Dual-Time-Scale Optimization Control Method for a Four-Switch Buck-Boost Converter

Abstract

The present invention discloses a dual-time-scale optimization control method for a four-switch buck-boost converter in the field of buck-boost converters, including S1: obtaining the input voltage sampling value, output voltage sampling value, inductor current value during the operation of the four-switch buck-boost converter within the current switching period, and the duration T1 of the common conduction of the switching transistors Q1 and Q4, and the duration T2 of the common conduction of the switching transistors Q1 and Q3; S2: the inductor current optimizer calculates the difference in the effective values of the inductor current between the current period and the previous period to obtain the adjustment coefficient K a , K b and outputs it to the output voltage regulator; S3: the output voltage regulator combines the output voltage reference value, output voltage sampling value, and input voltage sampling value, and calculates T1 and T2 through the PI algorithm, outputs them to the inductor current optimizer, and outputs a control signal for controlling the on-off states of the switching transistors Q1 to Q4 in the next switching period; S4: repeat S1 to S3. The present invention can not only ensure the performance of the converter but also reduce the requirements of the algorithm for the controller.

Description

Technical Field

[0001] The present invention relates to the field of buck-boost converters, and specifically to a dual-time-scale optimization control method for a four-switch buck-boost converter. Background Art

[0002] In recent years, China's space industry has been developing at an accelerated pace, and the demand for power supply from space equipment such as space stations and large satellites has been increasing. The main way to obtain space energy is photovoltaic panels, and the main storage medium for space energy is a storage battery. The biggest problem in the application of photovoltaic panels and storage batteries is that the output voltage varies within a wide range during operation, which cannot meet the stability requirements of the load for the supply voltage. Therefore, high-efficiency converters applicable to wide-range input voltage variations and capable of realizing buck-boost functions have received extensive attention. Among them, a topology called a four-switch buck-boost converter has the advantages of low voltage / current stress of switching devices, few passive components, and high efficiency, and has broad application prospects.

[0003] As Figure 1 shown, the switched-capacitor buck-boost converter includes switching transistors Q1 to Q2 and an inductor L. Among them, switching transistors Q1 and Q2 form the front bridge, and switching transistors Q3 and Q4 form the rear bridge. A capacitor Co and a load RL are connected in parallel at both ends of the rear bridge. In order to give full play to the topological advantages of the four-switch buck-boost converter and improve the power density and efficiency of the power supply, scholars at home and abroad have conducted in-depth research on the control strategy of this topology in recent years. Generally, it can be divided into two categories: hard-switching multi-mode PWM modulation strategy and quadrilateral inductor current soft-switching modulation strategy. The hard-switching control timing is relatively simple, but the loss of the switching device is relatively high, and it is not suitable for high-frequency and high-power density applications. The quadrilateral soft-switching modulation divides the modulation time of the inductor current into four segments, generates a reverse current, and realizes zero-voltage switching of the four switching transistors. This type of modulation method can not only achieve zero-voltage switching of the four switching transistors but also further improve the efficiency of the converter by optimizing the effective value of the inductor current.

[0004] After obtaining the expression of I Lrms as Figure 4 shown, for different load currents, there always exists a value of T2 such that the effective value of the inductor current I Lrms is minimized. And the smaller the effective value of the inductor current I Lrms , the smaller the loss of the inductor, and the highest conversion efficiency of the four-switch buck-boost converter.

[0005] Therefore, in order to improve the efficiency of the converter, the four-switch buck-boost converter needs to adopt the method of controlling the minimum effective value of the inductor current. Theoretically, the specific values of T2 (the duration of the common conduction of the switching transistors Q1 and Q3) and T1 (the duration of the common conduction of the switching transistors Q1 and Q4) can be calculated by differentiating the formula for T2. However, this process requires a large number of multiplication, square root, and division operations. In a switching power supply, the switching frequency is usually in the range of dozens of kilohertz to hundreds of kilohertz, and each switching cycle is only dozens of microseconds or even a few microseconds. Completing a large number of multiplication, square root, and division operations within such a short time requires the controller to have extremely high digital signal processing capabilities. Therefore, low-cost controllers on the market are difficult to meet the requirements of real-time online calculation of T2 and T1.

[0006] To achieve the control of the minimum effective value of the inductor current, the currently more advanced and practical control methods mainly include the look-up table method and the critical current approximation variable frequency control method.

