A passive control method for DC-DC boost converter based on adaptive unscented Kalman filter

Abstract

A passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter belongs to the technical field of power systems. In order to develop a more accurate and real-time passive control method for a boost converter, the present invention includes constructing an average dynamic model of the DC-DC boost converter, then designing it into a state-space expression, and then using the Euler equation to discretize the state-space expression to obtain a discretized state-space expression; designing an adaptive unscented Kalman filter, performing two unscented transformations based on the discretized state-space expression, predicting the mean and covariance of the system state quantities, and obtaining estimated values of the input voltage and constant-power load power; constructing a passive controller; inputting the estimated values of the input voltage and constant-power load power into the passive controller, generating a PWM signal to drive the switch tube to work, and controlling the DC-DC boost converter. The present invention improves the robustness of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of power systems, and in particular relates to a passive control method for a DC-DC boost converter based on adaptive unscented Kalman filtering. Background Art

[0002] In recent years, amidst a wave of technological advancements in the power system sector, DC microgrids, a new type of integrated power system integrating renewable energy, DC loads, and energy storage systems, offer advantages over existing AC microgrids, including a simpler structure, relatively easy control, and no reactive power or phase issues. Distributed control in DC microgrids also offers a higher response speed. As one of the most critical components of a DC microgrid, the DC / DC boost converter must exhibit high bidirectional energy conversion efficiency. Therefore, it is crucial to develop novel control strategies for boost converters that minimize disturbances and rapidly regulate voltage under constant power load conditions.

[0003] Common power electronic loads in DC microgrids can be primarily categorized as resistive loads and constant-power loads. Resistive loads, due to their inherent passivity, generate positive damping inputs in DC microgrid systems. Conversely, constant-power loads, due to their constant power, exhibit negative impedance characteristics at the system input. This characteristic interacts with other converters in the system, reducing system damping, impacting system stability, and creating the risk of oscillation. With the increasing penetration of renewable energy, the use of constant-power loads also poses greater stability risks to DC microgrid systems. Power system stability analysis can be categorized into large-signal stability and small-signal stability, depending on the operating scenario and disturbance magnitude. Current research primarily focuses on small-signal stability analysis. Two main control strategies exist: passive damping, which directly improves system damping by adding passive components to the load circuit, increases system cost and weight, contradicting the trend toward lightweighting. Active damping, on the other hand, modifies the control system structure to improve system impedance. However, this approach is limited to small-signal linear models and can only be applied near the equilibrium point, resulting in a limited scope of application.

[0004] Based on the nonlinear characteristics of the power electronic converters used and the negative impedance of the constant-power load, appropriate nonlinear control strategies are needed to improve the global stability of the DC microgrid system, such as passive control, backstepping control, model predictive control, sliding mode control, and deep learning. These control strategies effectively ensure the system's large-signal stability, but nonlinear control strategies place high demands on the system model. Consequently, the system inevitably experiences steady-state deviations under random disturbances of the source and load. To reduce or eliminate this output deviation, integral compensation is added at appropriate locations, but this strategy also affects the system's dynamic response speed. Alternatively, a Kalman filter can be added to the passive control system to accurately predict and estimate load power. However, traditional and extended Kalman filters impose a high computational load on model predictive control, with the iterative calculation of the Jacobian matrix taking up a significant amount of time. Summary of the Invention

[0005] The problem to be solved by the present invention is to develop a more accurate and real-time boost converter passive control method, and propose a DC-DC boost converter passive control method based on adaptive unscented Kalman filtering.

[0006] To achieve the above object, the present invention is implemented through the following technical solutions:

[0007] A passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter comprises the following steps:

[0008] S1. Construct an average dynamic model of the DC-DC boost converter, design it into a state-space expression, and then discretize the state-space expression using the Euler equation to obtain the discretized state-space expression.

[0009] S2. Design an adaptive unscented Kalman filter and perform two unscented transformations based on the discretized state-space expression to predict the mean and covariance of the system state variables and obtain estimates of the input voltage and constant-power load power.

