and 0.7s (48V→78V). in The change of represents the change of FC voltage reference value. Under LD-PIC, v o There are serious voltage fluctuations, v C1 , v C2 , v C3 , v C4 , v C5 Can be balanced at V in / 6,2V in / 6,3V in / 6,4V in / 6,5V in / 6, but FC voltage is affected by V in Under the proposed ID-ASMC, the FC voltage and v o V in The impact of the change is minimal, which verifies that the proposed ID-ASMC achieves the decoupling of FC voltage and output voltage.

[0109] Figure 6 Given V oref The simulation waveforms of the two control methods are shown in Figure 1 when the change time is 0.85s (30V→40V) and 0.95s (40V→25V). Under LD-PIC, v C1 , v C2 , v C3 , v C4 , v C5 Both are affected by V oref The impact of the change, v o There is voltage overshoot. However, under the proposed ID-ASMC, the FC voltage is almost not affected by V oref The impact of the change, v o Can smoothly track V oref The jump of FC voltage provides a constant 25W power to CPL, which further verifies that the proposed ID-ASMC can effectively decouple the FC voltage and output voltage.

[0110] In order to verify the correctness of the inverse decoupling adaptive sliding mode control method proposed in this invention, a three-level DC-DC Buck converter prototype with CPL was built and compared with LD-PIC. The system parameters are designed as follows: V in =28V, V oref =11V, R=50Ω, C=940μF, C1=220μF, L=330μH, f s =50kHz, P=7W. The control parameters are: c1=900, c 11 =1×10 6 , β=6×10 3 , γ=1×10 3 , c oo =1×10 7 .

[0111] Figure 7 The experimental results are given when R periodically jumps between 50Ω and 100Ω. C1 Balanced at 14V, that is, V in / 2,v o Stable at 11V. When R jumps, under LD-PIC, v C1 and v o There are certain fluctuations, and under the proposed ID-ASMC, v C1 and v o It is almost unaffected by changes in R and exhibits strong anti-interference performance.

[0112] Figure 8 The experimental results are given when P periodically jumps between 7W and 3W. Under LD-PIC, v C1After severe oscillation, it is balanced at 14V, v o There is a certain voltage fluctuation, and under the proposed ID-ASMC, v C1 and v o It is minimally affected by changes in P and exhibits strong robust performance.

[0113] Fig. 9 Given V in The experimental results are shown below when the voltage changes between 28V→26V→30V. C1 Balanced at V in / 2,v o Stable at 11V. When V in When the ID-ASMC is changed, it has a better control effect than the LD-PIC, and v o V in The impact of the change is minimal, achieving v C1 and v o decoupling.

[0114] Fig.10 Given V oref The experimental results of the transition between 11V→8V→11V are shown below. Obviously, under ID-PIC, v C1 and v o There are voltage fluctuations in both cases, and the proposed ID-ASMC has better dynamic regulation performance than LD-PIC, v C1 Almost unaffected by V oref The impact of the change further verifies that v o and v C1 decoupling.

[0115] The present invention proposes an inverse decoupling adaptive sliding mode control method, which solves the balancing problem of the flying capacitor voltage of the CPL MD Buck converter and the decoupling problem of the flying capacitor voltage and the output voltage, offsets the influence of the negative impedance characteristics of the CPL, and effectively improves the stability of the converter. After simulation comparison with the LD-PIC method, the results show that the ID-ASMC proposed in the present invention has stronger anti-interference performance and robustness, and improves the dynamic response performance of the system. Finally, an experimental comparison with the LD-PIC is carried out, and the results show that the ID-ASMC proposed in the present invention has better control performance and shows greater superiority. The inverse decoupling adaptive sliding mode control method proposed in the present invention can be extended to other CPL MD DC-DC converter control applications, and provides a reference for the nonlinear stability control of the CPL MD DC-DC converter.

[0116] The foregoing description of specific exemplary embodiments of the present invention is for the purpose of illustration and demonstration. These descriptions are not intended to limit the present invention to the precise form disclosed, and it is clear that many changes and variations can be made based on the above teachings. The purpose of selecting and describing the exemplary embodiments is to explain the specific principles of the present invention and its practical application, so that those skilled in the art can realize and utilize various different exemplary embodiments of the present invention and various different selections and changes. The scope of the present invention is intended to be limited by the claims and their equivalents.