[0007] 1. Look-up table method

[0008] As Figure 5 shown, in the look-up table method, the values of T2 and T1 are calculated by differentiating the expression of I i with respect to T2 under different input voltages V o , output voltages V Lrms , and output currents Iout in advance, and then stored in a multi-dimensional data table.

[0009] During operation, the controller first calculates the error signal between the given value Vref of the output voltage and the output voltage V o , then uses a proportional-integral (PI) regulator to obtain the given value Iout * of the output current, and then combines the input voltage V i , output voltage V o to load the values of T2 and T1 into the PWM signal generator through look-up table, and then achieve the control of the minimum effective value of the inductor current to reduce the inductor loss and improve the efficiency of the converter.

[0010] 2. Critical current approximation variable frequency control method

[0011] As Figure 6 shown, this method realizes the control goal of the minimum effective value of the inductor current by controlling the inductor current value at the moment of T1 or T1 + T2 to be exactly equal to the critical current that satisfies the soft-switching condition of the switching device.

[0012] The main difference between this method and the look-up table method is that the output of the PI regulator is not used as the reference value of the output current, but as a time control quantity T u . T u is related to the input voltage V i, the output voltage V o are used as the inputs of the control algorithm to calculate the values of T2 and T1, and then load them into the PWM generator to generate four PWM signals to achieve the control of the converter.

[0013] The control logic flow chart of the critical current approaching variable frequency control method is as Figure 7 shown, and the waveform change diagram of the inductor current under different conditions is as Figure 8 shown.

[0014] From Figure 7 and 8 , it can be seen that this method determines the calculation methods of T1 and T2 according to the mutual relationship between the input and output voltages, and according to the relationship between T1, T2 and T3 (T3 is the duration of the common conduction of the switching tubes Q2 and Q3) and the minimum switching period T min , there are three working conditions, and their characteristics can be summarized in the following table:

[0015]

[0016] From the summary in the table, it can be seen that when the load current is small (T1 + T2 + T2 ≤ T min ), it always ensures that I1 (V i > V o ) or I2 (V i < V o ) is equal to the minimum resonance current I z required for soft switching, so as to achieve the control goal of minimizing the effective value of the inductor current. However, as the load current increases, the sum of T1, T2, and T3 will exceed T min , and at this time, by increasing T1 and keeping T2 unchanged, more energy is transferred by the inductor to meet the load demand.

[0017] The above look-up table method needs to pre-calculate the values of T2 and T1. The method is only applicable to the control mode with a fixed switching frequency and cannot meet the requirements of variable switching frequency control, resulting in a narrow operating range of the converter. At the same time, it is very difficult for the pre-calculated values of T2 and T1 to adapt to the errors caused by factors such as the operating temperature change of the converter and manufacturing parameter deviations, and there are also data holes between each data point, resulting in a decrease in control accuracy. In addition, this method needs to establish a multi-dimensional data table about the input voltage V i , the output voltage V o , and the output current I out . The multi-dimensional data table requires a large amount of memory resources, which will increase the cost and volume of the controller.

[0018] The critical current approaching variable frequency control method when T1 + T2 + T3 ≤ T minWhen it is possible to achieve the minimum effective value control of the inductor current, but when T1 + T2 + T3 > T min it may not necessarily achieve the optimal effective value of the inductor current. In addition, the calculation accuracy of the T2 and T1 values in this method still depends on the reference measurement of the inductance, and it is still difficult to adapt to the errors caused by factors such as the working temperature change of the converter and the manufacturing parameter deviation. Summary of the Invention

[0019] The purpose of the present invention is to provide a dual-time-scale optimization control method for a four-switch buck-boost converter to solve the problems raised in the above background technology.

