A power balancing method for multi-module DC converter based on model predictive control

Abstract

The present invention relates to a multi-module DC converter power balancing method based on model predictive control, which uses model predictive control to quickly control the output voltage of the multi-module DC converter, so as to be suitable for output voltage steady-state control under different modulation strategies, and uses proportional integral control to evenly control the input voltage of the multi-module DC converter, so as to achieve power balance between multiple modules, thereby achieving system output voltage control and power balance control of each module at the same time. The method is conducive to achieving power balance between modules of the multi-module DC converter.

Description

Technical Field

[0001] The present invention belongs to the field of multi-module DC converter controller design, and in particular relates to a multi-module DC converter power balancing method based on model predictive control. Background Art

[0002] With the booming new energy industry and the widespread adoption of devices such as battery energy storage power stations, electric vehicles, and AC / DC charging stations, new demands are being placed on DC / DC converters with high power density, high efficiency, excellent electrical isolation, and bidirectional energy transmission. The integration of a large number of new energy generation devices and microgrid systems into the power grid has led to significant changes in the grid's architecture and layout as it absorbs this energy. The deep integration of multiple voltage levels of electrical energy, various AC / DC power sources, and multiple information networks is driving the realization of an energy internet characterized by interactive supply and demand and efficient energy conversion. As a key medium for direct energy exchange with the DC grid, DC / DC converters are garnering increasing attention.

[0003] With the increasing voltage and power levels of DC microgrids, multilevel single-module DC converters are becoming widely used. However, multilevel circuits require a large number of switches, making control more complex. If a switch fails, the entire converter ceases operation, preventing energy transfer and conversion. With the development and maturity of modular converter technology, modular series-parallel DC converter systems are gaining increasing attention. For high-voltage input applications, converter modules are connected in series to divide the voltage; for high-current output applications, converter modules are connected in parallel to divide the current. Modular DC converter input and output cascade configurations include input-series-output series, input-series-output parallel, input-parallel-output series, input-parallel-output parallel, and independent input-series-output, further expanding the converter's transmission capacity. The input-parallel-output combination is one of the earliest converter combinations studied in the industry and is also gaining increasing attention in energy exchange scenarios between high-voltage and low-voltage distribution networks.

[0004] Due to the limitations of process manufacturing, it is impossible to ensure that the parameters of each module are exactly the same. Therefore, when combining modules, certain control measures need to be adopted to achieve power balance between modules to prevent power imbalance in a single module. Summary of the Invention

[0005] The object of the present invention is to provide a multi-module DC converter power balancing method based on model predictive control, which is conducive to achieving power balance between modules of the multi-module DC converter.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a multi-module DC converter power balancing method based on model predictive control, which adopts model predictive control to quickly control the output voltage of the multi-module DC converter to be suitable for output voltage steady-state control under different modulation strategies, and adopts proportional-integral control to evenly control the input voltage of the multi-module DC converter to achieve power balance among multiple modules, thereby simultaneously achieving system output voltage control and power balance control of each module.

[0007] Furthermore, the output voltage stabilization controller of each module in the multi-module DC converter is a model predictive controller, and the input voltage stabilization controller of each module in the multi-module DC converter is a proportional-integral controller, and the control quantities output by the two controllers are combined as the final control quantity of a single module.

[0008] Furthermore, the model predictive controller calculates the predicted value of the output voltage by collecting the module output voltage value, module output current value and load current value at the current moment, and then obtains the module optimal control quantity at the next sampling moment.

[0009] Furthermore, the proportional-integral controller takes the average of the system input voltage as the reference value of the input voltage of each module, thereby achieving equal distribution of the input voltage of the module; by establishing a small signal model of a single converter and a transfer function between the input voltage and the control quantity, the parameters of the proportional-integral controller are designed, thereby achieving good control effect.

[0010] Furthermore, the method is applicable to module combinations of various circuit forms including dual active bridge circuit modules, and does not limit the modulation method of the circuit modules.

