Robust Adaptive Disturbance Rejection Control Method for DC-DC Converters Based on Immersion and Invariant Theory

Abstract

The present invention relates to a robust adaptive disturbance rejection control method for DC converters based on immersion and invariant theory. Based on the original linear active disturbance rejection controller, this method utilizes the immersion and invariant theory to achieve online estimation by measuring the output voltage online and using the total disturbance estimated in real time by an extended state observer and the control variable calculated by the voltage outer loop. This method is not aimed at providing accurate system parameters, but can make the given system parameters change online in the direction of improving the system dynamic response according to the change trend of the controlled variable, thereby enhancing the ability of the entire system to resist external disturbances. The present invention starts from the two major conditions of the converter, namely healthy and faulty, and designs input voltage step disturbances, output load step disturbances, and disturbances at the moment of fault. Experimental results show that the proposed method has stronger robustness than the traditional linear active disturbance rejection controller in both healthy and faulty states.

Description

Technical Field

[0001] The present invention belongs to the field of control science and control engineering, and proposes a robust disturbance rejection controller with adaptive ability and applies it to a DC converter. Background Technique

[0002] With the gradual depletion of fossil fuels, clean energy power generation systems such as wind power generation systems, photovoltaic power generation systems, and fuel cell power generation systems have received extensive research attention. In these renewable energy systems, DC converters always play an important role in energy conversion. However, DC converters are strongly nonlinear systems, and it is difficult to design a suitable controller for them. In addition, for new energy systems such as photovoltaic power generation and wind power generation, their output voltage and power are greatly affected by the external weather, resulting in unpredictable disturbances in the input voltage and current of the DC converter, which poses a severe challenge to the stable and optimal control of the converter. During the actual working process, the working performance of the converter will also be severely disturbed by the output load and the parameter uncertainty disturbance caused by the aging of power devices. In addition, in a more severe working environment, the DC converter will also cause the entire system to shut down due to device failure. Therefore, in order to improve the working performance of the entire new energy system, it is necessary to design a controller with high robustness on the basis of considering all these emergencies.

[0003] In the development process of control theory in the past few decades, a series of advanced nonlinear control strategies have been developed, mainly including sliding mode control, robust control, adaptive control, optimal control, disturbance rejection control, and even fuzzy logic control and artificial intelligence control. Although in academic research and laboratory environments, these control theories can achieve far better performance than traditional controllers, they are often difficult to be implemented in actual engineering practices. At present, the active disturbance rejection control method is given great hope by a large number of engineering practitioners. The active disturbance rejection control was first proposed by Mr. Han Jingqing. It is a generalization based on a profound consideration of the traditional proportional-integral controller and the concepts of state space and observer in modern control theory. The control performance shown by the active disturbance rejection control established by Mr. Han Jingqing is very good, but in the actual design process, it involves the design of many parameters, and the complexity of the parameter tuning process hinders its application in engineering practices. Later, with the efforts of Teacher Gao Zhiqiang, the active disturbance rejection controller was successfully modified into a linear form, and the number of controller parameters was reduced to two, which greatly promoted the application of this theory.

[0004] Theoretical and practical results show that the control performance of linear active disturbance rejection control (LADRC) depends to a large extent on the performance of its observer. The traditional LADRC uses a linear extended state observer with fixed gains and fixed frequencies, which cannot guarantee the operating performance of the controlled object under large disturbances and wide-frequency external disturbances. Therefore, in the literature "Active Disturbance Rejection Voltage Control of a Floating Interleaved DC–DC Boost Converter With Switch Fault Consideration" by S. Zhuo, A. Gaillard, L. Guo, L. Xu, D. Paire and F. Gao, in IEEE Transactions on Power Electronics, vol. 34, no. 12, pp. 12396-12406, Dec. 2019, a method is proposed in which the gains of the extended state observer adaptively change with the operating state of the system. Compared with the traditional LADRC, this method can significantly improve the operating performance of the converter system. However, the design of this method relies heavily on the physical model of the controlled object. For different control objects, different physical models need to be constructed, so it is not general. Summary of the Invention

[0005] Technical Problems to be Solved

[0006] In order to overcome the limitations of the traditional LADRC in the control of DC converters, the present invention proposes a more general robust adaptive disturbance rejection control method that does not rely on the actual physical model of the controlled object. This method can improve the ability of the DC converter to resist internal parameter perturbations and external disturbances in the normal operating state, and at the same time ensure that the DC converter can still maintain a good operating state in the fault state.