[0020] To achieve the above purpose, the present invention provides the following technical solutions:

[0021] A dual-time-scale optimization control method for a four-switch buck-boost converter, including an output voltage regulator and an inductor current optimizer. The method includes:

[0022] S1: Obtain the input voltage sampling value V i , output voltage sampling value V o , inductor current value I L and the duration T1 of the common conduction of the switching transistors Q1 and Q4, and the duration T2 of the common conduction of the switching transistors Q1 and Q3 during the operation of the four-switch buck-boost converter in the current switching period;

[0023] S2: The inductor current optimizer reads the data in S1, calculates the effective value I Lrms (k) of the inductor current in the current switching period of the four-switch buck-boost converter and the effective value I Lrms (k - 1) of the inductor current in the previous switching period, and calculates the adjustment coefficients K a , K b by taking the difference. The inductor current optimizer outputs the adjustment coefficients K a , K b to the output voltage regulator;

[0024] S3: The output voltage regulator combines the output voltage reference value V ref , output voltage sampling value V o , input voltage sampling value V i , and calculates the duration T1 of the common conduction of the switching transistors Q1 and Q4 and the duration T2 of the common conduction of the switching transistors Q1 and Q3 through the PI algorithm. The output voltage regulator outputs the duration T1 of the common conduction of the switching transistors Q1 and Q4 and the duration T2 of the common conduction of the switching transistors Q1 and Q3 to the inductor current optimizer, and outputs a control signal for controlling the on-off states of the switching transistors Q1 to Q4 in the next switching period;

[0025] S4: Repeat S1 to S3.

[0026] In some embodiments, the step of obtaining the adjustment coefficients K a 、K b is as follows:

[0027] S2.1: Obtain the effective value I Lrms (k) of the inductor current, and the difference ΔI(k) between I Lrms (k - 1);

[0028] S2.2: Determine whether ΔI(k)>0 holds. If so, proceed to S2.3; if not, proceed to S2.4;

[0029] S2.3: Determine whether DK = 1 holds. If so, proceed to S2.5; if not, proceed to S2.6;

[0030] S2.4: Determine whether DK = 1 holds. If so, proceed to S2.6; if not, proceed to S2.5;

[0031] S2.5: Calculate the values of the adjustment coefficients K a 、K b according to the following formula:

[0032]

[0033] S2.6: Calculate the values of the adjustment coefficients K a 、K b according to the following formula:

[0034]

[0035] where K i represents the optimization step size adjustment coefficient; abs() represents the absolute value function; DK represents the optimization direction of the previous iteration, DK = 0 indicates that the optimization direction points to the optimal point, and DK = 1 indicates that the optimization direction deviates from the optimal point.

[0036] In some embodiments, in S3, the control method of the output voltage regulator is as follows:

[0037] The output voltage reference value V ref and the output voltage sampled value V o are subtracted to obtain the output voltage error signal E o . The output voltage error signal E o is processed by a PI algorithm to obtain the time signal T u . The time signal T u is respectively multiplied by the adjustment coefficients K a 、K b to obtain the control quantities T a 、T b ;

[0038] Sampled value V of input voltage i and the sampled value V of output voltage o are subtracted to obtain the difference signal E between the input voltage and the output voltage io , and it is multiplied by the feed - forward regulation coefficient K c to obtain the control quantity T c ;

[0039] The difference signal E io and the feed - forward regulation coefficient K of the PI regulator d are used to obtain the control quantity T d ; The feed - forward regulation coefficients K c , K d are obtained through actual debugging;

[0040] The control quantity T a is subtracted from T c to obtain the duration T1 of the common conduction of Q1 and Q4, and T b is added to T d to obtain the duration T2 of the common conduction of the switching tubes Q1 and Q3;

[0041] The duration T1 of the common conduction of Q1 and Q4, the duration T2 of the common conduction of the switching tubes Q1 and Q3, the sampled value V of the input voltage i and the sampled value V of the output voltage o are used to calculate the duration T3 of the common conduction of the switching tubes Q2 and Q3 through the following formula:

[0042]

[0043] The duration T1 of the common conduction of the switching tubes Q1 and Q4, the duration T2 of the common conduction of the switching tubes Q1 and Q3, and the duration T3 of the common conduction of the switching tubes Q2 and Q3 are used to calculate the duration T4 of the common conduction of Q2 and Q4 according to the following formula:

[0044]

[0045] In the formula, T smin represents the minimum switching period of the four - switch buck - boost converter;

[0046] The conduction times of the switching tubes Q1 - Q4 in the next switching period are calculated through T1, T2, T3, and T4.

[0047] In some embodiments, the calculation formula for the effective value of the inductor current is:

[0048]

[0049] where, I zrepresents the minimum resonant current required for the switching transistors Q1 to Q4 to achieve soft switching, T s represents the switching period.

