Switching resonant converter state track control state switching parameter prediction calculation method

Abstract

The invention discloses a prediction and calculation method for state track control state switching parameters of a switch resonant converter. The method comprises the following steps: S1, selecting parameters to design an orthogonal experiment table; s2, according to the state of the switch resonant conversion circuit, establishing a formula to analyze the input voltage and the load; s3, establishing a mathematical model with unknown parameters, input voltage and load; and S4, solving unknown parameters by using an LM algorithm to obtain a switching parameter calculation formula. Therefore, the dynamic response speed of the converter is improved.

Description

Technical Field

[0001] The invention relates to the field of converters, and in particular to a method for predicting and calculating state switching parameters of a switch resonant converter state trajectory control. Background Art

[0002] A switching resonant converter is an electronic component that uses capacitors and switching tubes to achieve voltage conversion, and conventional switching resonant converters are equipped with multiple switching modes to achieve fast output voltage startup and fast load dynamic response. In order to improve the dynamic performance of the converter, it is necessary to design the switching conditions of the state trajectory corresponding to the switching mode. Figure 2 The switching resonant converter shown has four switching modes. The converter switches cyclically among the four switching modes in sequence. When switching from Mode 1 to Mode 2, and when switching from Mode 3 to Mode 4, the switching condition is that the resonant current is greater than or equal to the corresponding switching parameter. However, the originally fixed switching parameters cannot adapt to different load conditions, which affects the power transmission efficiency of the converter. Summary of the invention

[0003] The present invention aims to solve the technical problems existing in the prior art, and in particular innovatively proposes a method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control, thereby improving the dynamic performance of the converter and thus improving the dynamic response speed of the converter.

[0004] In order to achieve the above object, the present invention provides a method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control state, comprising the following steps:

[0005] S1: Select parameters to design orthogonal experimental table;

[0006] S2: According to the state of the switch resonant conversion circuit, establish a formula to analyze the input voltage and load;

[0007] S3: Establish a mathematical model with unknown parameters and input voltage and load;

[0008] S4: Use the LM algorithm to solve the unknown parameters and obtain the switching parameter calculation formula.

[0009] In the above scheme: step S2 also includes the following steps:

[0010] S2-1: Set the steady-state output voltage to a constant value of V o ;

[0011] S2-2: According to the charge and discharge balance principle of the output capacitor, the average output current I out Equal to the average load current I Load , and the average output current and output voltage satisfy the following relationship:

[0012] I out =I Load =V o / R;

[0013] S2-3: Calculate the average value of the resonant inductor voltage of the switching resonant conversion circuit during Mode 2 and Mode 4;

[0014] According to Kirchhoff's voltage law, the average value of the resonant inductor voltage during Mode 2 is (V in / 2-V o );

[0015] The average value of the resonant inductor voltage during Mode 4 is (V o -V in / 2);

[0016] Among them, V in is the input voltage of the switching resonant converter circuit, V o To maintain a constant value for steady-state output voltage;

[0017] S2-4: Calculate the inductor current slope during Mode 2 and Mode 4;

[0018] The average value of the resonant inductor voltage during Mode 2 is (V in / 2-V o ) / L R ;

[0019] Among them, V in is the input voltage of the switching resonant converter circuit, V o To keep the steady-state output voltage constant, L R is the resonant inductance value;

[0020] The average value of the resonant inductor voltage during Mode 4 is (V o -V in / 2) / L R ;

[0021] S2-5: Using the geometric analysis method, the inductor current slope is obtained as:

[0022] d L / dt=(V in / 2-V o ) / L r =(I T -I S ) / (T sw / 2-T s )

[0023]

[0024] Among them, L / dt is the inductor current slope, I T is the resonant current when switching from Mode 2 to Mode 3, T s is the timing difference between the upper and lower half bridges of the switched capacitor resonant converter; T sw is the switching cycle, I S The actual physical meaning is the maximum forward resonant current of Mode 1, which is the conditional parameter for switching from Mode 1 to Mode 2. out is the average output current;

[0025] Average output current I out , switching cycle T sw And the time difference between the upper and lower half bridges T s The following constraints exist:

[0026]

[0027] where f sw is the switching frequency and f sw =1 / T sw ;

[0028] S2-6: Simplify the calculation formula of step S2-5.