[0007] Technical Solution

[0008] A robust adaptive disturbance rejection control method for a DC converter based on the immersion and invariance theory, characterized by the following steps:

[0009] Step 1: Construct the state-space equations of the DC converter in the healthy state and the switch-fault state respectively;

[0010] Step 2: Design a cascaded control structure based on the voltage outer loop and the current inner loop for the DC converter; first, use the sliding mode control theory to design the current inner loop, and give the closed-loop transfer function of the converter current inner loop by using the concept of ideal sliding mode linearization;

[0011] Step 3: For the current inner-loop closed-loop transfer function obtained in Step 2, design an extended state observer using the traditional linear active disturbance rejection control theory to estimate the total internal and external disturbances of the system;

[0012] Step 4: For the total disturbance of the system obtained in Step 3, design a proportional controller according to the pre-selected control bandwidth to complete the design process of the entire traditional active disturbance rejection controller;

[0013] Step 5: On the basis of the traditional linear active disturbance rejection controller, realize the online real-time estimation of the system gain parameters based on the immersion and invariance theory to complete the design process of the proposed robust adaptive disturbance rejection control method.

[0014] Further technical solution of the present invention: The state-space equation described in Step 1:

[0015] When the converter operates in a healthy state:

[0016]

[0017] Among them, L1 and L2 are the input inductance values of the two-phase circuit respectively; R L1 , R L2 are the parasitic resistances of the two input inductors respectively; C1 and C2 are the output capacitance values of the two circuits respectively; i L1 , i L2 are the currents flowing through the two inductors respectively; v c1 , v c2 are the voltages across the two output capacitors respectively; L1 = L2 = L, C1 = C2 = C, v c1 = v c2 = v c , i L1 = i L2 = i L ; d H is the duty cycle definition in the healthy state, and R refers to the equivalent load resistance;

[0018] When the converter operates in a faulty state:

[0019]

[0020] Among them, d F represents the duty cycle of the converter after fault reconstruction, v o = v c1 + v c2 ;

[0021] Further technical solution of the present invention: The inner-loop closed-loop transfer function described in Step 2:

[0022] When the converter operates in a healthy state:

[0023]

[0024] Among them, V in is the input current, and I Lref is the reference current;

[0025] When the converter is operating in a fault state:

[0026]

[0027] Among them,

[0028] A further technical solution of the present invention: The total internal and external disturbances of the system described in step three are designed as follows:

[0029]

[0030] Among them, when the converter is in a healthy state, b = b1, y = v o ; when the converter is in a fault state, b = b2, y = v c ; b represents the system parameter of the converter, and f represents the total disturbance of the converter system.

[0031] A further technical solution of the present invention: The proportional controller described in step four:

[0032]

[0033] Among them, k p is the proportional control term gain, and V ref is the reference voltage of the converter.

[0034] Beneficial effects

[0035] A robust adaptive disturbance rejection control method for a DC converter based on the immersion and invariance theory proposed by the present invention. Based on the original linear active disturbance rejection controller, this method uses the immersion and invariance theory to achieve online estimation by measuring the output voltage online and using the total disturbance estimated in real time by the extended state observer and the control variable calculated by the voltage outer loop. This method is not intended to give accurate system parameters, but can make the given system parameters change online in the direction of improving the system dynamic response according to the change trend of the controlled variable, improving the ability of the entire system to resist external disturbances. This method can be applied to various DC converters to improve the working performance of the controlled object under internal and external disturbances and unknown potential faults.