[0050] Beneficial effects: In the present invention, the complex inductor current optimization calculation is placed within the long-time scale algorithm, and the simple output voltage regulation control is placed within the short-time scale algorithm, which can not only ensure the performance of the converter, but also reduce the computational requirements of the algorithm for the controller, and thus can reduce the cost of the four-switch buck-boost converter. Description of the drawings

[0051] Figure 1 is the topology diagram of the four-switch buck-boost transformer;

[0052] Figure 2 is the working timing waveform diagram of the four-switch buck-boost transformer under the control of the quadrilateral inductor current modulation strategy;

[0053] Figure 3-1 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t0 - t1 stage;

[0054] Figure 3-2 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t1 - t2 stage;

[0055] Figure 3-3 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t2 - t3 stage;

[0056] Figure 3-4 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t3 - t4 stage;

[0057] Figure 3-5 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t4 - t5 stage;

[0058] Figure 3-6 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t5 - t6 stage;

[0059] Figure 3-7 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t6 - t7 stage;

[0060] Figure 3-8 is for the four-switch buck-boost transformer in Figure 2 the commutation path during the t7 - t8 stage;

[0061] Figure 4It is a function image of the effective value of the inductor current obtained by a four-switch buck-boost transformer under different load currents;

[0062] Figure 5 It is a principle block diagram of the prior art for realizing the minimum inductor current effective value control of a four-switch buck-boost transformer according to the look-up table method;

[0063] Figure 6 It is a principle block diagram of the prior art for realizing the minimum effective value control of the inductor current of a four-switch buck-boost transformer according to the critical current approximation frequency conversion control method;

[0064] Figure 7 It is a control logic flowchart of the prior art for realizing the critical current approximation frequency conversion control method;

[0065] Figure 8 It is a waveform change diagram of the inductor current under different conditions when the prior art realizes the critical current approximation frequency conversion control method;

[0066] Figure 9 It is a method block diagram of the output voltage regulator of the present invention;

[0067] Figure 10 It is a principle block diagram of the output voltage regulation of the present invention;

[0068] Figure 11 It is a flowchart of the inductor current optimizer of the present invention for realizing iterative optimization;

[0069] Figure 12 It is a regulation process diagram of the output voltage regulator of the present invention when the input voltage remains unchanged and the output current suddenly increases;

[0070] Figure 13 It is a regulation process diagram of the output voltage regulator of the present invention when the input voltage remains unchanged and the output current suddenly decreases;

[0071] Figure 14 It is a waveform diagram of the regulation process when the present invention iteratively optimizes the inductor current. Specific embodiments

[0072] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0073] When the four-switch buck-boost converter adopts the quadrilateral inductor current modulation strategy, as Figure 2 shown, it is a working timing waveform diagram of the four-switch buck-boost converter.

[0074] See Figure 3-1 , during the time period from t0 to t1, the switching transistors Q1 and Q4 are turned on, and the voltage V across the inductor L is the input power supply voltage V i , and when the inductor current rises from I z to I1, the switching transistor Q4 is turned off.

[0075] See Figure 3-2 , during the time period from t1 to t2, the switching transistor Q4 is turned off, the inductor current charges the parasitic capacitance of the switching transistor Q4 and discharges the parasitic capacitance of the switching transistor Q3, the voltage across the switching transistor Q3 drops, and when the voltage across the switching transistor Q3 drops to 0, the anti-parallel diode of the switching transistor Q3 conducts, and this mode ends.

[0076] See Figure 3-3 , during the time period from t2 to t3, the switching transistor Q3 is turned on with zero voltage at t2, and the voltage V across the inductor L = V i - V o . When V i > V o , the inductor current I L rises linearly; when V i < V o , the inductor current I L drops linearly.

[0077] See Figure 3-4 , during the time period from t3 to t4, the switching transistor Q1 is turned off, and the inductor current I L charges the parasitic capacitance of the switching transistor Q1 and discharges the parasitic capacitance of the switching transistor Q2, the voltage across the switching transistor Q2 drops, and when the voltage across the switching transistor Q2 drops to 0, the anti-parallel diode of the switching transistor Q2 conducts, and this mode ends.

[0078] See Figure 3-5 , during the time period from t4 to t5, the switching transistor Q2 is turned on with zero voltage at t4, and the voltage V across the inductor L = - V o , and the inductor current I L drops linearly to - I z when the switching transistor Q3 is turned off.

