METHOD FOR CONTROLLING THE INPUT VOLTAGE FREQUENCY OF A DC-DC CONVERTER

Description

[0001] The present invention relates to the field of electric battery chargers, in particular for electric or hybrid motor vehicles.

[0002] More specifically, the invention relates to a method for frequency control of the input voltage of a DC-DC converter for an electric battery charger.

[0003] Electric vehicle battery chargers, more commonly known as chargers, require significant charging power, for example up to 22kW in three-phase, or 7kW in single-phase.

[0004] These chargers generally include two power conversion stages: a first power factor correction stage, more commonly known by its English name Power Factor Correction, usually abbreviated as PFC, ensuring the conversion of AC / DC network voltages to a DC bus, and a second DC-DC conversion stage, known as DCDC, ensuring the control of the output current required to charge the battery as well as galvanic isolation of the charger through a transformer.

[0005] With reference to the figure 1 In prior art, two DC output voltage buses, across the terminals of the output capacitors, are coupled to a DCDC converter each.

[0006] The DCDC can notably be of the LLC type, as shown figure 2 including a 22 transformer providing galvanic isolation of the charger.

[0007] There figure 3 represents a simplified diagram of the DCDC converter assembly of the figure 2 , comprising a capacitor Cr and two inductors Lr and Lm. The input voltage corresponds to the DC bus and the output voltage is the battery voltage. The gain then corresponds to the ratio of the two voltages.

[0008] The first MOSFET bridge of the LLC-type DCDC converter operates with a 50% duty cycle and is frequency-controlled. Frequency control allows the DCDC converter's gain to be adjusted and the DC bus voltage at the charger's input to a predetermined value. Depending on the battery voltage and power demand, the frequency can fluctuate, for example, between 60 kHz and 200 kHz.

[0009] The solutions proposed in the prior art for controlling this type of DCDC converter generally feature output voltage regulation such as that disclosed in the publication DRGOŇ A, Peter, FRIVALDSKÝ, Michal, and SIMONOVÁ, Anna. A New Approach of Control System Design for LLC Resonant Converter. In: MATLAB for Engineers-Applications in Control, Electrical Engineering, IT and Robotics. InTech, 2011 , in which the DCDC output voltage is controlled using the switching frequency. A transfer function between the duty cycle and the output voltage is derived from identification methods using a PSPICE hardware model simulating the dynamics of the output voltage responses to a frequency step. A regulator is then designed based on the previously derived transfer function.

[0010] The transfer function can also be obtained using the so-called "small signal" method, which consists of deducing a transfer function from an excitation around a function point and from the measurement of the DC / DC converter's response, as described in the doctoral thesis of YANG, Bo. Topology investigation of front end DC / DC converter for distributed power system. 2003 . However, this transfer function is only valid at the given operating point and becomes obsolete with each change of operating point. It is therefore necessary to recalculate it each time. Consequently, such a solution is relatively complex to implement and computationally expensive.

[0011] We also know of DC current regulation controls if the output voltage varies over a small range.

[0012] Finally, we also know from the publication FANG, Zhijian, WANG, Junhua, DUAN, Shanxu, et al. Control of an LLC Resonant Converter Using Load Feedback Linearization. IEEE Transactions on Power Electronics, 2018, vol. 33, no. 1, p. 887-898 a regulation built by linearizing control (also called in English feedback linearization This publication describes a 7-state nonlinear model, subsequently reduced to 2 states, and proposes PI loop control for controlling the output voltage of a DCDC LLC. However, such a solution requires complex and costly hardware and software adaptations.

[0013] Sometimes the output voltage is dictated by the battery. Furthermore, particularly in electric vehicle applications, this output voltage often varies over a wide range, for example between 250V and 430V.

[0014] Also, regulation of the input DC voltage as disclosed in document US 2017 / 033701 A1 is desirable, as it allows a DC voltage to be imposed across the capacitors, at the output of the PFC.

[0015] However, regulating the DC voltage at the input of the LLC DCDC converter is a subject for which the prior art provides no satisfactory solution.

