Model-free robust predictive control method for dual-active bridge DC-DC converter

Abstract

The invention discloses a model-free robust predictive control method for a dual-active-bridge DC-DC converter, and belongs to the technical field of converters. The method comprises the following steps: constructing an autoregressive moving average model and determining an order; designing an extended state observer and carrying out gain selection and discretization; identifying and updating parameters of the autoregressive moving average model on line, then generating a control signal, and filtering the control signal; and calculating and constraining a phase shift angle. According to the method, a dead-beat prediction control strategy based on data driving is provided, the dual-active bridge DC-DC converter is estimated by using an autoregression moving average model, and model parameters are identified online through a least square method with a forgetting factor, so that robustness improvement of converter output voltage control under any disturbance is realized.

Description

Technical Field

[0001] The invention relates to a model-free robust predictive control method for a dual active bridge DC-DC converter, belonging to the technical field of converters. Background Art

[0002] As an important technology in the field of power electronics, the precise control of the output voltage of the dual active bridge DC-DC converter is crucial for the stable operation of the system. The working principle of this type of converter is based on the prediction principle, that is, the predictive controller predicts and determines the next switching state based on the current sampling value and the known discrete model, thereby achieving dynamic multi-objective optimization. However, due to various factors such as model inaccuracy, ignored dead time and parameter errors, the performance of the controller is often reduced, resulting in an increase in the steady-state error of the reference voltage, and the dynamic performance is also significantly impaired.

[0003] Existing technologies usually use model-based solutions for compensation. Among them, the disturbance observer (DOB)-based method estimates the deviation between the nominal model and the actual controlled object by designing a special observer, and uses this estimation result for system compensation to reduce the error caused by parameter mismatch. For example, the use of extended state observer (ESO) for lumped disturbance estimation is an effective DOB method.

[0004] On the other hand, the parameter identification-based method calculates the parameters accurately online and updates the nominal model in the controller in real time, thereby eliminating the impact of parameter mismatch on system performance. This method, such as the recursive least squares (RLS) method, can identify key parameters such as transformer leakage inductance online and significantly improve control accuracy.

[0005] However, although the above methods have alleviated the problems caused by parameter mismatch to a certain extent, their anti-disturbance performance is still affected by the use of parameter models. Model-based compensation technology is difficult to completely eliminate all disturbances, especially in complex and changeable actual working conditions, and its effect is often limited.

[0006] Therefore, how to further improve the output voltage control accuracy of the dual active bridge DC-DC converter, especially in the presence of parameter errors and model inaccuracies, has become a technical problem that needs to be solved urgently. Summary of the invention

[0007] In order to solve the problems existing in the background technology, the present invention provides a model-free robust predictive control method for a dual active bridge DC-DC converter.

[0008] To achieve the above object, the present invention adopts the following technical solution: a model-free robust predictive control method for a dual active bridge DC-DC converter, the method comprising the following steps:

[0009] S1: Construct an autoregressive moving average model and design a data-driven model for a dual active bridge DC-DC converter using the autoregressive moving average model;

[0010] The autoregressive moving average model described in S1 is as follows:

[0011]

[0012] In formula (1):

[0013] p represents the order of the autoregressive part of the model;

[0014] q represents the order of the sliding average part of the model;

[0015] θ x and φ y (x∈[0,1,...,q],y∈[1,...,p]) all represent the coefficients of the autoregressive moving average model;

[0016] ε represents the output voltage following error;

[0017] Represents the output voltage following error at time t=k+1, where: represents the observed value of the output voltage at time t = k + 1, represents the reference value of the output voltage at time t=k+1;

[0018] represents the control signal calculated at time t=k-1 and applied to the converter at time t=k.

[0019] S2: Determine the order of the autoregressive moving average model;

[0020] S201: Determine the order q of the sliding average part of the model;

[0021] S202: Determine the order p of the autoregressive part of the model:

[0022] When the partial autocorrelation coefficient between the output voltage following errors ε(k) and ε(kp) at time k and time kp in the time series is less than the threshold value, it is considered that the signals corresponding to time k and time kp are not correlated. The time difference between time k and time kp is the value of the order p of the autoregressive part of the model.