[0079] See Figure 3-6 , during the time period from t5 to t6, after the switching transistor Q3 is turned off, the inductor current I L charges the parasitic capacitance of the switching transistor Q3 and discharges the parasitic capacitance of the switching transistor Q4. When the voltage across the switching transistor Q4 drops to 0, the anti-parallel diode of the switching transistor Q4 conducts.

[0080] See Figure 3-7, during the time period t6 - t7, the switching transistor Q4 conducts with zero voltage, and the voltage V across the inductor L = 0. Ignoring the equivalent series resistance of the circuit, the inductor current I L is maintained at -I z .

[0081] See Figure 3-8 . During the time period t7 - t8, the switching transistor Q2 is turned off, and the inductor current discharges the parasitic capacitance of the switching transistor Q1 and charges the parasitic capacitance of the switching transistor Q2. The voltage across the switching transistor Q1 linearly decreases. When the voltage across the switching transistor Q1 drops to 0, the switching transistor Q1 conducts with zero voltage, starting a new switching cycle.

[0082] According to the above analysis, in order to achieve ZVS turn-on of Q1, Q2, Q3, and Q4, referring to Figure 3-1 to 3-8 , the inductor current I L should satisfy the following constraint conditions:

[0083]

[0084] where t d is the dead time between the switching transistors Q1 and Q2 (Q3 and Q4), and C1, C2, C3, and C4 respectively represent the parasitic capacitances of the switching transistors Q1, Q2, Q3, and Q4.

[0085] According to Figure 2 , ignoring the dead time, the expression of the inductor current I L in one switching cycle is:

[0086]

[0087] where T1 is the duration of the common conduction of the switching transistors Q1 and Q4, T2 is the duration of the common conduction of the switching transistors Q1 and Q3, T3 is the duration of the common conduction of the switching transistors Q2 and Q3, and T s is the switching cycle. I1 is the inductor current value at time T1, and I2 is the inductor current value at time T1 + T2.

[0088] According to the circuit operating principle and the relationships between physical quantities such as the input voltage, output voltage, output current, and switching frequency, the expression of the effective value of the inductor current can be derived as:

[0089]

[0090] The expression of the output current is:

[0091]

[0092] It can be seen from Equation (3) that the output current I out is determined by the input voltage Vi 、 Output voltage V o is determined by the functions of T1 and T2. When V i and V o are constant, a constant output current I out corresponds to different combinations of T1 and T2 values. It can be seen from Equation (2) that different combinations of T1 and T2 values will generate different effective values of inductor current I Lrms . Substituting Equation (3) into (2) to eliminate T1, the function of I Lrms is as follows:

[0093]

[0094] As Figure 9 shown, a dual-time-scale optimal control method for a four-switch buck-boost converter includes an output voltage regulator and an inductor current optimizer. The method includes:

[0095] S1: Obtain the sampled value V i of the input voltage, the sampled value V o of the output voltage, the value I L of the inductor current, and the duration T1 of the common conduction of switching transistors Q1 and Q4, and the duration T2 of the common conduction of switching transistors Q1 and Q3 during the operation of the four-switch buck-boost converter in the current switching period; in the initial state, the data values obtained in this step are all 0.

[0096] S2: The inductor current optimizer reads the data in S1, calculates the effective value I Lrms (k) of the inductor current in the current switching period of the four-switch buck-boost converter and the effective value I Lrms (k - 1) of the inductor current in the previous switching period, calculates ΔI(k) by taking the difference, and determines the adjustment coefficients K a , K b according to ΔI(k);

[0097] As Figure 11 shown, the steps to obtain the adjustment coefficients K a , K b include the following:

[0098] S2.1: Obtain the difference ΔI(k) between the effective values I Lrms (k) and I Lrms (k - 1) of the inductor current;

[0099] S2.2: Judge whether ΔI(k) > 0 holds. If it is, go to S2.3; if not, go to S2.4;

[0100] S2.3: Judge whether DK = 1 holds. If it is, go to S2.5; if not, go to S2.6;

[0101] S2.4: Determine whether DK = 1 holds. If it does, proceed to S2.6; if not, proceed to S2.5.