[0016] Therefore, a solution exists for quickly and reliably controlling the DC input voltage of the LLC DC-DC converter. We propose a frequency-controlled method for controlling the input voltage of a frequency-controlled LLC DC-DC converter operating with a 50% duty cycle, comprising: a preliminary step of defining a setpoint voltage value, a step of calculating a control frequency value of said DC-DC converter, obtained by mathematical inversion of the expression of the gain of said DC-DC converter, as a function of an output battery voltage, an input power setpoint and said input setpoint voltage; and a step of applying the control frequency thus calculated to said converter.

[0017] Thus, a DCDC input command can be obtained in a relatively simple and quick manner.

[0018] Advantageously, and without limitation, said DC-DC converter is of the resonant series LLC type defined by the parameters of an equivalent circuit comprising two inductors and one capacitor; said control frequency value being a function of said two inductors and said capacitor. Thus, the calculation of the control frequency is obtained by an approximation of the DC-DC operation, simplifying the calculations and making the process faster.

[0019] The step of applying the calculated control frequency includes: the definition of a frequency increment step; a step of initializing the control frequency to an initial control value corresponding to the control frequency thus calculated; the definition of a first threshold value and a second threshold value, the opposite value of the first threshold value and the opposite value of the second threshold value; a step of calculating an error value between a measured input voltage value and said setpoint input voltage; and a step of comparing said error value and said threshold values;the process comprising a control step in which: when said error value is between the first threshold value and the opposite value of the first threshold value, and when said error is greater than the second threshold value or less than the opposite value of the second threshold value, the initial control frequency is incremented by the frequency increment step; when said error value is between the second threshold value and the opposite value of the second threshold value, the control frequency is maintained at its previous value; if none of these conditions are met, the initial control value is applied as the control frequency.

[0020] Thus, the process includes a relatively simple, fast and robust frequency control

[0021] According to a particular embodiment of the invention, the method includes feedback regulation of the control frequency.

[0022] The invention also relates to a device for implementing a process as described above.

[0023] The invention also relates to an electric battery charger comprising a power factor correction stage, at least one DC-DC converter, and a device as described above.

[0024] The invention also relates to a motor vehicle comprising an electric battery charger as described above.

[0025] Other features and advantages of the invention will become apparent from the following description of a particular embodiment of the invention, given by way of example but not limitation, with reference to the attached drawings in which: there figure 1 is a schematic view of an electric battery charger known from the prior art; the figure 2 is a detailed view of a DC-DC converter of a charger according to the figure 1 ; there figure 3 is a simplified diagram of an LLC circuit of a DC-DC converter according to the figure 2 ; and the figure 4 is a flowchart of the control process according to one embodiment of the invention.

[0026] THE figures 1 à 4 relating to the same embodiment, they will be commented on simultaneously.

[0027] With reference to the figure 1 , a charger 1 of electric accumulators 13 connected to a three-phase electrical network 10 includes a power factor correction stage 11, also called PFC stage 11, and DC-DC converters 12a and 12b each comprising an inverter 212.

[0028] The three-phase electrical network 10 is mounted to an input filter 14 transmitting filtered input currents to the PFC stage 11.

[0029] At the output of PFC 11 two DC voltage buses, connected to the terminals of the output capacitors of the PFC 11 stage, are each coupled to a DCDC converter 12a, 12b, connected at the output in parallel to a battery of accumulators 13.

[0030] Each DCDC 12a, 12b, of which only one example is represented figure 2 , includes an input MOSFET bridge 120, an LLC circuit 121, a simplified equivalent representation of which is shown figure 3 , a 22 transformer and a 122 output diode bridge.

[0031] The charger 1 further includes control means 15 for the DC-DC converters 12 capable of implementing a control method 4 according to the invention.

[0032] The control method 4 according to the invention aims to control the frequency of the input voltages of the DC-DC converters 12.

[0033] To this end, the method according to the invention includes calculating a DCDC switching frequency.