[0029] In the above scheme: the simplified calculation formula of step S2-6 is as follows:

[0030]

[0031] Among them, V in and L r are known circuit parameters, R is the resistance, which can be obtained based on load detection, V o Determined by the input reference value of the controller.

[0032] In the above scheme: Step S3 establishes the model as follows:

[0033] I s =k*(V in +a)*(R+b) 2 +c;

[0034] Among them, I s is the switching parameter, k, a, b and c are the model parameters.

[0035] In the above scheme: step S4 also includes the following steps:

[0036] S4-1: storing the data of multiple sets of input voltage Vin, resistance R and corresponding switching parameter Is in MATLAB vectors respectively;

[0037] S4-2: According to the data and experimental experience, the initial guess values ​​are set to k = 0.000000123381, a = 118.338189676784, b = 1334.346495829290, c = -28;

[0038] S4-3: Define mathematical model and establish function;

[0039] S4-4: define the residual function;

[0040] S4-5: Fitting model parameters.

[0041] In the above scheme: the function established in step S4-3 is as follows:

[0042] modelFunc=@(params,Vin,R)params(1)*(Vin+params(2)).*(R+params(3)).^2+params(4). This function accepts model parameters and input data, namely input voltage Vin, resistance R, and outputs the calculated switching parameter Is. This function implements the mathematical model Is=k*(Vin+a)*(R+b)^2+c.

[0043] Here, @() means creating an anonymous function, params(1) corresponds to k, params(2) corresponds to a, params(3) corresponds to b, and params(4) corresponds to c.

[0044] In the above scheme: Step S4-3 also includes comparing the observed data Is in the orthogonal experimental table in Step S1 with the Is value calculated by the model, and calculating the difference between the model prediction value and the observed data value;

[0045] In the above scheme: step S4-4 also includes the following steps:

[0046] S4-4-1: Perform nonlinear least squares fitting of the model using the lsqnonlin function of MATLAB; adjust the model parameters k, a, b, and c based on the provided initial guess values ​​so that the difference between the calculated current value and the observed data is minimized;

[0047] S4-4-2: Select the Levenberg-Marquardt algorithm and estimate the optimal parameters by minimizing the residual function. The following fitting results are obtained:

[0048] k=1.18707*10 -7 , a=119.12241, b=1341.51766, c=-28.09194

[0049] S4-4-3: Substitute the fitting results into the following calculation formula:

[0050] I s =1.18707*10 7 (V in +119.12241)*(R+1341.51766) 2 -28.09194.

[0051] In summary, the beneficial effects of the present invention are: it can adapt to different loads and input voltages, obtain suitable resonant current stress, ensure the soft development operation of the switched capacitor resonant converter, reduce the switching loss of the converter, improve the transmission efficiency of the converter, and speed up the dynamic adjustment speed of the output voltage of the switched capacitor resonant converter. At the same time, the calculation parameters can be updated according to different load and voltage data, and have an adaptive effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of the present invention.

[0053] Figure 2 It is the circuit diagram of a switching resonant converter.

[0054] Figure 3 It is the plane trajectory diagram of the vout-iL state during the switching capacitor resonant operation process.

[0055] Figure 4 In the figure, a is the waveform of the resonant current under the input voltage of 20V, the output voltage of 10V, and the load of 5Ω; b is the waveform of the resonant voltage under the input voltage of 20V, the output voltage of 10V, and the load of 5Ω; c is the waveform of the output voltage under the input voltage of 20V, the output voltage of 10V, and the load of 5Ω.

[0056] Figure 5 In the figure, a is the waveform of the resonant current under the input voltage of 30V, the output voltage of 15V and the load of 10Ω; b is the waveform of the resonant voltage under the input voltage of 30V, the output voltage of 15V and the load of 10Ω; c is the waveform of the output voltage under the input voltage of 30V, the output voltage of 15V and the load of 10Ω.