[0102] S2.5: Calculate the adjustment coefficient K according to the following formula a and K b value:

[0103]

[0104] S2.6: Calculate the adjustment coefficient K according to the following formula a and K b value:

[0105]

[0106] where K i represents the optimization step size adjustment coefficient. To ensure stability, it generally takes values between 0.1 and 0.2; abs() represents the absolute value function; DK represents the optimization direction of the previous iteration. DK = 0 indicates that the optimization direction points to the optimal point, and DK = 1 indicates that the optimization direction deviates from the optimal point.

[0107] In the above steps, by judging whether ΔI(k)>0 and DK = 1 hold, the perturbation direction of this cycle is judged to make the effective value I of the inductor current Lrms whether it is far from the optimal point. For example, let the value of I Lrms (k - 1) be on the left and the value of I Lrms (k) be on the right. If ΔI(k)>0 holds, it means that it is lower on the left and higher on the right. If DK = 1 holds, it indicates that the perturbation direction generated in the current cycle makes the effective value I of the inductor current Lrms far from the optimal point, and it should optimize towards the lower left. K b needs to be reduced, that is, proceed to S2.5; if ΔI(k)>0 holds but DK = 1 does not hold, it indicates that the optimization direction of the effective value I of the inductor current under the perturbation direction of the current cycle Lrms is correct, and it should optimize towards the higher right. K b needs to be increased, and proceed to S2.6.

[0108] Another example, if ΔI(k)>0 does not hold, it means that it is higher on the left and lower on the right. If DK = 1 holds, it indicates that it should optimize towards the higher left. K b needs to be increased, and proceed to S2.6. If DK = 1 does not hold, it indicates that it should optimize towards the lower right. K b needs to be reduced, and proceed to S2.5.

[0109] In the above embodiments, the gradient descent method can be used, or other applicable methods such as the Newton iteration method can also be used, which is not limited. When adjusting the coefficient Ka , K b After the optimization calculation, the inductor current optimizer outputs the adjustment coefficient K a , K b to the output voltage regulator.

[0110] S3: As Figure 9 shown, in order to achieve the fast calculation of the output voltage regulator, the voltage regulator adopts a PI regulator. The output voltage regulator combines other input signals to output the voltage reference value V ref , the output voltage sampling value V o , and the input voltage sampling value V i . Through the PI algorithm, the control signals (PWM_Q1, PWM_Q2, PWM_Q3, PWM_Q4) of T1, T2, and the four switching tubes are calculated.

[0111] As Figure 10 shown, the control method of the output voltage regulator is as follows:

[0112] The voltage reference value V ref and the output voltage sampling value V o are subtracted to obtain the output voltage error signal E o . The output voltage error signal E o is processed through the PI algorithm to obtain the time signal T u . The time signal T u is respectively multiplied by the adjustment coefficients K a , K b to obtain the control quantities T a , T b ;

[0113] The input voltage sampling value V i is subtracted from the output voltage sampling value V o to obtain the difference signal E io between the input voltage and the output voltage, and it is multiplied by the feedforward adjustment coefficient K c to obtain the control quantity T c ;

[0114] The difference signal E io and the feedforward adjustment coefficient K d of the PI regulator are used to obtain the control quantity T d ; The feedforward adjustment coefficients K c , K c are all determined by the trial-and-error method during the debugging process.

[0115] The control quantity T a is subtracted from T c to obtain T1, and T b is added to T d to obtain T2;

[0116] T1, T2, and the input voltage sampling value V i and the output voltage sampling value V o The duration T3 of the common conduction of switching transistors Q2 and Q3 is obtained by calculation according to the following formula:

[0117]

[0118] T1, T2, and T3 are used to calculate the duration T4 of the common conduction of switching transistors Q2 and Q4 according to the following formula:

[0119]

[0120] In the formula, T smin represents the minimum switching period of the four-switch buck-boost converter.

[0121] A traditional PI regulator is used in the output voltage regulator, and only 6 multiplication operations are required for other signal operation processes, which has a low computing power requirement for the controller and can complete the calculation within a few microseconds.

[0122] The output voltage regulator outputs T1 and T2 to the inductor current optimizer and outputs a control signal for controlling the on / off states of switching transistors Q1 to Q4 in the next switching period.

[0123] S4: As Figure 11 shown, repeat S1 to S3. The inductor current optimizer continuously updates and iterates the variables, and the output voltage regulator outputs the control signals of switching transistors Q1 to Q4 to continuously maintain the effective value of the inductor current at the minimum.