[0036] To verify the superiority of the proposed method, a typical input-parallel output-series boost converter is taken as the research object. Starting from the healthy and faulty conditions of the converter respectively, input voltage step disturbance, output load step disturbance, and disturbance at the moment of fault are designed. The experimental results show that the proposed method has stronger robustness than the traditional linear active disturbance rejection controller in both healthy and faulty states. Compared with the existing technologies, as follows:

[0037] (1) Compared with the traditional linear active disturbance rejection control method, the proposed method can improve the ability of the DC converter to resist internal parameter perturbation, external input voltage disturbance, and external load disturbance under normal working conditions;

[0038] (2) Compared with the traditional linear active disturbance rejection control method, the proposed method can greatly improve the ability of the DC converter to resist internal and external disturbances in the fault reconstruction state, and can also improve the performance of the converter itself at the moment of fault occurrence. Description of the Drawings

[0039] The drawings are only for the purpose of showing specific embodiments and are not considered as a limitation of the present invention. Throughout the drawings, the same reference signs represent the same components.

[0040] Figure 1 Input-parallel output-series DC boost converter topology, (a) Topology in healthy state, (b) Fault reconstruction topology under S1 fault, (c) Fault reconstruction topology under S2 fault;

[0041] Figure 2 The overall control block diagram of the proposed control method in the input-parallel output-series DC boost converter;

[0042] Figure 3 The experimental test platform for a typical input-parallel output-series DC boost converter;

[0043] Figure 4 The experimental diagram of input voltage step disturbance of the converter in healthy state: (a) Traditional active disturbance rejection controller, (b) The proposed adaptive disturbance rejection controller;

[0044] Figure 5 The experimental diagram of input voltage step disturbance of the converter in faulty state: (a) Traditional active disturbance rejection controller, (b) The proposed adaptive disturbance rejection controller;

[0045] Figure 6 The experimental diagram of load current step disturbance of the converter in healthy state: (a) Traditional active disturbance rejection controller, (b) The proposed adaptive disturbance rejection controller;

[0046] Figure 7Load current step disturbance experimental diagram of the converter in the fault state: (a) Traditional active disturbance rejection controller, (b) Proposed adaptive disturbance rejection controller;

[0047] Figure 8 Disturbance rejection ability verification experimental diagram of the converter at the moment of fault: (a) Traditional active disturbance rejection controller, (b) Proposed adaptive disturbance rejection controller. Specific implementation manners

[0048] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0049] The present invention proposes a robust adaptive disturbance rejection control method based on immersion and invariant theory, which is divided into the following steps.

[0050] Step 1: Respectively construct the state space equations of the DC converter in the healthy state and the switch tube fault state, laying a foundation for the subsequent design of the controller;

[0051] Step 2: Design a cascaded control structure based on the voltage outer loop and the current inner loop for the DC converter. Since the proposed method mainly focuses on the control of the voltage outer loop, the sliding mode control theory is first used to design the current inner loop, and the closed-loop transfer function G ie (s) of the converter current inner loop is given by using the concept of ideal sliding mode linearization;

[0052] Step 3: For the current inner loop closed-loop transfer function obtained in Step 2, use the traditional linear active disturbance rejection control theory to design an extended state observer to estimate the total internal and external disturbances f of the system;

[0053] Step 4: For the total disturbance of the system obtained in Step 3, design a proportional controller according to the pre-selected control bandwidth to complete the design process of the entire traditional active disturbance rejection controller;

[0054] Step 5: On the basis of the traditional linear active disturbance rejection controller, realize the online real-time estimation of the system gain parameter b based on the immersion and invariant theory to complete the design process of the proposed robust adaptive disturbance rejection control method.

[0055] In order to enable those skilled in the art to better understand the present invention, the present invention will be described in detail below with reference to specific embodiments.

[0056] Refer to the attached Figure 1 , attached Figure 2 and attached Figure 3, the implementation steps of the present invention are as follows.

[0057] The present invention proposes a robust adaptive disturbance rejection control method for a DC converter based on immersion and invariant theory, which mainly includes the following five steps.