[0011] Furthermore, voltage-stabilizing capacitors are connected in parallel on both the DC input side and the DC output side to achieve DC voltage stability; and by making the output-side parallel capacitors equivalent, the circuit structure and the output voltage prediction model are further simplified.

[0012] Furthermore, the method specifically comprises the following steps:

[0013] Step S1: Measure the current system input voltage U1 and the input voltage U of module i. 1i , divide the system input voltage U1 by the number of modules m to obtain the module input voltage reference value U 1ref ; The input voltage U of module i 1i and the module input voltage reference value U 1ref After the operation, the proportional integral controller is input to obtain the output control quantity d of the input voltage proportional integral controller. ip ;

[0014] Step S2: Measure the output voltage U of module i at the current moment2i (k), output current I si (k) and load current I2(k), the output voltage U2(k) and the output reference voltage U 2ref The calculation is performed to obtain the optimal step length ΔD, and then the current control quantity D i (k) based on the optimal sequence, the number of elements in the optimal sequence is N, the optimal sequence is D i [N];

[0015] Step S3: Substitute the generated control quantity optimization sequence into the corresponding power expression to obtain the transmission power sequence, and then substitute it into the current prediction formula to obtain the module output current prediction value sequence I si(k+1) [N];

[0016] Step S4: obtaining the module output current prediction value sequence I under different phase shift control amounts si(k+1) [N], optimize the N control variables to be optimized, first optimize the first element, substitute the output current prediction value into the output voltage prediction formula to obtain the module output voltage prediction value U after the next two switching cycles 2p (k+2); Due to the error in the prediction process, the converter output voltage is compensated for the error;

[0017] Step S5: Substitute the calculated output voltage prediction value under the current cycle into the cost function J, initialize the minimum value of the cost function to infinity, compare the cost function value obtained under the current cycle with the minimum cost function, and retain the minimum cost function value J. min As the optimal cost function value;

[0018] Step S6: Determine whether the current cycle is the last cycle. If so, exit the current optimization cycle. If not, enter the next cycle and obtain the cost function value under the next optimization element. When all optimization sequence element cycles are completed, the optimal cost function value is obtained and the optimal predicted voltage value U corresponding to the cost function is recorded. 2pi (k+2) and the optimal control quantity d io ;

[0019] Step S7: The obtained model predictive control optimal output control quantity d io The proportional-integral controller outputs the control quantity d ip Perform calculations to obtain the optimal phase shift control amount d of module i at the next moment i .

[0020] Compared with the prior art, the present invention has the following beneficial effects: it provides a power balancing method for a multi-module DC converter based on model predictive control, which can simultaneously realize the control of the system output voltage and the power balance between each module, so that the multi-module converter can not only achieve normal power output, but also ensure the power transmission balance between each module, and has good steady-state performance and dynamic capabilities, meeting the actual application requirements of high-power multi-module DC converters. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 1 is a structural block diagram of a multi-module dual active bridge DC converter according to an embodiment of the present invention;

[0022] Figure 2 1 is a small signal model diagram of a multi-module dual active bridge DC converter according to an embodiment of the present invention;

[0023] Figure 3 is an equivalent circuit diagram of a multi-module dual active bridge DC converter according to an embodiment of the present invention;

[0024] Figure 4 1 is a simplified circuit diagram of a multi-module dual active bridge DC converter according to an embodiment of the present invention;

[0025] Figure 5 This is a control block diagram of a power balancing method for a multi-module DC converter according to an embodiment of the present invention;

[0026] Figure 6 This is a block diagram of the implementation of the model predictive control of the output voltage of the multi-module DC converter in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0028] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0029] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0030] This embodiment provides a multi-module DC converter power balancing method based on model predictive control. The method uses model predictive control to quickly control the output voltage of the multi-module DC converter to be suitable for output voltage steady-state control under different modulation strategies. Proportional-integral control is used to evenly control the input voltage of the multi-module DC converter to achieve power balance among multiple modules, thereby simultaneously achieving system output voltage control and power balance control of each module.