[0057] Figure 6 In the figure, a is the waveform of the resonant current under the input voltage of 50V, the output voltage of 25V and the load of 20Ω; b is the waveform of the resonant voltage under the input voltage of 50V, the output voltage of 25V and the load of 20Ω; c is the waveform of the output voltage under the input voltage of 50V, the output voltage of 25V and the load of 20Ω. DETAILED DESCRIPTION

[0058] The present invention will be further described below by way of embodiments and in conjunction with the accompanying drawings:

[0059] like Figure 1 As shown, a method for calculating state switching parameters of a switching resonant converter state trajectory control comprises the following steps:

[0060] S1: Select parameters to design orthogonal experimental table;

[0061] According to the data of multiple experimental results, the orthogonal experimental table is shown in Table 1;

[0062] Table 1: Orthogonal experiment table

[0063] Input voltage resistance Is value 30V 10Ω 4.5AA 115V 55Ω 28.5A 200V 100Ω 50A 30V 55Ω 5.7A 115V 100Ω 28.9A 200V 10Ω 41.5A 30V 100Ω 7.4A 115V 10Ω 19.2A 200V 55Ω 47.8A

[0064] S2: According to the state of the switch resonant conversion circuit, establish a formula to analyze the input voltage and load;

[0065] S2-1: Set the steady-state output voltage to a constant value of V o ;

[0066] S2-2: According to the charge and discharge balance principle of the output capacitor, the average output current I out Equal to the average load current I Load , and the average output current and output voltage satisfy the following relationship:

[0067] I out =I Load =V o / R;

[0068] S2-3: Calculate the average value of the resonant inductor voltage of the switching resonant conversion circuit during Mode 2 and Mode 4;

[0069] According to Kirchhoff's voltage law, the average value of the resonant inductor voltage during Mode 2 is (V in / 2-V o );

[0070] The average value of the resonant inductor voltage during Mode 4 is (V o -V in / 2);

[0071] Among them, V in is the input voltage of the switching resonant converter circuit, V o To maintain a constant value for steady-state output voltage;

[0072] S2-4: Calculate the inductor current slope during Mode 2 and Mode 4;

[0073] The average value of the resonant inductor voltage during Mode 2 is (V in / 2-V o ) / LR ;

[0074] Among them, V in is the input voltage of the switching resonant converter circuit, V o To keep the steady-state output voltage constant, L R is the resonant inductance value.

[0075] The average value of the resonant inductor voltage during Mode 4 is (V o -V in / 2) / L R ;

[0076] S2-5: Using geometric analysis method, we can get the mathematical relationship:

[0077] d L / dt=(V in / 2-V o ) / L r =(I T -I S ) / (T sw / 2-T s )

[0078]

[0079] Among them, L / d t is the inductor current slope, I T is the resonant current when switching from Mode 2 to Mode 3, T s is the timing difference between the upper and lower half bridges of the switched capacitor resonant converter; T sw is the switching cycle, I S The actual physical meaning is the maximum forward resonant current of Mode 1, which is the conditional parameter for switching from Mode 1 to Mode 2. out is the average output current;

[0080] Average output current I out , switching cycle T sw And the time difference between the upper and lower half bridges T s The following constraints exist:

[0081]

[0082] where f sw is the switching frequency and f sw = 1 / T sw ;

[0083] S2-6: The calculation formula of step S2-6 is simplified and solved as follows:

[0084]

[0085] Among them, V in and L r are known circuit parameters, R is the resistance, which can be obtained based on load detection, V o Determined by the input reference value of the controller; therefore, under the proposed control method, in the dynamic process, the switching parameter I is designed S With input voltage V in , and the load R.

[0086] S3: Establish a mathematical model with unknown parameters and input voltage and load;

[0087] The model is established as follows:

[0088] I s =k*(V in +a)*(R+b) 2 +c;

[0089] Among them, I s is the switching parameter, k, a, b and c are the model parameters.