[0010] S3. Build a passive controller;

[0011] S4. Input the input voltage obtained in step S2 and the estimated value of the constant power load power into the passive controller constructed in step S3 to generate a PWM signal to drive the switch tube to control the DC-DC boost converter.

[0012] Furthermore, the specific implementation method of step S1 includes the following steps:

[0013] S1.1. Construct an average dynamic model of the DC-DC boost converter, expressed as:

[0014]

[0015] Among them, L, C, E, v o 、i L , μ, R, and P CPL They are the inductor value, output side capacitor value, input voltage, output voltage, inductor current, switch duty cycle, resistive load and constant power load power of the DC-DC boost converter;

[0016] Set the input voltage E and constant power load power P CPL is the estimated state variable;

[0017] S1.2. Design the average dynamic model of the DC-DC boost converter as follows:

[0018]

[0019] Among them, the state variable matrix x=[i L v0] T , S is the system matrix, B is the input matrix, d is the nonlinear term matrix, h is the output matrix, y is the output quantity, and ρ is the influence factor based on the capacitor voltage;

[0020]

[0021] S1.3. Design and add the constant power load power as one of the state variables to the state space expression. The state space expression based on the constant power load power is as follows:

[0022]

[0023] in, is the state variable matrix with constant power load power added, u(t) is the time-varying system input, f is the Euler equation of the dynamic model after adding the output power to the state variable matrix, ω is the Gaussian white noise with zero mean in the state variable matrix, and ν is the Gaussian white noise with zero mean in the output;

[0024] S1.4. Use the Euler equation to discretize the state-space expression obtained in step S1.3 to obtain the discretized state-space expressions for time k+1 and time k:

[0025]

[0026] Among them, f(x aukf,k ,u(t) k ) represents the Euler equation of the dynamic model after the output power is added to the state variable matrix at time k, T s is the sampling time.

[0027] Furthermore, the specific implementation method of step S2 includes the following steps:

[0028] S2.1. Design an adaptive unscented Kalman filter. Determine the size of the sigma point set to 2n+1 based on the dimension n of the state variable. Select the sigma point set and design the unscented transformation process as follows:

[0029]

[0030] in, is the sigma point set at time k, is the mean value of the state variable at time k, P k / k is the covariance matrix at time k, and λ represents the distance from each point in the sigma point set to the mean point;

[0031] S2.2. Based on the sigma point set selected in step S2.1, perform predictions and substitute the state space expression to calculate the state variables at time k+1. Predict the mean and covariance matrix of the state variables at time k+1, and obtain the expression:

[0032]

[0033] Among them, ω i are the weights corresponding to the mean and variance, Q is the covariance matrix of the Gaussian white noise with zero mean of the state variable matrix;

[0034] S2.3. Perform an untraceable transformation on the mean and variance of the predicted state variables at time k+1 obtained in step S2.2 to obtain the sigma point set at time k+1:

[0035]

[0036] Substitute the sigma point set at time k+1 into the observation equation to obtain the predicted observation value, where Z is the discretized observation value:

[0037]

[0038] S2.4. Calculate the mean and covariance matrix of the predicted observations based on the predicted observations:

[0039]

[0040] Among them, R is the covariance matrix of Gaussian white noise with zero mean of the output, is the mean of the observations at time k+1, is the covariance matrix of the observation at time k+1, is the covariance matrix of the state variables and observations at time k+1;

[0041] S2.5. Calculate the Kalman gain K at time k+1 k+1 By adding feedforward compensation, the mean value of the system state variable at time k+1 and the covariance matrix at time k+1 are output:

[0042]

[0043] The estimated state variables obtained by iteration are the input voltage E and the power P of the constant power load CPL CPL

[0044] Furthermore, the specific implementation method of step S3 includes the following steps:

[0045] S3.1. Based on the design of the passive controller, the average dynamic model of the DC-DC boost converter is constructed in matrix form, and the expression is obtained:

[0046]

[0047] Where H is the transient quantity in the passive controller, X is the state variable matrix in the passive controller, Γ is the input voltage matrix in the passive controller, G is the voltage transformation coefficient matrix in the passive controller, and R(x) is the system impedance;

[0048]

[0049] S3.2. Design a passive controller including the energy reshaping stage and the damping injection stage;

[0050] S3.2.1. Energy Reshaping Phase: Adoption Perform the transformation and get the expression:

[0051]

[0052] Among them, X d is the state quantity of the equilibrium point, Represents the interference value of X, where R(x m ) is the actual quantity matrix of system impedance, R(x d ) is the reference value of the system impedance at the equilibrium point;

[0053] S3.2.2. Energy reshaping stage: by injecting a virtual damping matrix R V X modifies the system dissipation function and obtains the expression:

[0054]

[0055] Among them, R p is the system impedance after injecting virtual damping, R v is the virtual damping coefficient;

[0056]

[0057] S3.2.3. To ensure that the system injected with the virtual damping matrix in step S3.2.2 satisfies Lyapunov stability, set the global stable equilibrium point of the system to Set the energy function of the system for:

[0058]

[0059] The expression of the time derivative of the energy function is:

[0060]

[0061] S3.3. Based on the global stability convergence of the system, set the injected virtual damping coefficient R v Let the matrix R p All elements are positive, and the expression of the control law of the passive controller is:

[0062]

[0063] Furthermore, the specific implementation method of step S4 is to substitute the estimated values of the input voltage and the constant power load power obtained in step S2 into the control law of the passive controller obtained in step S3 to obtain the expression:

[0064]

[0065] Among them, i L,ref 、v ref is the reference value of the state quantity;

[0066] According to the properties of the DC-DC boost converter, a PWM signal is generated to drive the switch tube to work, and the control result of the switch duty cycle μ is obtained. The expression is:

[0067]

[0068] Beneficial effects of the present invention:

[0069] The present invention describes a passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter. By designing an adaptive unscented Kalman filter, the method performs online estimation of system parameters, including output power, input voltage, and other state variables. Based on the state variables output by the filter, an adaptive passive model prediction technique is proposed for controlling a DC / DC boost converter with a constant power load. Furthermore, simple current limiting can be achieved through a saturation current limiting link prior to the current loop. This method, combined with a passive control loop, offers advantages such as a simple structure and large-signal stability, significantly improving system robustness.

[0070] The passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter, described in this invention, not only significantly improves the system's large-signal stability and robustness, but also provides precise current limiting and rapid voltage regulation, effectively ensuring the stability and reliability of the DC microgrid under various complex operating conditions. The proposed nonlinear control loop avoids the challenges of obtaining nominal parameters and performing complex multi-degree-of-freedom parameter adjustments in traditional control methods, resulting in a relatively simple and practical control structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 This is a flow chart of a passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter according to the present invention;

[0072] Figure 2 This is the block diagram of the unscented Kalman filter passive control system of the present invention;

[0073] Figure 3 This is the simulation result diagram of the load mutation of the unscented Kalman filter passive control of the present invention. DETAILED DESCRIPTION

[0074] In order to make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present invention and are not intended to limit the present invention. That is, the specific embodiments described herein are only some embodiments of the present invention, not all embodiments. Generally, the components of the specific embodiments of the present invention described and illustrated in the drawings herein can be arranged and designed in various different configurations, and the present invention can also have other embodiments.

[0075] Therefore, the following detailed description of the specific embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but is merely representative of selected specific embodiments of the present invention. All other specific embodiments obtained by those skilled in the art based on the specific embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0076] In order to further understand the content, features and effects of the present invention, the following specific embodiments are given as examples, and the attached Figure 1 -Attached Figure 3 The detailed instructions are as follows:

[0077] Example 1:

[0078] A passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter comprises the following steps:

[0079] S1. Construct an average dynamic model of the DC-DC boost converter, design it into a state-space expression, and then discretize the state-space expression using the Euler equation to obtain the discretized state-space expression.