[0058] Step 1: Construct the state - space models of the input - parallel output - series DC boost converter in healthy and faulty states. It can be seen that whether it is the fault of switch S1 or switch S2, the topological structure of the DC converter after fault reconstruction is the same. Therefore, the state - space model of the converter in the faulty state can be uniformly expressed. Figure 1 When the converter is operating in the healthy state:

[0059] When the converter is operating in the healthy state:

[0060]

[0061] Among them, L1 and L2 are the input inductance values of the two - phase circuit respectively; R L1 , R L2 are the parasitic resistances of the two input inductors respectively; C1 and C2 are the output capacitance values of the two circuits respectively; i L1 , i L2 are the currents flowing through the two inductors respectively; v c1 , v c2 are the voltages across the two output capacitors respectively; d1 and d2 are the duty cycles of the two - phase circuit. In the present invention, the control of the two - phase circuit is realized by separate controllers respectively, and the control of the system after fault is completed by the control of the healthy - phase circuit. For convenience, the requirement that the voltages across the output capacitors are exactly equal is weakened, and a single voltage controller is used to adjust the voltages of the two - phase circuit. At the same time, for the convenience of subsequent analysis and design, it is assumed that the two - phase circuit is completely the same, with L1 = L2 = L, C1 = C2 = C, R L1 = R L2 = r, v c1 = v c2 = v c , i L1 = i L2 = i L , d1 = d2 = d, then Equation (1) can be simplified to:

[0062]

[0063] Among them, in order to distinguish the duty cycle of the converter in healthy and faulty states, the duty cycle in the healthy state is defined as d H , and R refers to the equivalent load resistance.

[0064] When the converter is operating in the faulty state:

[0065] After a fault occurs, due to the effect of the fault-tolerant control strategy, the original input-parallel output-series DC boost converter is reconstructed into a traditional DC boost converter. From the perspective of the inner-loop controller, the inductance value of the input inductor and the capacitance value of the output capacitor of the converter are reduced by half after the fault reconstruction, while the output voltage is doubled. Based on this phenomenon, the state-space equation of the converter after the fault reconstruction can be obtained as follows:

[0066]

[0067] where d F represents the duty cycle of the converter after the fault reconstruction, v o = v c1 + v c2 ;

[0068] Step 2: For the current inner-loop and voltage outer-loop cascade control structure adopted by the DC converter, apply the sliding-mode variable control theory to design the current inner-loop controller. Then, use the ideal sliding-mode linearization method to obtain the inner-loop current closed-loop transfer function of the DC converter. Since the inner-loop current controller based on the sliding-mode theory is not the protection point of the present invention, the detailed process is omitted here, and the inner-loop current closed-loop transfer functions G ie (s) of the DC converter in the healthy and faulty states are directly given as follows:

[0069] (1) When the converter is operating in the healthy state:

[0070]

[0071] where V in is the input current and I Lref is the reference current;

[0072] (2) When the converter is operating in the faulty state:

[0073]

[0074] where V in is the input current and I Lref is the reference current;

[0075] It can be obtained therefrom that before and after the fault: a1 = 2a2, b1 = 2b2, c1 = 2c2. Accordingly, appropriate sliding-mode controller parameters can be selected to ensure that the DC converter can work well before and after the fault.

[0076] Step 3: For the inner-loop current closed-loop transfer function obtained in Step 2, use the traditional linear active disturbance rejection control theory to design an extended state observer to estimate the total internal and external disturbance f of the system. The detailed design process is as follows.

[0077] According to the derivation of the closed-loop transfer function of the current inner loop of the converter in step 2, a unified expression form for the outer loop controller can be obtained:

[0078]

[0079] Define the output of the converter in the healthy state as y = v c and the output under fault reconstruction as y = v o , and the control variable as u = I Lref . Therefore, the following expression can be obtained:

[0080]

[0081] where, when the converter is in the healthy state, b = b1, y = v o ; when the converter is in the fault state, b = b2, y = v c . b represents the system parameter of the converter, and f represents the total disturbance of the converter system.

[0082] To facilitate the design of the controller, equation (7) is transformed into the state-space form. Define the variables of this state space as x = [x1 x2] T = [y f], so equation (7) can be rewritten as:

[0083]

[0084] where h is the first derivative of the total system disturbance, which is an unknown but bounded variable.

[0085] To construct a suitable linear active disturbance rejection controller, a linear extended state observer can be designed to observe the total system disturbance and state variables in real time:

[0086]

[0087] where z1 and z2 are the observed variables of variables y and f respectively.