[0124] In one embodiment, when the input voltage V i remains unchanged and the output current I out suddenly increases, the adjustment process of the output voltage regulator is as Figure 12 shown. In this embodiment, the output current I out increases, the energy storage capacitor discharges, resulting in a decrease in the output voltage V o , the output voltage error signal E o increases, and after being processed by the PI algorithm, the time signal T u increases. Therefore, the control quantities T a and T b increase. Since the output voltage V o decreases, the difference signal E io between the input voltage and the output voltage increases, and the control quantity T d multiplied by the feedforward adjustment coefficient K d also increases.

[0125] The control quantity T a is subtracted from T c to obtain T1, Tb Added to T d to obtain T2. The values of T1 and T2 are input into the inductor current optimizer, which prompts the inductor current optimizer to calculate a new K through iterative optimization calculation a and K b , and then optimized T1 and T2 are obtained to obtain the minimum effective value of the inductor current and maximize the efficiency of the converter.

[0126] In one embodiment, when the input voltage remains unchanged and the output current suddenly decreases, the adjustment process of the output voltage regulator is as follows Figure 13 shown.

[0127] It should be noted that in several cycles after the output current mutation (generally 2 to 4 times the switching cycle), in order to ensure the power supply demand of the load, the output voltage regulator still calculates T1 and T2 times according to the original K a and K b , but the combination of T1 and T2 at this time may not be optimal and the inductor current is relatively large. After the output voltage stabilizes, the inductor current optimizer calculates a new K through iterative optimization calculation a and K b , and then optimized T1 and T2 are obtained, reducing the effective value of the inductor current. The schematic diagram of its adjustment process is as follows Figure 14 shown. (a) is the inductor current waveform in a certain switching cycle before the output load mutation, (b) is the inductor current waveform in a certain switching cycle dozens of microseconds after the load mutation, and (c) is the inductor current waveform in a certain switching cycle obtained after iterative optimization from dozens of milliseconds to hundreds of milliseconds. It can be seen from the figure that the effective value of the inductor current is significantly reduced compared to Figure (b).

[0128] In the above control method, a traditional PI regulator is used in the output voltage regulator, and only 6 multiplication operations are required for other signal operation processes, with a low computing power requirement for the controller and can be completed within a few microseconds. Therefore, the output voltage regulator can operate at a relatively high frequency, usually the same as or close to the switching frequency of the converter, and its operating cycle is usually set from a few microseconds to dozens of microseconds, with the setting standard being able to complete the operation. The inductor current optimizer needs to perform online calculation on the effective value of the inductor current. Although there are many multiplication calculations in the calculation process, the inductor current optimizer can operate at a relatively low frequency, and its operating cycle is usually set from a few milliseconds to dozens of milliseconds. Since the operating frequency of the optimizer is usually set relatively low, its requirement for the computing ability of the controller is still low.

[0129] Generally speaking, the present invention places the complex inductor current optimization calculation within the long-time scale algorithm and the simple output voltage regulation control within the short-time scale algorithm, which can not only ensure the performance of the converter, but also reduce the computational requirements of the algorithm for the controller, thereby reducing the cost of the four-switch buck-boost converter.

[0130] Meanwhile, compared with the prior art, the present invention uses a PI regulator, so it can automatically adjust the values of T1, T2, T3 and T4 according to the load conditions, without the need to establish a large number of data tables, eliminating the data holes between data points in the look-up table method and improving the control accuracy; since the inductor current optimizer adopts an iterative optimization algorithm, the technical solution of the present invention can minimize the effective value of the inductor current of the converter under all input and output states.

[0131] In addition, the adjustment process of T1, T2, T3 and T4 of the present invention only uses the sampled values V i 、V o of the input voltage and output voltage of the converter, and does not include physical quantities such as inductors and capacitors in the circuit that are easily affected by factors such as operating temperature or manufacturing errors. Therefore, the technical solution of the present invention can eliminate the negative impacts on the performance of the converter caused by factors such as temperature changes and manufacturing parameter deviations.

[0132] Although this specification is described according to embodiments, not every embodiment only contains an independent technical solution. This narrative way of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0133] Therefore, the above are only the preferred embodiments of the present application, and are not used to limit the scope of implementation of the present application; that is, all equivalent transformations made according to the scope of the claims of the present application are within the protection scope of the claims of the present application.