[0031] Specifically, the output voltage stabilization controller for each module in a multi-module DC converter is a model predictive controller, and the input voltage stabilization controller for each module in the multi-module DC converter is a proportional-integral controller. The control variables output by the two controllers are combined to form the final control variable for each module. The model predictive controller has a fast dynamic response capability, enabling the converter to quickly achieve output voltage stabilization. By collecting the current module output voltage, module output current, and load current values, a predicted output voltage value is calculated to obtain the optimal control variable for the module at the next sampling moment. The proportional-integral controller uses the mean of the system input voltage as a reference value for the input voltage of each module, thereby achieving equal distribution of the module input voltage. By establishing a small-signal model for a single converter and a transfer function between the input voltage and the control variable, the parameters of the proportional-integral controller are designed, thereby achieving good control results.

[0032] This method is applicable to module combinations of various circuit forms, including but not limited to dual active bridge circuit modules. The method is not limited to the modulation method of the circuit module; as long as the transmission power expression is derived, the power balancing method of the present invention can be implemented.

[0033] In this embodiment, the combined DC converter with dual active bridge as the basic module used in this method is as follows: Figure 1 As shown, the circuit consists of m basic modules, with their inputs connected in series and their outputs in parallel. The series connection divides the high input voltage, significantly reducing the voltage stress on the module's switching transistors. The parallel connection generates a larger current at the output, achieving high power output.

[0034] In order to reasonably design the input voltage proportional integral controller, it is necessary to model the converter circuit for small signals. First, for the input series output parallel type DC converter circuit structure, its input end is divided by multiple modules, the output end voltage is automatically balanced, and the sum of the output currents of each module is the load current. Figure 1 , where the input side voltage is U1 and the module input voltage is U 11 , the equivalent inductance of the circuit is L eq1 , the transformer ratio is n 11 :n 21 ; The input voltage of module m is U1m , the equivalent inductance of the circuit is L eqm , the transformer ratio is n 1m :n 2m , the output voltage of the two modules is U2; the switching frequency of the converter is f s Based on the small signal modeling of the converter system with a single module, we can get the circuit small signal model diagram of m modules with input in series and output in parallel, as shown in the figure: Figure 2 shown.

[0035] according to Figure 2 The basic circuit equation relationship of the input series and output parallel circuit can be obtained:

[0036]

[0037] Since the input terminals of the multi-module converter are connected in series, when the system is in steady state, the primary input voltages of each module are equal. Since the power between each module is equal, the input voltages can also be equal. The small signal expressions of the input current and output current of module one are written as:

[0038]

[0039] where d 1o d is the phase-shift duty cycle disturbance signal output by the output voltage stabilization loop of module 1, 1p It is the output phase shift duty cycle disturbance value of the input voltage equalizing loop of module 1.

[0040] The sum of the input voltages of each module is the system input voltage. When the voltage of a module rises, the sum of the voltages of other modules will decrease according to the voltage balance relationship. Therefore, the sum of the disturbance values ​​output by the input voltage grading loops of the m modules is zero, and we have:

[0041]

[0042] Further transformation of the above formula can be obtained:

[0043]

[0044] According to the above formula, the input voltage U of module 1 11 The output voltage U2 is the controlled object, and other small signal components are set to zero. The transfer function between the input voltage, output voltage and control quantity of module 1 can be obtained as follows:

[0045]

[0046] The above small signal model and transfer function derivation are based on the theory of dual-loop voltage balancing control, which includes an input voltage balancing loop and an output voltage stabilization loop. The two voltage loops work together to maintain the stability of the output voltage and the power balance between modules. Based on the above theory, the input voltage U 1i and control quantity d ip The transfer function is designed in the frequency domain, and the parameters of the input voltage proportional-integral controller are tuned and designed to further improve the anti-interference performance of the system.