[0034] We know, in reference to the figure 3 , that the transfer function of a DCDC LLC according to the invention is of the form: G = ηV bat V dc = V out V in

[0035] With G being the gain of the DCDC transfer function (or at least of the inverter part of the DC / DC going up to the primary of the transformer); η the transformation ratio of the DCDC transformer; Vbatt the voltage across the battery terminals, i.e. the output voltage of the DCDC, Vdc the DC input voltage of the DCDC; And by generic terminology: Vout the output voltage of the DCDC and Vin the input voltage of the DCDC.

[0036] With reference to the figure 3 In a simplified view of DC-DC, the equivalent resistance R of the DC-DC transformer corresponds to the battery charge referred to the primary side of the transformer. Therefore, R is calculated using the following equation: R = 8 π 2 N P N s 2 V bat 2 P

[0037] With NP and Ns respectively the number of turns in the primary and secondary windings of the transformer. P is the power in the primary winding of the transformer and V is the voltage in the secondary winding of the transformer.

[0038] Therefore, we write the transfer function of equation (1) as follows: The transfer function of this circuit is written: V out V in = R L m c r s 2 R L m C r s 2 + L r C r s 2 + 1 R + L m s

[0039] Also, to calculate the gain of the DCDC transfer function, we calculate: G s = V out V in = ηV bat V DC = R L m C r s 2 L m L r C r s 3 + RC r L m + L r s 2 + L m S + R

[0040] This equation (4) is rewritten in terms of the angular frequency ω (ω = 2πf sw ), by setting s=j ω.

[0041] Therefore, the gain equation can be written using the following equations: G s = V out V in = ηV bat V DC = R L m C r jw 2 L m L r C r jw 3 + RC r L m + L r jw 2 + L m jw + R Or G s = V out V in = ηV bat V DC = R L m C r j 2 πf sw 2 L m L r C r j 2 πf sw 3 + RC r L m + L r j 2 πf sw 2 + L m j 2 πf sw + R

[0042] By calculating the transfer gain G, to obtain an expression for the control frequency f sw according to the equation: f sw ω = fct Vbat , Preq , Vdc consigne

[0043] With Vbat the battery voltage, Vdc the input voltage of the DCDC, and Preq a power setpoint at the input of the DCDC.

[0044] Indeed, by replacing V dc in the expression for G(s) with a setpoint value of V dc, we can calculate the frequency at which the DC bus converges to a given voltage, for example 450V.

[0045] The gain G is calculated as the ratio of ηVbat / Vdc, i.e., in this embodiment G = ηVbat / 450V

[0046] We deduce a third-order equation depending on (ω = 2πf sω ), ω 3 + A ω 2 + B ω + C = 0

[0047] With parameters A, B, C as functions of Vbat, Preq, Lm and Lr the inductance values ​​of the equivalent DCDC circuit, and Cr the capacitance value of the equivalent DCDC circuit.

[0048] Solving equation (6) in ω allows calculation by anticipatory control, known in English as feedforward, the control frequency f sw (ω) of the DCDC.

[0049] Due to parametric variations and calculation precision, as well as the simplifying assumptions made in writing the DC-DC transfer function, applying this direct calculation is insufficient to eliminate the steady-state error between the measured DC voltage and the setpoint. However, the error remains insignificant and has a maximum of 30V.

[0050] To address this problem, with reference to the figure 4 A regulator has been added to the previous feedforward circuit. It operates by incrementing or decrementing the frequency until the static error is eliminated, thus further adjusting the initial frequency generated by the previous calculation for improved accuracy.

[0051] The regulator according to the first embodiment is a discrete regulator in which: eps1 is a threshold value from which the frequency increment / decrement begins. eps2 is a threshold value for which the control frequency is fixed.

[0052] Also, according to an example of implementation with reference to the figure 4 , in the first step we calculate 40 the control frequency f sw (ω), also called switching frequency f sw (ω), as described previously, as a function of a setpoint voltage V DC req< , for example 450V, a required power P req< , for example a power between 0< P req< <11kW, and as a function of a battery voltage such as 250V <V bat <430V.

[0053] The control frequency value is initialized to 41. f sw (k) at the initial frequency value f sw_feedforward previously calculated.

[0054] Next, we calculate 44 an error value ε between the setpoint voltage V DC req< and the measured voltage V dc measured< at the input of the DCDC.