Claims

1. A dual-time-scale optimization control method for a four-switch buck-boost converter. The switching tube buck-boost converter includes switching tubes Q1 to Q2 and an inductor L, where, The switching transistors Q1 and Q2 form the front bridge, and the switching transistors Q3 and Q4 form the rear bridge. The switching transistors Q1 and Q3 are the upper bridge arms, and the switching transistors Q2 and Q4 are the lower bridge arms. It is characterized in that it includes an output voltage regulator and an inductor current optimizer. The method includes: S1: Obtain the sampled value V of the input voltage during the operation of the four-switch buck-boost converter in the current switching period, i the sampled value V of the output voltage, o the inductor current value I, L and the duration T1 of the common conduction of the switching transistors Q1 and Q4, and the duration T2 of the common conduction of the switching transistors Q1 and Q3; S2: The inductor current optimizer reads the data in S1, and calculates the effective value I of the inductor current in the current switching period of the four-switch buck-boost converter Lrms (k) and the effective value I of the inductor current in the previous switching period Lrms (k-1), and calculates the adjustment coefficient K through subtraction a 、K b , and the inductor current optimizer outputs the adjustment coefficients K a 、K b to the output voltage regulator; S3: The output voltage regulator combines the output voltage reference value V ref , the output voltage sampled value V o , the input voltage sampled value V i , and through the PI algorithm, calculates T1 and T2. The output voltage regulator outputs T1 and T2 to the inductor current optimizer and outputs a control signal for controlling the on / off states of the switching transistors Q1 to Q4 in the next switching cycle; S4: Repeat S1 to S3; Obtain the adjustment coefficient K in S2 a , K b The steps are as follows: S2.1: Obtain the effective value I of the inductor current Lrms (k), I Lrms (k - 1) difference ΔI(k); S2.2: Determine whether ΔI(k)>0 holds. If it is, go to S2.3; if not, go to S2.4; S2.3: Determine whether DK = 1 holds. If it is, go to S2.5; if not, go to S2.6; S2.4: Determine whether DK = 1 holds. If it is, go to S2.6; if not, go to S2.5; S2.5: Calculate the adjustment coefficient K according to the following formula a , K b value: S2.6: Calculate the adjustment coefficient K according to the following formula a , K b value: Among them, K i represents the optimization step size adjustment coefficient; abs() represents the absolute value function; DK represents the optimization direction of the previous iteration, DK = 0 indicates that the optimization direction points to the optimal point, and DK = 1 indicates that the optimization direction deviates from the optimal point; In S3, the control method of the output voltage regulator is as follows: Output voltage reference value V ref and output voltage sampled value V o are subtracted to obtain output voltage error signal E o . The output voltage error signal E o is processed by a PI algorithm to obtain time signal T u . The time signal T u is respectively multiplied by regulation coefficients K a and K b to obtain control quantities T a and T b ; Input voltage sampling value V i is subtracted from the output voltage sampling value V o to obtain the difference signal E between the input voltage and the output voltage io , and is multiplied by the feedforward adjustment coefficient K c to obtain the control quantity T c ; Difference signal E io and the feedforward adjustment coefficient K of the PI regulator d to obtain the control quantity T d ; The feedforward adjustment coefficients K c , K d are obtained through actual debugging; Control quantity T a Subtract from T c to obtain the conduction duration T1 during which Q1 and Q4 conduct simultaneously, T b Add to T d to obtain the conduction duration T2 during which switching transistors Q1 and Q3 conduct simultaneously; The duration T1 during which Q1 and Q4 are conducting together, the duration T2 during which switching transistors Q1 and Q3 are conducting together, the input voltage sampled value V i and the output voltage sampled value V o The duration T3 during which switching transistors Q2 and Q3 are conducting together is obtained by calculation using the following formula: The duration T1 of the common conduction of the switching transistors Q1 and Q4, the duration T2 of the common conduction of the switching transistors Q1 and Q3, and the duration T3 of the common conduction of the switching transistors Q2 and Q3 are used to calculate the duration T4 of the common conduction of Q2 and Q4 according to the following formula: where, T smin represents the minimum switching period of the four-switch buck-boost converter; The conduction times of the switching transistors Q1 to Q4 in the next switching cycle are obtained by calculating T1, T2, T3, and T4.

Images