Claims

1. An inverse decoupling adaptive sliding mode control method, characterized in that: The steps include: S1: Establish a nonlinear mathematical model of a multi-level DC-DC Buck converter with a constant power load, use the inverse system method to verify the reversibility of the system, and derive the inverse system expression: S2: Decouple the system linearization into multiple pseudo-linear subsystems to offset the impact of negative impedance characteristics: S3: Introduce an adaptive mechanism to update the sliding mode switching gain in real time and design adaptive sliding mode controllers for each subsystem; S4: Verify the stability and robustness of the closed-loop system based on Lyapunov theory; S5: Compare the simulation with the existing control method to verify that the inverse decoupling adaptive sliding mode control method has stronger robustness and better superiority; S6: Build a prototype; In step S1, With (v Ck ,i L , v o ) is the state variable, with d=(d1,...,d k , ..., d p-1 , d p ) is the input variable, and the nonlinear mathematical model of the multi-level DC-DC Buck converter with constant power load established by the state space average modeling method is: Among them, v Ck C k Voltage across the terminals, i L is the inductor current, d k Q k The duty cycle of the switch tube Q is the basic unit. k 、Diode D k and the flying capacitor C k Composition, and Q k and D k Not conducting each other; for C k , k = 1, 2, ..., p-1; and for Q k and D k , k=1,2,...,p; The multi-level DC-DC Buck converter with constant power load is composed of p basic units connected in series, including p switches, p diodes and p-1 flying capacitors; In an ideal state, the duty cycles of the p switches are equal, all d, and the phases are turned on in sequence with a difference of 2π / p. At this time, the FC voltage is balanced at kV in / p, output voltage v o =d×V in , where V in is the input voltage; in order to make the system have a good tracking effect, the FC voltage and the output voltage are taken as the output function respectively, that is, For y in which the FC voltage and the output voltage are used as the output function k and o Find the first and second derivatives respectively, As the input of the inverse system, the inverse system expression is obtained as follows: Among them, K and F satisfy:

2. The inverse decoupling adaptive sliding mode control method according to claim 1, characterized in that: In step S2, the inverse system is connected in series to the nonlinear mathematical model of the multi-level DC-DC Buck converter with constant power load established in step S1 to form a pseudo-linear subsystem. Where s is the Laplace operator. After being processed by the inverse system method, the FC voltage and the output voltage are linearly decoupled into p-1 first-order linear subsystems and one second-order linear subsystem, respectively.

3. The inverse decoupling adaptive sliding mode control method according to claim 2, characterized in that: When designing an adaptive sliding mode controller for the subsystem in step S3, the errors of the p-1 first-order linear subsystems are defined as Among them, V Ck for y k The reference value of p-1 first-order linear subsystems is obtained by taking the first-order derivative of the error The sliding surface and exponential reaching law are designed as s1=e1,…,s k =e k ,…,s p-1 =e p-1 , Combining the above equations, we can get the control laws of p-1 FC voltages: Define the error of the second-order linear subsystem as e o =y o -V oref , and the derivative is Where V oref for y o The reference value of the second-order sliding surface and sliding mode control law are Where β is a constant greater than zero; Combining the above equations, the control law of the output voltage is: Introduce an adaptive law to update c k and c o , and Respectively represent Q k and the estimated value of M, the adaptive law is designed as Among them, ρ k >0 and γ>0.

4. The inverse decoupling adaptive sliding mode control method according to claim 3 is characterized in that: In step S4, the Lyapunov function is taken as but The control law d of the multi-level DC-DC Buck converter with constant power load is designed as 5. The inverse decoupling adaptive sliding mode control method according to claim 4, characterized in that: In step S5, the existing control method linear decoupling proportional integral control is simulated and compared with the inverse decoupling adaptive sliding mode control method.

6. The inverse decoupling adaptive sliding mode control method according to claim 5, characterized in that: In step S6, a three-level DC-DC Buck converter prototype with a constant power load is built and compared with the linear decoupled proportional integral control.

Images