[0055] We compare this error value ε to two error threshold values ​​eps1 and eps2.

[0056] If (condition 1) the error ε is between the limits of eps1 and -eps1, for example between 10V and -10V, and if in addition the error ε is greater than eps2 or less than -eps2, these thresholds being for example 5V and -5V, the initial frequency value is incremented by 43 f sw_feedforward by an increment of one frequency increment step ΔF, i.e.: f sw k = f sw _ feedforward + ΔF k being a time integer.

[0057] After this step 43, we loop back to step 44.

[0058] If (condition 2) after step 44 the error ε is between the limits of eps2 and -eps2, we freeze and maintain 45 the value of the frequency f sw (k) which ensures a 5V DC bus near the setpoint at the previous value, i.e.: f sw k = f sw k − 1

[0059] The value f sw (k - 1) being equal à f sw_feedforward if condition 1 has not been met previously, or à f sw_feedforward + k * ΔF if step 45 takes place after k previous steps 43.

[0060] If none of these conditions are met in step 44, the frequency value 46 is used. f sw (k) calculated by feedforward in step 40. This value is updated periodically. The command will continue to apply the frequency calculated by feedforward as long as no error condition is met, with steps 43, 45, and 46 looping back to step 44.

[0061] The invention is not limited to the given example values ​​of error thresholds eps1 and eps2. In particular, eps2 can be set to 1 or 0 V, depending on the feasibility of the operating point.

[0062] This method ensures stable frequency convergence, ensured by the feedforward action, and efficient convergence, thanks to the action of the regulator which finishes canceling the static error and makes the DC bus converge precisely to the setpoint value.

[0063] The invention is not limited to the type of regulator described in the first embodiment. A Proportional-Integral or Proportional-Integral-Derivative type regulator can also be provided, the implementation of which is known to those skilled in the art, although its adjustment is more complex than the regulator of the first embodiment of the invention.

Claims

1. Method (4) for controlling the frequency of the input voltage of an LLC DC-to-DC converter (12) which operates with a duty cycle of 50% and is frequency-controlled, comprising: - a preliminary step of defining a setpoint input voltage value (Vdcreq), - a step (40) of calculating a control frequency value (fsw(ω)) for said DC-to-DC converter (12), and - a step of applying the control frequency calculated in this way to said converter, said control frequency value (fsw(ω)) being obtained by solving a 3rd-order frequency equation expressing the gain of said DC-to-DC converter (12), on the basis of an output battery voltage (Vbat), an input power setpoint (Preq) and said setpoint input voltage (Vdcreq), said method being characterized in that the step of applying the control frequency calculated in this way comprises: - defining a frequency increment step (ΔF); - a step (41) of initializing the control frequency (fsw(k)) to an initial control value (fsw_feedforward) that corresponds to the control frequency calculated in this way; - defining a first threshold value (eps1) and a second threshold value (eps2), and the additive inverse of the first threshold value (-eps1) and the additive inverse of the second threshold value (-eps2); - a step (44) of calculating an error value (ε) between a measured input voltage value (Vdcmeasured) and said setpoint input voltage (Vdcreq); and - a step (42) of comparing said error value with said threshold values (eps1, -eps1, eps2, -eps2); - the method comprising a regulating step during which: • when and as long as said error value is between the first threshold value (eps1) and the additive inverse of the first threshold value (-eps1), and when said error is higher than the second threshold value (eps2) or lower than the additive inverse of the second threshold value (-eps2), the control frequency (fsw(k)) is incremented (43) by the frequency increment step (ΔF); • when said error value is between the second threshold value (eps2) and the additive inverse of the second threshold value (-eps2), the control frequency (fsw(k)) is kept (45) at its previous value; • if none of these conditions is met, the initial control value is applied as the control frequency.

2. Method (4) according to Claim 1, characterized in that said DC-to-DC converter is an LLC series resonant DC-to-DC converter, which is defined by parameters of an equivalent circuit comprising two inductors (Lm, Lr) and a capacitor (Cr); said control frequency value (fsw(ω)) being a function of the values of said two inductors (Lm, Lr) and said capacitor (Cr).