[0080] Furthermore, the specific implementation method of step S1 includes the following steps:

[0081] S1.1. Construct an average dynamic model of the DC-DC boost converter, expressed as:

[0082]

[0083] Among them, L, C, E, v o 、i L , μ, R, and P CPL They are the inductor value, output side capacitor value, input voltage, output voltage, inductor current, switch duty cycle, resistive load and constant power load power of the DC-DC boost converter;

[0084] Set the input voltage E and constant power load power P CPL is the estimated state variable;

[0085] S1.2. Design the average dynamic model of the DC-DC boost converter as follows:

[0086]

[0087] Among them, the state variable matrix x=[i L v0] T , S is the system matrix, B is the input matrix, d is the nonlinear term matrix, h is the output matrix, y is the output quantity, and ρ is the influence factor based on the capacitor voltage;

[0088]

[0089] Based on the unscented Kalman filtering method proposed in this invention, the power of the constant power load needs to be considered as one of the state variables and added into the state space expression to obtain a new state variable matrix:

[0090]

[0091] Due to the power characteristics of the constant power load, it is easy to conclude that P is equal to 0, and the state variable matrix becomes:

[0092]

[0093] Where u(t) is the time-varying system input, and f is the relationship between u(t) and x. aukfThe function represents the Euler equation of the dynamic model after the output power is added to the state variable matrix. In addition, the output of the UKF algorithm is updated based on the change of the state variable:

[0094]

[0095] S1.3. Design and add the constant power load power as one of the state variables to the state space expression. The state space expression based on the constant power load power is as follows:

[0096]

[0097] Among them, x aukf is the state variable matrix with constant power load power added, u(t) is the time-varying system input, f is the Euler equation of the dynamic model after adding the output power to the state variable matrix, ω is the Gaussian white noise with zero mean in the state variable matrix, and ν is the Gaussian white noise with zero mean in the output;

[0098] S1.4. Use the Euler equation to discretize the state-space expression obtained in step S1.3 to obtain the discretized state-space expressions for time k+1 and time k:

[0099]

[0100] Among them, f(x aukf,k ,u(t) k ) represents the Euler equation of the dynamic model after the output power is added to the state variable matrix at time k, T s is the sampling time.

[0101] S2. Design an adaptive unscented Kalman filter and perform two unscented transformations based on the discretized state-space expression to predict the mean and covariance of the system state variables and obtain estimates of the input voltage and constant-power load power.

[0102] Furthermore, the specific implementation method of step S2 includes the following steps:

[0103] S2.1. Design an adaptive unscented Kalman filter. Determine the size of the sigma point set to 2n+1 based on the dimension n of the state variable. Select the sigma point set and design the unscented transformation process as follows:

[0104]

[0105] in, is the sigma point set at time k, is the mean value of the state variable at time k, P k / k is the covariance matrix at time k, and λ represents the distance from each point in the sigma point set to the mean point;

[0106] Furthermore, it should be noted that (n+λ)P must be a positive semidefinite matrix when given;

[0107] S2.2. Based on the sigma point set selected in step S2.1, perform predictions and substitute the state space expression to calculate the state variables at time k+1. Predict the mean and covariance matrix of the state variables at time k+1, and obtain the expression:

[0108]

[0109] Among them, ω i are the weights corresponding to the mean and variance, Q is the covariance matrix of the Gaussian white noise with zero mean of the state variable matrix;

[0110] Furthermore, we get the predicted mean and variance. It can be seen that after one iteration, the two weights are equal, where λ = α 2 (N+ε)-N, assuming the noise is still white Gaussian distributed, take N+ε=3, ε is the parameter to be selected, here it is 0, α affects the distribution of the parameter and is usually a small positive number. β is a hyperparameter that reflects the high-order state history information. For Gaussian distribution, β is 2. The weight calculation is as follows:

[0111]

[0112] S2.3. Perform an untraceable transformation on the mean and variance of the predicted state variables at time k+1 obtained in step S2.2 to obtain the sigma point set at time k+1:

[0113]

[0114] Substitute the sigma point set at time k+1 into the observation equation to obtain the predicted observation value, where Z is the discretized observation value:

[0115]

[0116] S2.4. Calculate the mean and covariance matrix of the predicted observations based on the predicted observations:

[0117]

[0118] Among them, R is the covariance matrix of Gaussian white noise with zero mean of the output, is the mean of the observations at time k+1, is the covariance matrix of the observation at time k+1, is the covariance matrix of the state variables and observations at time k+1;

[0119] S2.5. Calculate the Kalman gain K at time k+1k+1 By adding feedforward compensation, the mean value of the system state variable at time k+1 and the covariance matrix at time k+1 are output:

[0120]

[0121] The estimated state variables obtained by iteration are the input voltage E and the power P of the constant power load CPL CPL

[0122] S3. Build a passive controller;

[0123] Furthermore, the specific implementation method of step S3 includes the following steps:

[0124] S3.1. Based on the design of the passive controller, the average dynamic model of the DC-DC boost converter is constructed in matrix form, and the expression is obtained:

[0125]

[0126] Where H is the transient quantity in the passive controller, X is the state variable matrix in the passive controller, Γ is the input voltage matrix in the passive controller, G is the voltage transformation coefficient matrix in the passive controller, and R(x) is the system impedance;

[0127]

[0128] Furthermore, G = -G T , R=R T ,The matrix form of the average dynamic model of the constructed DC-DC boost converter satisfies the characteristics of the Euler-Lagrangian model;

[0129] S3.2. Design a passive controller including the energy reshaping stage and the damping injection stage;

[0130] S3.2.1. Energy Reshaping Phase: Adoption Perform the transformation and get the expression:

[0131]

[0132] Among them, X d is the state quantity of the equilibrium point, Represents the interference value of X, where R(x m ) is the actual quantity matrix of system impedance, R(x d ) is the reference value of the system impedance at the equilibrium point;

[0133] S3.2.2. Energy reshaping stage: by injecting a virtual damping matrix R V X modifies the system dissipation function and obtains the expression:

[0134]

[0135] Among them, R p is the system impedance after injecting virtual damping, R v is the virtual damping coefficient;

[0136]

[0137] S3.2.3. To ensure that the system injected with the virtual damping matrix in step S3.2.2 satisfies Lyapunov stability, set the global stable equilibrium point of the system to Set the energy function of the system for:

[0138]

[0139] The expression of the time derivative of the energy function is:

[0140]

[0141] S3.3. Based on the global stability convergence of the system, set the injected virtual damping coefficient R v Let the matrix R p All elements are positive, and the expression of the control law of the passive controller is:

[0142]

[0143] S4. Input the input voltage obtained in step S2 and the estimated value of the constant power load power into the passive controller constructed in step S3 to generate a PWM signal to drive the switch tube to control the DC-DC boost converter.

[0144] Furthermore, the specific implementation method of step S4 is to substitute the estimated values of the input voltage and the constant power load power obtained in step S2 into the control law of the passive controller obtained in step S3 to obtain the expression:

[0145]

[0146] i L,ref 、v ref It is the reference value of the state quantity.

[0147] According to the properties of the DC-DC boost converter, a PWM signal is generated to drive the switch tube to work, and the control result of the switch duty cycle μ is obtained. The expression is:

[0148]

[0149] By applying a current limiting link after the current reference, the peak current during startup and interference can be suppressed, thereby reducing the current stress of the equipment. By combining the passive control voltage loop and the unscented Kalman filter, a nonlinear controller with simple structure, high robustness and low computational burden can be formed. Combining the above controller design, the adaptive unscented Kalman filter passive control method of the present invention is finally as follows: Figure 2 shown.