[0088] From equation (9), the characteristic equation of the linear extended state observer can be obtained as:

[0089] p(s) = s 2 + β1s + β2 (10)

[0090] To make both characteristic roots of the observer characteristic equation located at -ω0 (ω0 is the bandwidth value of the linear extended state observer), the values of the observer parameters β1 and β2 can be set as:

[0091]

[0092] To more intuitively design an appropriate bandwidth value for the linear extended state observer, Equation (9) can be rewritten in the z-domain form as follows:

[0093]

[0094] where, e rr (k) = y(k) - z1(k), β 1s = T s β1, β 2s = T s β2. β 1s and β 2s The values of β and β will affect the distribution of the system's closed-loop poles, and thus affect the stability of the system. Therefore, these two parameters need to be carefully designed to meet the system stability and the overall controller effect. According to Equation (12), the transfer function of the extended state observer can be obtained as follows:

[0095]

[0096] The characteristic equation of Equation (13) is:

[0097] z 2 + (β 1s - 2)z + 1 - β 1s + β 2s T s = 0 (14) Considering β 1s = 2ω0T s , Therefore, the characteristic roots of the transfer function G(z) can be solved as follows:

[0098] z 1,2 = 1 - ω0T s (15) Therefore, the bandwidth of the linear extended state observer can be obtained as follows:

[0099]

[0100] At this time, the bandwidth value of the system is determined by analyzing the unit circle effect in the z-domain. So far, the unknown disturbance f in Equation (7) can be well estimated and effectively suppressed by the method of feedforward compensation.

[0101] Step 4: For the observed value z2 of the total disturbance f of the system obtained in Step 3, design a proportional controller according to the pre-selected control bandwidth to complete the design process of the entire traditional active disturbance rejection controller. The detailed design process is as follows.

[0102] According to Equation (7) and the observed value z2 of f obtained in Step 3, the controller expression of the outer voltage loop of the converter can be defined as:

[0103]

[0104] where u0 is the feedback control mechanism introduced by the system to ensure the closed-loop stability of the linear active disturbance rejection controller. It is assumed that the feedback controller of the outer voltage loop of the converter is only a proportional link, that is:

[0105] u0 = k p (V ref - z1) ≈ k p (V ref - v o ) (18)

[0106] where V ref is the reference voltage of the converter. Thus, when the linear extended state observer can track the total disturbance of the system well, Equation (7) can be simplified as:

[0107]

[0108] where k p is the gain of the proportional control term, and its value can be determined by the predefined controller bandwidth of the system.

[0109] Since whether the converter is in a healthy state or a fault state, y = v o is satisfied. Therefore, it can be concluded that under the action of the designed linear active disturbance rejection controller of the outer voltage loop, the output voltage of the converter can always achieve global asymptotic stability regardless of whether there is a fault or not.

[0110] Step 5: Based on the immersion and invariance theory, online real-time estimation of the system gain parameter b is realized on the basis of the traditional linear active disturbance rejection controller, and the design process of the proposed robust adaptive disturbance rejection control method is completed. The detailed design process is as follows.

[0111] Let the estimated value of the system parameter b be b I + b p (y), then the error between the estimated value and the true value can be solved as:

[0112] ξ = b I + b p (y) - b (20)

[0113] When the error ξ between the estimated value of the parameter b and its true value converges to zero, the estimated value can be directly used in the design process of the linear active disturbance rejection controller.

[0114] Taking the derivative of both sides of Equation (20) respectively, we have:

[0115]

[0116] If we take

[0117]

[0118] then Equation (21) can be further simplified to:

[0119]

[0120] To ensure that the estimation error ξ can achieve asymptotic convergence globally, b p (y) is defined as follows:

[0121] b p = ρy (24)

[0122] where ρ is a constant greater than zero, representing the convergence speed of the estimation error ξ.