[0047] Next, we describe the design and implementation of a model predictive controller for the converter's output voltage. It's worth noting that, unlike model predictive control for a single module, connecting multiple circuits in series on the input side and in parallel on the output side changes the equivalent circuit structure. Figure 3 The equivalent circuit diagram of a multi-module dual-active bridge DC converter is shown, providing theoretical guidance for further analysis of the output voltage prediction expression in the combined mode. A single converter module is equivalent to a current source. A portion of this current flows through the output voltage-stabilizing capacitor, while the rest flows to the load. The modules are connected in parallel, and the output voltage-stabilizing capacitors of each module are also connected in parallel with each other and with the DC module. Therefore, it can be equivalent to Figure 4 The simplified circuit diagram of the multi-module combined DC converter is shown in FIG. Figure 3 Equivalent to:

[0048]

[0049] The output voltage and current equation is:

[0050]

[0051] The output voltage prediction value of the converter system at the next moment is:

[0052]

[0053] The output voltage prediction values ​​of the converter system at the next two moments are further obtained as follows:

[0054]

[0055] Where k represents the current moment, k+1 represents the next switching cycle moment, and so on.

[0056] The predicted value of the output voltage involves the parameters of m parallel modules, and the parameters are coupled with each other, which makes it difficult to implement a decentralized control strategy. Therefore, it is necessary to decouple the above output voltage prediction value expression. According to the superposition principle of the circuit, the above module can be equivalent to a circuit with two current sources acting independently. Then, the output quantities are superimposed to achieve the decoupling of the output voltage. According to this idea, the output voltage prediction disturbance value represented by formula (9) is disassembled to obtain the predicted voltage values ​​of m modules separately, and then superimposed:

[0057]

[0058] The load current is shared by m circuit modules. Based on the steady-state input voltage equalization, the system achieves power balance across all modules. Since the output voltages of the parallel modules are equal, the output load currents of each module are also equal, allowing the load current to be divided into m equal parts. Thus, the output voltages of the m parallel modules are decoupled, implementing a decentralized control strategy. The control variables between the modules do not interfere with each other, providing theoretical guidance for further system stability. The predicted output voltages of each module at the next two moments are obtained:

[0059]

[0060] According to formula (11), in order to obtain the output voltage prediction value of a single module, it is necessary to know the output current value I of a single module at the next moment. si (k+1), which can be derived from the transmission power:

[0061]

[0062] Among them, P i is the power value transmitted by module i, and U2 is the module output voltage value.

[0063] Next, we will establish an optimal set of control variables to select the appropriate phase-shift control variable, solve for the module's transmission power at different duty cycles, and then obtain the output current values ​​of different modules. This allows us to calculate the predicted output voltage of the system under the current control variable. The predicted system output voltage values ​​for different values ​​are compared with the output voltage reference value, and the control variable that closest approaches the voltage reference value is selected to achieve output voltage control.

[0064] Figure 5 The control block diagram of this method is to achieve the purpose of power sharing by realizing voltage sharing on the input side of the system. The input voltage is controlled by a proportional integral controller, where the reference value of the input voltage is the ratio of the input voltage U1 to the number of modules m. The input voltage controller of each module outputs the control component d ipThe model predictive control controls the output voltage and selects the control quantity that makes the voltage closest to the voltage reference value. The controller outputs the control component d io The final phase shift control quantity of a single module is the control component d io With the control component d ip The difference between the input voltage and the output voltage can be controlled simultaneously.

[0065] Figure 6 For the specific model predictive control optimization process of a single module, combined with the input voltage proportional integral control, the specific implementation steps of the multi-module DC converter power balancing method based on model predictive control proposed by the present invention are as follows:

[0066] Step S1: Measure the current system input voltage U1 and the input voltage U of module i 1i , divide the system input voltage by the number of modules m to get the module input voltage reference value U 1ref . The input voltage U of module i 1i and the module input voltage reference value U 1ref After the operation, the proportional integral controller is input to obtain the output control quantity d of the input voltage proportional integral controller. ip .