[0150] In order to verify the effectiveness of the method of this embodiment, the method of this embodiment is simulated and verified, and the parameters are set as follows: converter input voltage E = 100V, L = 2mH, C = 940μF, R = 100Ω. First, the performance of the controller proposed in this embodiment is simulated and verified. The initial startup parameters of the system are input voltage 100V and constant power load 100W. The simulation waveform is as follows: Figure 3 As shown in the figure, the output voltage achieves precise convergence. At 0.2s, the constant power load power increases to 500W, at 0.6s, the constant power load decreases to 100W, and at 1.0s, the resistance R decreases to 50Ω. The simulation results show that the control method proposed in this invention can quickly stabilize the bus voltage when the load suddenly changes.

[0151] The control method proposed in this embodiment primarily combines unscented Kalman filtering technology, passive control, and model predictive control. First, based on converter information, a designed unscented Kalman filter controller accurately predicts the unknown input voltage and load power, taking into account time-varying characteristics. This predicted information is fed into the proposed passive controller, which generates a PWM signal to drive the switch, thereby controlling a boost converter with a constant-power load. The proposed method significantly improves the system's large-signal stability and robustness, and also provides precise current limiting control and rapid voltage regulation capabilities, effectively ensuring the stability and reliability of the DC microgrid under various complex operating conditions. Furthermore, the proposed nonlinear control loop avoids the challenges of obtaining nominal parameters and performing complex multi-degree-of-freedom parameter adjustments in traditional control methods, resulting in a relatively simple and practical control structure. To further evaluate the superiority of the proposed control method, simulation studies were conducted, and the results demonstrated that the control method proposed in this embodiment exhibits excellent control performance.

[0152] It should be noted that relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.

[0153] Although the present application has been described above with reference to specific embodiments, various modifications may be made thereto and components may be substituted with equivalents without departing from the scope of the present application. In particular, as long as there are no structural conflicts, the various features of the embodiments disclosed herein may be combined with each other in any manner, and the omission of an exhaustive description of these combinations in this specification is solely for the sake of space and resource conservation. Therefore, the present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions within the scope of the claims.

Claims

1. A passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter, characterized in that: The steps include: S1. Construct an average dynamic model of the DC-DC boost converter, design it into a state-space expression, and then discretize the state-space expression using the Euler equation to obtain the discretized state-space expression. The specific implementation method of step S1 includes the following steps: S1.

1. Construct an average dynamic model of the DC-DC boost converter, expressed as: Among them, L, C, E, v o 、i L , μ, R, and P CPL They are the inductor value, output side capacitor value, input voltage, output voltage, inductor current, switch duty cycle, resistive load and constant power load power of the DC-DC boost converter; Set the input voltage E and constant power load power P CPL is the estimated state variable; S1.

2. Design the average dynamic model of the DC-DC boost converter as follows: Among them, the state variable matrix x=[i L v o ] T , S is the system matrix, B is the input matrix, d is the nonlinear term matrix, h is the output matrix, y is the output quantity, and ρ is the influence factor based on the capacitor voltage; S1.

3. Design and add the constant power load power as one of the state variables to the state space expression. The state space expression based on the constant power load power is as follows: in, is the time derivative of the state variable matrix with constant load power, u(t) is the time-varying system input, f is the Euler equation of the dynamic model after adding the output power to the state variable matrix, ω is the zero-mean Gaussian white noise of the state variable matrix, and ν is the zero-mean Gaussian white noise of the output; S1.

4. Use the Euler equation to discretize the state-space expression obtained in step S1.3 to obtain the discretized state-space expressions for time k+1 and time k: Among them, f(x aukf,k ,u(t) k ) represents the Euler equation of the dynamic model after the output power is added to the state variable matrix at time k, T s is the sampling time; S2. Design an adaptive unscented Kalman filter and perform two unscented transformations based on the discretized state-space expression to predict the mean and covariance of the system state variables and obtain estimates of the input voltage and constant-power load power. S3. Build a passive controller; S4. Input the input voltage obtained in step S2 and the estimated value of the constant power load power into the passive controller constructed in step S3 to generate a PWM signal to drive the switch tube to control the DC-DC boost converter.