[0123] Substituting the value of b p (y) into Equation (22), the dynamic equation of the observer can be obtained as:

[0124]

[0125] So far, the present invention has solved the problems that it is difficult to accurately estimate the system parameter b in the traditional linear active disturbance rejection controller, the online identification is complex and lacks generality. By using the immersion and invariance theory, the real-time online estimation of the parameter b can be realized based on Equation (7). The detailed design block diagram is as shown in Figure 2 as shown. This estimation method does not depend on the system model and does not rely on a large amount of data. The parameter adjustment is simple and practical, and has wide generality.

[0126] The effect of the present invention is further illustrated by the following experiments.

[0127] (1) Experimental conditions

[0128] The robust adaptive disturbance rejection control method for DC converters based on the immersion and invariance theory proposed by the present invention will be applied to a typical input parallel output series DC boost converter. The topology structure diagram of this converter is as shown in Figure 1 as shown. The main parameters of the converter in the experiment are: inductance L1 = L2 = 1000 μH, capacitance C1 = C2 = 470 μF, output voltage V in = 48 V and switching frequency 25 kHz. The proposed control method will run in dSPACE 1007. The experimental platform diagram is as shown in Figure 2 as shown.

[0129] (2) Experimental content

[0130] Taking a typical input-parallel output-series DC boost converter as the test object, the advantages of the proposed robust adaptive disturbance rejection control method for DC converters based on the immersion and invariance theory compared with the traditional linear active disturbance rejection controller are verified. It mainly includes five experimental contents: the input voltage step disturbance of the converter in the healthy state, the input voltage step disturbance of the converter in the faulty state, the load current step disturbance of the converter in the healthy state, the load current step disturbance of the converter in the faulty state, and the verification of the disturbance rejection ability of the converter at the moment of fault.

[0131] Experimental content 1:

[0132] In the healthy state, the boost ability of the converter is higher. In this invention, the input voltage is designed to step periodically between 6V and 14V, with each voltage lasting for 500ms and the period being 1s. At this time, the output reference voltage is still 48V, and the load demand current is set to 1A.

[0133] Under the healthy state of the converter, the experimental results corresponding to the traditional linear active disturbance rejection controller are as Figure 4 shown in (a). It can be seen from the figure that the traditional linear active disturbance rejection controller can ensure that the output voltage is maintained at the reference voltage of 48V when the input voltage is stable. When the input voltage steps from 6V to 14V, the output voltage increases, with the highest voltage being 52.8V and the recovery time being 66.5ms; when the input voltage steps from 14V to 6V, the output voltage decreases, with the lowest voltage being 45.6V and the recovery time being 45.6ms.

[0134] Under the healthy state of the converter, the experimental results corresponding to the proposed controller are as Figure 4 shown in (b). It can be seen from this that under the given input voltage step disturbance, the proposed adaptive disturbance rejection controller can also ensure that the output voltage is maintained at the reference voltage of 48V when the input voltage is stable, and due to the dynamic change of b (fluctuating from 488 to 872 with a period of 1s), its voltage fluctuation amplitude is smaller. When the load current jumps from 2A to 1A, the output voltage increases, with the highest voltage being 50V and the recovery time being 31.1ms; when the load current jumps from 1A to 2A, the output voltage decreases, with the lowest voltage being 46.8V and the recovery time being 39.4ms.

[0135] Experimental content 2:

[0136] After the converter fails, its boost ability decreases. At this time, the input voltage step disturbance is set as a periodic disturbance between 10V and 18V, with each voltage level lasting for 500ms and the period being 1s. The reference value of the output voltage is still set to 48V, and the load demand current is 1A.

[0137] Under the converter fault condition, the experimental results corresponding to the traditional linear active disturbance rejection controller are as follows Figure 5 As shown in (a), it can be seen from the figure that under the given input voltage step disturbance, the traditional linear active disturbance rejection controller can ensure that the output voltage is maintained at the reference voltage of 48V when the input voltage is stable. When the input voltage jumps from 10V to 18V stepwise, the output voltage increases, and the highest voltage is 51.6V, and the recovery time is 47.1ms; when the input voltage jumps from 18V to 10V stepwise, the output voltage decreases, and the lowest voltage is 45.6V, and the recovery time is 62.2ms.