[0067] Step S2: Measure the output voltage U of the module at the current moment 2i (k), output current I si (k) and load current I2(k), the output voltage and the reference voltage are calculated to obtain the optimal step length ΔD, which can be used to calculate the optimal step length ΔD at the current control quantity D i (k) is used to generate an optimization sequence. The number of elements in the optimization sequence is N. The specific optimization sequence D i The sum of the [N] element set is:

[0068]

[0069] Step S3: Substitute the generated control quantity optimization sequence into the corresponding power expression. The power expression changes with the different phase shift modulation methods. After obtaining the transmission power sequence, substitute it into formula (7) to obtain the module output current prediction value sequence I si(k+1) [N].

[0070] Step S4: Obtain module output current prediction value sequence I under different phase shift control amounts si(k+1) [N], optimize the N control variables to be optimized, first optimize the first element, substitute the output current prediction value into formula (11) to obtain the module output voltage prediction value U after the next two switching cycles 2p(k+2). Since there are errors in the prediction process, the error compensation of the converter output voltage is performed according to the error iteration principle. The error compensation formula is:

[0071] U 2r (k+2)=U 2p (k+2)+γ1(U2(k-1)-U 2p (k-1))+γ2(U2(k)-U 2p (k)) (14)

[0072] Among them U 2r (k+2) is the predicted value of the output voltage after compensation, U 2p (k+2) is the output voltage prediction value before compensation. U2(k) is the sampled value of the module output voltage at the current moment, U 2p (k) is the predicted output voltage at the current moment, based on the last two switching cycles. The same applies to other representations. The two coefficients used in the output voltage correction process are γ1 and γ2. γ1 corrects the comparison between the predicted and output values ​​from the previous cycle, while γ2 compares the current output voltage with the predicted value. Together, they correct the predicted output voltage for the next two cycles.

[0073] Step S5: Substitute the calculated output voltage prediction value under the current cycle into the cost function, initialize the minimum value of the cost function to infinity, compare the cost function value obtained under the current cycle with the minimum cost function, and retain the minimum cost function value as the optimal cost function value. The cost function expression is:

[0074] J=δ1J cost1 +δ2J cost2 =δ1|U 2ref -U 2r (k+2)|+δ2|U2(k)-U 2r (k+2)| (15)

[0075] Among them, δ1 and δ2 are weight factors used for control and prediction in the cost function. δ1 is used to characterize the degree of difference between the circuit predicted output voltage and the output voltage reference value; δ2 is used to characterize the degree of difference between the circuit actual output voltage and the output voltage predicted values ​​of the next two stages.

[0076] Step S6: Determine whether the current cycle is the last cycle. If so, exit the current optimization cycle; if not, enter the next cycle and obtain the cost function value under the next optimization element. When all optimization sequence element cycles are completed, the optimal cost function value is obtained, and the optimal predicted voltage value U corresponding to the cost function is recorded. 2pi (k+2) and the optimal control quantity d io .

[0077] Step S7: The obtained model predictive control optimal output control quantity d io The proportional-integral controller outputs the control quantity d ip By performing calculations, we can obtain the phase shift control amount d of module i at the next moment. i . Thus, the multi-module DC converter power balancing method based on model predictive control proposed by the present invention is realized.

[0078] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other manner. Any person skilled in the art may utilize the above-disclosed technical content to modify or modify the present invention into equivalent embodiments. However, any simple modifications, equivalent variations, and modifications to the above embodiments that do not depart from the technical content of the present invention and are based on the technical essence of the present invention remain within the scope of protection of the present invention.