[0090] S4: Use the LM algorithm to solve the unknown parameters and obtain the switching parameter calculation formula;

[0091] S4-1: storing the data of multiple sets of input voltage Vin, resistance R and corresponding switching parameter Is in MATLAB vectors respectively;

[0092] S4-2: In order to fit the model using the Levenberg-Marquardt algorithm, it is necessary to make initial guesses for the unknown parameters k, a, b, and c. Based on the data and experimental experience, the initial guess values ​​are set to k = 0.000000123381, a = 118.338189676784, b = 1334.346495829290, c = -28;

[0093] S4-2: Define the mathematical model and establish the function as follows:

[0094] modelFunc=@(params,Vin,R)params(1)*(Vin+params(2)).*(R+params(3)).^2+params(4). This function accepts model parameters and input data, namely input voltage Vin, resistance R, and outputs the calculated switching parameter Is. This function implements the mathematical model Is=k*(Vin+a)*(R+b)^2+c.

[0095] Here, @() means creating an anonymous function, params(1) corresponds to k, params(2) corresponds to a, params(3) corresponds to b, and params(4) corresponds to c.

[0096] S4-3: define the residual function;

[0097] The residual function is used to calculate the difference between the model predictions and the observed data.

[0098] S4-3-1: Compare the observed data Is in the orthogonal experimental table in step S1 with the Is value calculated by the model, and calculate the difference between the observed data value and the model predicted value, that is, the difference V = observed data value - model predicted value; that is, the residual, and the parameters will be fitted by minimizing this residual.

[0099] S4-4: Fitting model parameters;

[0100] S4-4-1: Perform nonlinear least squares fitting of the model using the lsqnonlin function of MATLAB; adjust the model parameters k, a, b, and c based on the provided initial guess values ​​so that the difference between the calculated current value and the observed data is minimized;

[0101] S4-4-2: Select the Levenberg-Marquardt algorithm and estimate the optimal parameters by minimizing the residual function. The following fitting results are obtained:

[0102] k=1.18707*10 -7 , a=119.12241, b=1341.51766, c=-28.09194

[0103] S4-4-3: Substitute the fitting results into the following calculation formula:

[0104] I s =1.18707*10 7 (V in +119.12241)*(R+1341.51766) 2 -28.09194.

[0105] like Figure 2 The switched capacitor resonant converter shown in the figure has four dominant switching modes, namely Mode 1, Mode 2, Mode 3 and Mode 4. The vout-iL state plane trajectory during the switched capacitor resonant operation of the switched capacitor resonant converter is shown in FIG. Figure 3 shown.

[0106] Mode 1: S1 and S4 are turned on, while S2 and S3 are turned off. The power supply, resonant inductor, switched capacitor, S1, and S4 form a loop. When the polarity of the resonant current is positive, the power supply provides energy to the LC resonant circuit; when the polarity of the resonant current is negative, the LC resonant circuit feeds back energy to the power supply. The output capacitor keeps discharging to the load.

[0107] Mode 2: S1 and S3 are turned on, while S2 and S4 are turned off. The power supply, resonant inductor, switched capacitor, S1, S3, output capacitor, and load form a loop. When the polarity of the resonant current is positive, the power supply and the LC resonant circuit provide energy to the output capacitor and the load; when the polarity of the resonant current is negative, the output capacitor feeds back energy to the power supply and the LC resonant circuit and provides energy to the load.

[0108] Mode 3: S2 and S3 are turned on, while S1 and S4 are turned off. The resonant inductor, switched capacitor, S2, and S3 form a loop. When the current polarity is positive, energy is transferred from the resonant inductor to the switched capacitor; when the current polarity is negative, energy is transferred from the switched capacitor to the resonant inductor. The output capacitor keeps discharging to the load.

[0109] Mode 4: S2 and S4 are turned on, while S1 and S3 are turned off. The resonant inductor, switched capacitor, S2, S4, output capacitor, and load form a loop. When the polarity of the resonant current is positive, the output capacitor feeds back energy to the power supply and the LC resonant circuit and provides energy to the load; when the polarity of the resonant current is negative, the LC resonant circuit provides energy to the output capacitor and the load.