[0138] Under the converter fault condition, the experimental results corresponding to the proposed controller are as follows Figure 5 As shown in (b), it can be seen from it that under the given input voltage step disturbance, the proposed adaptive disturbance rejection controller can also ensure that the output voltage is maintained at the reference voltage of 48V when the input voltage is stable, and due to the dynamic change of b (fluctuating from 408 to 632 with a period of 1s), its voltage fluctuation amplitude is smaller. When the input voltage jumps from 10V to 18V, the output voltage increases, and the highest voltage is 49.2V, and the recovery time is negligible; when the input voltage jumps from 18V to 10V, the output voltage decreases, and the lowest voltage is 47.6V, and the recovery time is also negligible.

[0139] Table 1 Analysis of experimental results of input voltage step disturbance

[0140]

[0141] According to the analysis of Experimental Content 1 and Experimental Content 2, it can be seen that both controllers can ensure that the output voltage of the converter is maintained near the reference voltage of 48V under the input voltage step disturbance, whether in the fault or healthy state. The specific differences are summarized in Table 1. Compared with the traditional controller, the proposed method shows stronger disturbance rejection ability, smaller voltage difference of the output voltage, and shorter recovery time.

[0142] Experimental Content 3:

[0143] Under the load current step disturbance, whether the converter is in the healthy or fault state, the load current is set to I o = 1A → 2A → 1A, and the input voltage at this time is 12V, and the output voltage reference value is 48V.

[0144] Under the healthy state of the converter, the experimental results corresponding to the traditional linear active disturbance rejection controller are as follows Figure 6As shown in (a), it can be seen from the figure that under a given step disturbance of the load current, the traditional linear active disturbance rejection controller can ensure that the output voltage is maintained near the reference voltage of 48V. The maximum voltage is 54.4V, and the recovery time is 37.9ms; the minimum voltage is 42.4V, the recovery time is 35.7ms, and the fluctuation period is 1s.

[0145] Under the healthy state of the converter, the experimental results corresponding to the proposed controller are as Figure 6 shown in (b). It can be seen from this that under a given step disturbance of the load current, the proposed adaptive disturbance rejection controller maintains the output bus voltage near 48V. And due to the dynamic change of b (fluctuating from 488 to 760 with a period of 1s), the voltage fluctuation amplitude is smaller. The maximum voltage is 50.8V, and the recovery time is 19.8ms; the minimum voltage is 45.2V, and the recovery time is 20.7ms.

[0146] Experiment Content 4:

[0147] Under the faulty state of the converter, the experimental results corresponding to the traditional linear active disturbance rejection controller are as Figure 7 shown in (a). It can be seen from the figure that under a given step disturbance of the load current, the traditional linear active disturbance rejection controller can ensure that the output voltage fluctuates slightly with 48V as the reference. Among them, the maximum voltage is 56.8V, and the recovery time is about 81.9ms; the minimum voltage is 40.4V, and the recovery time is 41.6ms.

[0148] Under the faulty state of the converter, the experimental results corresponding to the proposed controller are as Figure 7 shown in (b). It can be seen from this that under a given step disturbance of the load current, due to the dynamic change of b (fluctuating from 256 to 456 with a period of 1s), the output voltage of the proposed adaptive disturbance rejection controller is maintained near 48V with a smaller fluctuation amplitude. The maximum voltage is 50.1V, and the recovery time is 14.4ms; the minimum voltage is 46.1V, and the recovery time is 20.7ms.

[0149] Table 2 Analysis of Experimental Results of Load Current Step Disturbance

[0150]

[0151] According to the analysis of Experiment 3 and Experiment 4, it can be seen that whether in the fault state or in the healthy state, both controllers can ensure that when the converter is under a step disturbance of the output load, its output voltage can always fluctuate near the reference voltage of 48V. The specific differences are summarized in Table 2. Compared with the traditional controller, the proposed method shows stronger disturbance rejection ability, with a smaller voltage difference of the output voltage and a shorter recovery time, which reflects that the proposed method has stronger robustness than the traditional linear active disturbance rejection controller.