[0023] S3: Design an extended state observer to suppress noise in the sampling process, reduce the order of the data-driven model in the one-step delay compensation design, and estimate the accurate predicted value of the state variable at the next sampling period t=k+1, and then use the predicted value for coefficient identification of the data-driven model;

[0024] The extended state observer in S3 is as follows:

[0025]

[0026] In formula (2):

[0027] represents the first derivative of z1;

[0028] represents the first derivative of z2;

[0029] Represents the output voltage U o Observed value of

[0030] z2 represents the observed value of the system lumped disturbance;

[0031] Represents input gain, where: n represents the primary-to-secondary turns ratio of the high-frequency transformer, U in represents the input voltage, f s Represents the control frequency, L k Represents the leakage inductance of the high-frequency transformer, C o represents the output capacitance;

[0032] Both represent the extended state observer gain, where: ω n represents an adjustable parameter that determines the bandwidth of the extended state observer.

[0033] S4: Gain selection and discretization of the extended state observer;

[0034] S401: Using control cycle T s Discretizing equation (2), we can get:

[0035]

[0036] S402: Calculate the observed value of the output voltage at time t=k+1 and the observed value of the system lumped disturbance. S5: Online identification and update of autoregressive moving average model parameters;

[0037] The algorithm of online identification described in S5 is as follows:

[0038]

[0039] In formula (4):

[0040] represents the parameter vector;

[0041] represents the regression vector at time t = k;

[0042] represents the predicted output voltage following error at time t=k+1; P and K both represent intermediate variables;

[0043] λ represents the forgetting factor, 0≤λ≤1.

[0044] I represents the identity matrix.

[0045] S6: Generate a control signal based on the online identification result;

[0046] The generation formula of the control signal S6 is as follows:

[0047]

[0048] S7: The control signal is filtered before being input into the converter;

[0049] The filtering formula described in S7 is as follows:

[0050]

[0051] In formula (6):

[0052] represents the filtered control signal;

[0053] k LPF Represents the adjustable filter coefficient.

[0054] S8: Calculate and constrain the phase shift angle.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The present invention proposes a deadbeat predictive control strategy based on data drive, uses an autoregressive moving average model to estimate the dual active bridge DC-DC converter, and uses the least square method with a forgetting factor to identify the model parameters online, thereby achieving the improvement of the robustness of the converter output voltage control under any disturbance. It gets rid of the dependence of the traditional method on the converter parameters, and performs online learning and adaptive adjustment based on the output voltage error and control quantity output of the DAB converter at the historical moment. The converged model does not depend on any parameters of the DAB converter, and can achieve robust zero steady-state error voltage control. Even in the face of variable operating conditions and disturbance changes, it can maintain stable output voltage control, which has broad application prospects and important practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a control block diagram of the present invention;

[0058] Figure 2It is a schematic diagram of the waveform change of the parameters of the autoregressive sliding average model when the reference voltage changes stepwise from 80V to 60V;

[0059] Figure 3 It is a comparison diagram of experimental waveforms when the load resistance is reduced from 25Ω to 10Ω using the traditional PI control strategy and the present invention without parameter errors; the four waveforms from top to bottom in the figure respectively represent the DAB converter input voltage, the DAB converter port output voltage, the DAB converter port output current and the high-frequency transformer internal current, and the load resistance switching occurs at the step increase of the current waveform in the figure;

[0060] Figure 4 It is a schematic diagram of experimental waveforms when the load resistance is reduced from 25Ω to 10Ω in the presence of inductance error using the present invention. DETAILED DESCRIPTION

[0061] The technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0062] The method is applied to a dual active bridge DC-DC converter topology and adopts a dual phase shift strategy for control, that is, changing the phase shift angles between different bridge arms in two H bridges and between two different H bridges.

[0063] A dual active bridge DC-DC converter (DAB) model-free robust predictive control method of the present invention comprises the following steps:

[0064] S1: Construct an autoregressive moving average model (ARMA) with a suitable order and design a data-driven model of a dual active bridge DC-DC converter through the ARMA model;

[0065] The autoregressive moving average model described in S1 is as follows:

[0066]

[0067] In formula (1):

[0068] p represents the order of the autoregressive part (AR) of the model;

[0069] q represents the order of the moving average part (MA) of the model;

[0070] θ x and φ y(x∈[0,1,...,q],y∈[1,...,p]) all represent the coefficients of the autoregressive moving average model;

[0071] ε represents the output voltage following error;

[0072] Represents the output voltage following error at time t=k+1, where: represents the observed value of the output voltage at time t = k + 1, represents the reference value of the output voltage at time t=k+1;

[0073] represents the control signal calculated at time t=k-1 and applied to the converter at time t=k.

[0074] S2: Determine the order of the autoregressive moving average model;

[0075] S201: determining the order q of the sliding average part of the model according to the mathematical model of the DAB converter considering delay compensation;

[0076] Take the autoregressive moving average model constructed with the autoregressive model order p=5 and the moving average model order q=2 as an example.