[0110] When vout < Vref, to achieve fast dynamic response, it is necessary to prevent the resonant current from discharging the output capacitor during Mode 2 and Mode 4. That is to say, the polarity of the resonant current should be limited to iL > 0 during Mode 2 and iL < 0 during Mode 4. To achieve the conversion of the resonant current polarity, Mode 1 or Mode 3 is added between Mode 2 and Mode 4. Therefore, the state trajectory circulates in the order of "Mode 1, Mode 2, Mode 3, Mode 4, Mode 1...".

[0111] Similarly, when vout > Vref, to achieve fast dynamic response, it is necessary to prevent the resonant current from charging the output capacitor during Mode 2 and Mode 4. That is to say, the polarity of the resonant current should be limited to iL < 0 during Mode 2 and iL > 0 during Mode 4. Mode 1 or Mode 3 is added between Mode 2 and Mode 4 to achieve the conversion of the resonant current polarity. Therefore, the state trajectory is in the order of "Mode 1, Mode 4, Mode 3, Mode 2, Mode 1...".

[0112] In the dynamic adjustment process, in order to make the states of the switched capacitor resonant converter on the two state planes move according to the planned typical trajectory, it is necessary to design the switching conditions of the switching mode corresponding to the state trajectory. In the dynamic adjustment process, the switching parameter IS is designed to derive the switching conditions in the dynamic adjustment process. Among them, the actual physical meaning of IS is the maximum positive resonant current when switching from Mode1 to Mode2, and -IS is the maximum negative resonant current when switching from Mode3 to Mode4.

[0113] When the resonant current reaches the maximum value IS during Mode 1, the state trajectory reaches the switching point.

[0114] And switch from Mode 1 to Mode 2. The switching condition can be expressed as iL≥IS. When the resonant current reaches the negative maximum value -IS during Mode 3, the state trajectory reaches the switching point and switches from Mode 3 to Mode 4. The switching condition can be expressed as iL≤-IS. In order to adapt to different load conditions, the selection of the controller switching parameter IS depends on parameters such as the input voltage Vin, the output voltage Vo, and the load R. When the converter is lightly loaded, a larger switching parameter IS will result in a larger resonant current stress, affecting the converter power transmission efficiency.

[0115] In order to verify the rationality of the proposed state switching parameter calculation method for state trajectory control of a switching resonant converter. A switching capacitor resonant converter control experimental platform built with EP4CE10F17C8N FPGA chip was used. The resonant capacitor Cr selected in the resonant circuit was 9.4μF, and the resonant inductor Lr was 27μH. Four experimental groups were set up for testing. The proposed state switching parameter calculation method for state trajectory control of a switching resonant converter was used to calculate and set the switching parameters. The input voltages were 20V, 30V, 40V, and 50V, respectively. The output voltages were 10V, 15V, 20V, and 25V, respectively. The load resistances were 5Ω, 10Ω, 15Ω, and 20Ω, respectively. The resonant current change curve, resonant voltage change curve, and output voltage change curve of the test are shown in Figure 2. Figures 4 to 6 shown.

[0116] Depend on Figures 4 to 6 It can be seen that under multiple different conditions, the startup transient waveform of the converter is ideal, the output voltage quickly reaches the preset value, and the overshoot is small. This result verifies the rationality of the proposed state switching parameter calculation method for state trajectory control of the switching resonant converter, which can effectively calculate and set appropriate switching parameters for different situations.

Claims

1. A method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control state, characterized in that: The following steps are involved: S1: Select parameters to design orthogonal experimental table; S2: According to the state of the switch resonant conversion circuit, establish a formula to analyze the input voltage and load; S3: Establish a mathematical model with unknown parameters and input voltage and load; S4: Use the LM algorithm to solve the unknown parameters and obtain the switching parameter calculation formula.