[0152] Experiment 5:

[0153] When an open-circuit fault occurs in the main switch device of the power converter, due to the reconstruction effect, both the total equivalent input inductance and the equivalent output capacitance of the converter will suddenly drop by half in value, and the system parameter b for the linear active disturbance rejection controller will also change suddenly. These changes have a severe impact on the dynamic effect of the system control. Here, taking the open-circuit fault of power switch S2 as an example, the proposed method is tested for its ability to improve the system response at the moment of fault. At the moment of fault, the input voltage of the converter is 12V, the reference voltage is 48V, and the load demand current is set to 1A.

[0154] At the moment of the converter fault, the experimental results corresponding to the traditional linear active disturbance rejection controller are as Figure 8 shown in (a). It can be seen from the figure that when an open-circuit fault suddenly occurs in the converter, the output bus voltage drops rapidly from 48V to 41.6V, and it takes 28.3ms to recover to the target voltage value.

[0155] Correspondingly, at the moment of the converter fault, the experimental results corresponding to the proposed controller are as Figure 8 shown in (b). It can be seen from the figure that when an open-circuit fault suddenly occurs in the converter, the output bus voltage drops suddenly from 48V to 42.4V, and it takes 7.5ms to recover to the target voltage value.

[0156] The above also verifies that for the external disturbance at the moment of fault, the proposed controller has stronger robustness than the traditional disturbance rejection controller.

[0157] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present invention.

Claims

1. A robust adaptive disturbance rejection control method for DC converters based on immersion and invariant theory, characterized in that The steps are as follows: Step 1: Construct the state - space equations of the DC - DC converter in the healthy state and the switch - tube fault state respectively; Step 2: Design a cascaded control structure based on the voltage outer - loop and current inner - loop for the DC - DC converter. First, use the sliding - mode control theory to design the current inner - loop, and give the closed - loop transfer function of the converter current inner - loop by using the concept of ideal sliding - mode linearization; Step 3: For the closed - loop transfer function of the current inner - loop obtained in Step 2, use the traditional linear active disturbance rejection control theory to design an extended state observer to estimate the total disturbance inside and outside the system; Step 4: For the total disturbance of the system obtained in Step 3, design a proportional controller according to the pre - selected control bandwidth to complete the design process of the whole traditional active disturbance rejection controller; Step 5: On the basis of the traditional linear active disturbance rejection controller, realize the online real - time estimation of the system gain parameters based on the immersion and invariance theory to complete the design process of the proposed robust adaptive disturbance rejection control method.

2. The robust adaptive disturbance rejection control method for a DC converter based on the immersion and invariance theory according to claim 1, wherein: The state - space equations described in Step 1: When the converter is in the healthy state: Among them, L1 and L2 are the input inductance values of the two-phase circuit; R L1 , R L2 are the parasitic resistances of the two input inductors respectively; C1 and C2 are the output capacitance values of the two circuits respectively; i L1 , i L2 are the currents flowing through the two inductors respectively; v c1 , v c2 are the voltages across the two output capacitors respectively; L1 = L2 = L, C1 = C2 = C, R L1 = R L2 = r, v c1 = v c2 = v c , i L1 = i L2 = i L ; d H is the duty cycle definition in the healthy state, and R refers to the equivalent load resistance; When the converter is in the fault state: where d F represents the duty cycle of the converter after fault reconstruction, v o = v c1 + v c2 ; 3. The robust adaptive disturbance rejection control method for a DC converter based on immersion and invariant theory according to claim 2, characterized in that: The closed - loop transfer function of the inner - loop described in Step 2: When the converter is in the healthy state: Among them, V in is the input current, and I Lref is the reference current; When the converter is in the fault state: Among them, 4. The robust adaptive disturbance rejection control method for a DC converter based on the immersion and invariant theory according to claim 3, characterized in that: The total disturbance inside and outside the system described in Step 3 is designed as follows: Among them, when the converter is in a healthy state, b = b1, y = v o ; when the converter is in a faulty state, b = b2, y = v c ; b represents the system parameter of the converter, and f represents the total disturbance of the converter system.

5. The robust adaptive disturbance rejection control method for a DC converter based on the immersion and invariant theory according to claim 4, characterized in that: The proportional controller described in Step 4: where k p is the proportional control term gain, V ref is the converter reference voltage, and z1 is the observed variable of variable y.

Images