Claims

1. A multi-module DC converter power balancing method based on model predictive control, characterized in that: Model predictive control is used to quickly control the output voltage of multi-module DC converters to adapt to the steady-state control of output voltage under different modulation strategies. Proportional integral control is used to evenly control the input voltage of multi-module DC converters to achieve power balance among multiple modules, thereby achieving system output voltage control and power balance control of each module at the same time. The multi-module DC converter power balancing method based on model predictive control specifically comprises the following steps: Step S1: Measure the input voltage U1 of the current system and the input voltage U 1i , divide the system input voltage U1 by the number of modules m to get the module input voltage reference value U 1ref ; The input voltage U of module i 1i And the module input voltage reference value U 1ref After calculation, input the proportional integral controller to obtain the input voltage proportional integral controller output control quantity d ip ; Step S2: Measure the output voltage U of module i at the current moment 2i (k), output current I si (k) and load current I2(k), the output voltage U2(k) and the output reference voltage U 2ref The optimization step length ΔD is calculated, and then the current control quantity D i (k) generates an optimal sequence based on the number of elements in the optimal sequence, N, and the optimal sequence is D i [N]; Step S3: Substitute the generated control quantity optimization sequence into the corresponding power expression, and after obtaining the transmission power sequence, substitute it into the current prediction formula to obtain the module output current prediction value sequence I si(k+1) [N]; Step S4: obtaining the module output current prediction value sequence I under different phase shift control amounts si(k+1) [N], optimize the N control variables to be optimized, first optimize the first element, substitute the output current prediction value into the output voltage prediction formula to obtain the module output voltage prediction value U after the next two switching cycles 2p (k+2); Due to the error in the prediction process, the error compensation is performed on the converter output voltage; Step S5: Substitute the calculated output voltage prediction value under the current cycle into the cost function J, initialize the minimum value of the cost function to infinity, compare the cost function value obtained under the current cycle with the minimum cost function, and retain the minimum cost function value J. min As the optimal cost function value; Step S6: determine whether the current cycle is the last cycle. If so, exit the current optimization cycle; if not, enter the next cycle to obtain the cost function value of the next optimization element; when all optimization sequence element cycles are completed, obtain the optimal cost function value and record the optimal predicted voltage value U corresponding to the cost function. 2pi (k+2) and the optimal control quantity d io ; Step S7: The obtained model predictive control optimal output control quantity d io The proportional-integral controller outputs the control quantity d ip Perform calculations to obtain the optimal phase shift control amount d of module i at the next moment i .

2. The method for balancing power of a multi-module DC converter based on model predictive control according to claim 1, characterized in that: The output voltage stabilization controller of each module in the multi-module DC converter is a model predictive controller, and the input voltage stabilization controller of each module in the multi-module DC converter is a proportional integral controller. The control quantities output by the two controllers are combined as the final control quantity of a single module.

3. The method for balancing power of a multi-module DC converter based on model predictive control according to claim 2, characterized in that: The model predictive controller collects the module output voltage value, module output current value and load current value at the current moment, calculates the predicted value of the output voltage, and then obtains the optimal control value of the module at the next sampling moment.

4. The method for balancing power of a multi-module DC converter based on model predictive control according to claim 2, characterized in that: The proportional-integral controller takes the average value of the system input voltage as the reference value of the input voltage of each module, thereby realizing equal distribution of the input voltage of the module; by establishing a small signal model of a single converter and a transfer function between the input voltage and the control quantity, the parameters of the proportional-integral controller are designed, thereby achieving a good control effect.

5. The method for balancing power of a multi-module DC converter based on model predictive control according to claim 1, characterized in that: The method is applicable to module combinations of various circuit forms including dual active bridge circuit modules, and does not limit the modulation method of the circuit modules.

6. The method for balancing power of a multi-module DC converter based on model predictive control according to claim 1, characterized in that: Voltage-stabilizing capacitors are connected in parallel on both the DC input side and the DC output side to achieve DC voltage stability; the circuit structure and the output voltage prediction model are further simplified by making the output side parallel capacitor equivalent.

Images