2. A method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control state according to claim 1, characterized in that: Step S2 also includes the following steps: S2-1: Set the steady-state output voltage to a constant value of V o ; S2-2: According to the charge and discharge balance principle of the output capacitor, the average output current I out Equal to the average load current I Load , and the average output current and output voltage satisfy the following relationship: I out =I Load =V o / R; S2-3: Calculate the average value of the resonant inductor voltage of the switching resonant conversion circuit during Mode 2 and Mode 4; According to Kirchhoff's voltage law, the average value of the resonant inductor voltage during Mode 2 is (V in / 2-V o ); The average value of the resonant inductor voltage during Mode 4 is (V o -V in / 2); Among them, V in is the input voltage of the switching resonant converter circuit, V o To maintain a constant value for steady-state output voltage; S2-4: Calculate the inductor current slope during Mode 2 and Mode 4; The average value of the resonant inductor voltage during Mode 2 is (V in / 2-V o ) / L R ; Among them, V in is the input voltage of the switching resonant converter circuit, V o To keep the steady-state output voltage constant, L R is the resonant inductance value; The average value of the resonant inductor voltage during Mode 4 is (V o -V in / 2) / L R ; S2-5: Using the geometric analysis method, the inductor current slope is obtained as: go L / dt=(V in / 2-V o ) / L r =(I T -I S ) / (T sw / 2-T s ) Among them, L / dt is the inductor current slope, I T is the resonant current when switching from Mode 2 to Mode 3, T s is the timing difference between the upper and lower half bridges of the switched capacitor resonant converter; T sw is the switching cycle, I S is the maximum forward resonant current of Mode 1, which is the conditional parameter for switching from Mode 1 to Mode 2. out is the average output current; Average output current I out , switching cycle T sw And the time difference between the upper and lower half bridges T s The following constraints exist: where f sw is the switching frequency and f sw =1 / T sw ; S2-6: Simplify the calculation formula of step S2-5.

3. A method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control according to claim 2, characterized in that: The simplified calculation formula of step S2-6 is as follows: Among them, V in and L r are known circuit parameters, R is the resistance, which can be obtained based on load detection, V o Determined by the input reference value of the controller.

4. The method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control state according to claim 1, characterized in that: Step S3 builds the model as follows: I s =k*(V in +a)*(R+b) 2 +c; Among them, I s is the switching parameter, k, a, b and c are the model parameters.

5. The method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control state according to claim 1, characterized in that: Step S4 also includes the following steps: S4-1: storing the data of multiple sets of input voltage Vin, resistance R and corresponding switching parameter Is in MATLAB vectors respectively; S4-2: According to data and experimental experience, the initial guess value is set to k = 0.000000123381, a=118.338189676784, b=1334.346495829290, c=-28; S4-3: Define mathematical model and establish function; S4-4: define the residual function; S4-5: Fitting model parameters.

6. A method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control according to claim 1, characterized in that: The function established in step S4-3 is as follows: modelFunc=@(params,Vin,R)params(1)*(Vin+params(2)).*(R+ params(3)).^2+params(4); Here, @() means creating an anonymous function, params(1) corresponds to k, params(2) corresponds to a, params(3) corresponds to b, and params(4) corresponds to c.

7. A method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control according to claim 6, characterized in that: Step S4-3 also includes comparing the observed data Is in the orthogonal experiment table in step S1 with the Is value calculated by the model, and calculating the difference between the model prediction value and the observed data value.

8. The method for predicting and calculating state switching parameters of a switching resonant converter state trajectory control state according to claim 1, characterized in that: Step S4-4 also includes the following steps: S4-4-1: Perform nonlinear least squares fitting of the model using the lsqnonlin function of MATLAB; adjust the model parameters k, a, b, and c based on the provided initial guess values ​​so that the difference between the calculated current value and the observed data is minimized; S4-4-2: Select the Levenberg-Marquardt algorithm and estimate the optimal parameters by minimizing the residual function. The following fitting results are obtained: k=1.18707*10 -7 ,a=119.12241,b=1341.51766,c=-28.09194 S4-4-3: Substitute the fitting results into the following calculation formula: I s =1.18707*10 7 (V in +119.12241)*(R+1341.51766) 2 -28.09194。

Images