2. The passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter according to claim 1, characterized in that: The specific implementation method of step S2 includes the following steps: S2.

1. Design an adaptive unscented Kalman filter. Determine the size of the sigma point set to 2n+1 based on the dimension n of the state variable. Select the sigma point set and design the unscented transformation process as follows: in, is the sigma point set at time k, is the mean value of the state variable at time k, P k / k is the covariance matrix at time k, and λ represents the distance from each point in the sigma point set to the mean point; S2.

2. Based on the sigma point set selected in step S2.1, perform predictions and substitute the state space expression to calculate the state variables at time k+1. Predict the mean and covariance matrix of the state variables at time k+1, and obtain the expression: Among them, ω i are the weights corresponding to the mean and variance, Q is the covariance matrix of the Gaussian white noise with zero mean of the state variable matrix; S2.

3. Perform an untraceable transformation on the mean and variance of the predicted state variables at time k+1 obtained in step S2.2 to obtain the sigma point set at time k+1: Substitute the sigma point set at time k+1 into the observation equation to obtain the predicted observation value, where Z is the discretized observation value: S2.

4. Calculate the mean and covariance matrix of the predicted observations based on the predicted observations: Among them, R is the covariance matrix of Gaussian white noise with zero mean of the output, is the mean of the observations at time k+1, is the covariance matrix of the observation at time k+1, is the covariance matrix of the state variables and observations at time k+1; S2.

5. Calculate the Kalman gain K at time k+1 k+1 By adding feedforward compensation, the mean value of the system state variable at time k+1 and the covariance matrix at time k+1 are output: The estimated state variables obtained by iteration are the input voltage E and the power P of the constant power load CPL CPL .

3. The passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter according to claim 2, characterized in that: The specific implementation method of step S3 includes the following steps: S3.

1. Based on the design of the passive controller, the average dynamic model of the DC-DC boost converter is constructed in matrix form, and the expression is obtained: Where H is the transient quantity in the passive controller, X is the state variable matrix in the passive controller, Γ is the input voltage matrix in the passive controller, G is the voltage transformation coefficient matrix in the passive controller, and R(x) is the system impedance; S3.

2. Design a passive controller including the energy reshaping stage and the damping injection stage; S3.2.

1. Energy Reshaping Phase: Adopting Perform the transformation and get the expression: Among them, X d is the state quantity of the equilibrium point, Represents the interference value of X, where R(x m ) is the actual quantity matrix of system impedance, R(x d ) is the reference value of the system impedance at the equilibrium point; S3.2.

2. Energy reshaping stage: by injecting virtual damping matrix Modify the system dissipation function and get the expression: Among them, R p is the system impedance after injecting virtual damping, R v is the virtual damping coefficient; S3.2.

3. To ensure that the system injected with the virtual damping matrix in step S3.2.2 satisfies Lyapunov stability, set the global stable equilibrium point of the system to Set the energy function of the system for: The expression of the time derivative of the energy function is: S3.

3. Based on the global stability convergence of the system, set the injected virtual damping coefficient R v Let the matrix R p All elements are positive, and the control law of the passive controller is expressed as:

4. The passive control method for a DC-DC boost converter based on an adaptive unscented Kalman filter according to claim 3, characterized in that: The specific implementation method of step S4 is to substitute the estimated value of the input voltage and the constant power load power obtained in step S2 into the control law of the passive controller obtained in step S3 to obtain the expression: Among them, i L,ref 、v ref is the reference value of the state quantity; According to the properties of the DC-DC boost converter, a PWM signal is generated to drive the switch tube to work, and the control result of the switch duty cycle μ is obtained. The expression is:

Images