[0077] Since the model-free predictive control follows the deadbeat principle, the output voltage U at the current moment o (k) is determined by the control voltage calculated in the two-step control cycle two steps before the current moment, so q=2.

[0078] S202: Determine the order p of the autoregressive part of the model:

[0079] Since the partial autocorrelation coefficient value represents the correlation between the output voltage following errors ε(k) and ε(kp) at time k and time kp, when the partial autocorrelation coefficient value between the output voltage following errors ε(k) and ε(kp) at time k and time kp in the time series is less than the threshold value (such as 0.2), it is considered that the signals corresponding to time k and time kp are not correlated, and the time difference between time k and time kp is the value of the order p of the autoregressive part of the model.

[0080] S3: Design an extended state observer (ESO) to suppress noise in the sampling process, reduce the order of the data-driven model in the one-step delay compensation design, and estimate the accurate predicted value of the state variable at the next sampling period t=k+1, and then use the predicted value for coefficient identification of the data-driven model;

[0081] The extended state observer in S3 is as follows:

[0082]

[0083] In formula (2):

[0084] represents the first derivative of z1;

[0085] represents the first derivative of z2;

[0086] Represents the output voltage U o Observed value of

[0087] z2 represents the observed value of the system lumped disturbance;

[0088] Represents input gain, where: n represents the primary-to-secondary turns ratio of the high-frequency transformer, U in represents the input voltage, f s Represents the control frequency, L k Represents the leakage inductance of the high-frequency transformer, C o represents the output capacitance;

[0089] Both represent the extended state observer gain, where: ω n represents an adjustable parameter that determines the bandwidth of the extended state observer.

[0090] S4: Gain selection and discretization of the extended state observer;

[0091] S401: Using control cycle T s Discretize equation (2). In each control cycle, firstly, according to the output voltage U at time t=k, o (k) The output voltage observation z1(k+1) at the next moment is calculated by the extended state observer. It can be obtained:

[0092]

[0093] S402: Calculate the observed value of the output voltage at time t=k+1 and the observed value of the system lumped disturbance.

[0094] S5: Use the least square method with forgetting factor to online identify and update the parameters of the autoregressive moving average model;

[0095] The algorithm of online identification described in S5 is as follows:

[0096]

[0097] In formula (4):

[0098] represents the parameter vector;

[0099] represents the regression vector at time t = k;

[0100] represents the predicted output voltage following error at time t=k+1;

[0101] P and K both represent intermediate variables;

[0102] λ represents the forgetting factor, 0≤λ≤1.

[0103] I represents the identity matrix.

[0104] S6: Generate a control signal based on the online identification result;

[0105] The generation formula of the control signal S6 is as follows:

[0106]

[0107] S7: The control signal is filtered before being input to the converter to make the control smoother, suppress the jitter of the output voltage, and improve the voltage control performance and robustness;

[0108] According to the observed value of the output voltage at time t = k + 1 and the predicted value of the output voltage z1(k+1), calculate the corresponding control signal considering one-step delay compensation The control signal is applied to the converter after passing through the minimum current stress dual phase shift modulation module to achieve accurate and robust voltage control.

[0109] The filtering formula described in S7 is as follows:

[0110]

[0111] In formula (6):

[0112] represents the filtered control signal;

[0113] k LPF Represents the adjustable filter coefficient.

[0114] S8: Calculate and constrain the phase shift angle.

[0115] In order to realize the control of the DAB converter, the control signal should be modulated into an inner phase shift angle D1 and an outer phase shift angle D2, and its generation method can be calculated according to the existing minimum current stress double phase shift modulation strategy. In addition, in order to prevent damage to the DAB converter, the following constraints should be imposed on the output of the phase shift angle to limit the output range:

[0116]

[0117] In order to ensure the fitting accuracy of the model, a model order selection method based on the inherent characteristics of the DAB converter system and time series analysis is adopted. While maintaining the inherent fast dynamic characteristics of predictive control, the dependence on model parameters is completely eliminated.

[0118] like Figure 2 As shown in the figure, when the operating conditions of the dual active bridge DC-DC converter change, the autoregressive moving average model parameters will adaptively follow the new parameter values ​​and have a faster convergence speed.

[0119] like Figure 3 As shown, although both the traditional PI control strategy and the method of the present invention can achieve zero steady-state error output voltage following, the transient recovery time of the present invention is 4.1ms, which has the characteristics of predictive control fast dynamics. In contrast, the transient recovery time of the traditional PI control strategy is 20.5ms, which proves the superiority of the method of the present invention.

[0120] like Figure 4 As shown, the present invention can completely eliminate errors, has good steady-state and dynamic following performance, a transient recovery time of 4.2ms, and is completely unaffected by changes or inaccuracies in converter parameters, and the robustness of the system is significantly improved.

[0121] The present invention only requires the dual active bridge DC-DC converter to have a voltage sensor, and does not need a current sensor like in the existing method. In addition, the robustness of the output voltage control of the predictive controller can be improved without additional hardware design, thereby reducing hardware costs.

[0122] The algorithm can be implemented by programming in a digital control chip (DSP) and can be applied to a general dual active bridge DC-DC converter controller, and has extremely high application value and economic value.

[0123] The information required by the algorithm of the present invention is only the current voltage of the converter, and the controller does not need to measure any converter parameters in advance.

[0124] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations within the meaning and range of equivalents of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0125] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A model-free robust predictive control method for a dual active bridge DC-DC converter, characterized in that: The method comprises the following steps: S1: Construct an autoregressive moving average model and design a data-driven model of a dual active bridge DC-DC converter through the autoregressive moving average model; S2: Determine the order of the autoregressive moving average model; S3: Design an extended state observer to suppress noise in the sampling process, reduce the order of the data-driven model in the one-step delay compensation design, and estimate the accurate predicted value of the state variable at the next sampling period t=k+1, and then use the predicted value for coefficient identification of the data-driven model; S4: Gain selection and discretization of the extended state observer; S5: Online identification and updating of autoregressive moving average model parameters; S6: Generate a control signal based on the online identification result; S7: The control signal is filtered before being input into the converter; S8: Calculate and constrain the phase shift angle.

2. The method for model-free robust predictive control of a dual active bridge DC-DC converter according to claim 1, characterized in that: The autoregressive moving average model described in S1 is as follows: In formula (1): p represents the order of the autoregressive part of the model; q represents the order of the sliding average part of the model; θ x and φ y (x∈[0,1,...,q],y∈[1,...,p]) all represent the coefficients of the autoregressive moving average model; ε represents the output voltage following error; Represents the output voltage following error at time t=k+1, where: represents the observed value of the output voltage at time t = k + 1, represents the reference value of the output voltage at time t=k+1; represents the control signal calculated at time t=k-1 and applied to the converter at time t=k.

3. The method for model-free robust predictive control of a dual active bridge DC-DC converter according to claim 2, characterized in that: The S2 comprises the following steps: S201: Determine the order q of the sliding average part of the model; S202: Determine the order p of the autoregressive part of the model: When the partial autocorrelation coefficient between the output voltage following errors ε(k) and ε(kp) at time k and time kp in the time series is less than the threshold value, it is considered that the signals corresponding to time k and time kp are not correlated. The time difference between time k and time kp is the value of the order p of the autoregressive part of the model.

4. The method for model-free robust predictive control of a dual active bridge DC-DC converter according to claim 3, characterized in that: The extended state observer in S3 is as follows: In formula (2): represents the first derivative of z1; represents the first derivative of z2; Represents the output voltage U o Observed value of z2 represents the observed value of the system lumped disturbance; Represents input gain, where: n represents the primary-to-secondary turns ratio of the high-frequency transformer, U in represents the input voltage, f s Represents the control frequency, L k Represents the leakage inductance of the high-frequency transformer, C o represents the output capacitance; Both represent the extended state observer gain, where: ω n represents an adjustable parameter that determines the bandwidth of the extended state observer.

5. The method for model-free robust predictive control of a dual active bridge DC-DC converter according to claim 4, characterized in that: The S4 comprises the following steps: S401: Using control cycle T s Discretizing equation (2), we can get: S402: Calculate the observed value of the output voltage at time t=k+1 and the observed value of the system lumped disturbance.

6. The method for model-free robust predictive control of a dual active bridge DC-DC converter according to claim 5, characterized in that: The algorithm of online identification described in S5 is as follows: In formula (4): represents the parameter vector; represents the regression vector at time t = k; represents the predicted output voltage following error at time t=k+1; P and K both represent intermediate variables; λ represents the forgetting factor, 0≤λ≤1. I represents the identity matrix.

7. The method for model-free robust predictive control of a dual active bridge DC-DC converter according to claim 6, characterized in that: The generation formula of the control signal S6 is as follows:

8. The method for model-free robust predictive control of a dual active bridge DC-DC converter according to claim 7, characterized in that: The filtering formula described in S7 is as follows: In formula (6): represents the control signal after filtering; k LPF Represents the adjustable filter coefficient.

Images