DE102021108698A1 - Method for operating a converter module, converter device with a converter module and motor vehicle with a converter device - Google Patents

Abstract

The invention relates to a method for operating a converter module (K), which provides a coil arrangement (T) for intermediate energy storage and a switch arrangement (G) for controlled rectification. The switch arrangement (G) is operated according to a predetermined switching pattern. According to a predetermined prognosis algorithm (F), a switching time (tE) for switching through the switch arrangement (G) is adapted for the predetermined switching pattern. According to the prognosis algorithm (F), a time profile of a channel voltage (Vs) that can be provided at the switch arrangement (G) is predicted for a respective half-wave of a coil voltage applied to the coil arrangement (T). The prediction is made by calculating a time profile of a coil current (ilk), which is provided by the coil arrangement (T) as a function of the applied coil voltage, and the switching time (tE) is determined as a function of the coil current profile (ilk) and the switch arrangement ( G) is driven to switch through.

Description

The present invention relates to a method for operating a converter module. The invention also relates to a converter device with a corresponding converter module and a motor vehicle with such a converter device.

In the present sense, the converter module means, in particular, a voltage converter for converting electrical voltage. Such a converter module can be used in a motor vehicle, for example, in order to enable electrical energy transmission in an on-board network with different reference potentials. For example, there are motor vehicles in which a converter module is used to convert electrical energy between a high-voltage vehicle electrical system and a low-voltage vehicle electrical system.

A rectifier or a DC voltage converter (DC/DC converter), for example, can be used as a converter module in an on-board electrical system. This can be designed, for example, as an inductive converter with active rectification on the secondary side or on the output side. As a result of the rectification, an electrical DC voltage can thus be provided by the converter module at a connection. In order to enable active or controlled rectification, for example semiconductor switches can be interconnected in a suitable electrical configuration to form a rectifier circuit. The semiconductor switches can be operated in a switching mode according to a predetermined switching pattern.

En secondary-side rectifier of an inductive energy transmission system is, for example, from WO 2013/075896 A2 known. In this case, the rectifier uses a B2 bridge circuit with only one switching means in order to regulate a target output voltage for an energy transmission system.

From the DE 10 2015 107 960 A1 an active rectifier for a load circuit is also known. In this case, the active rectifier is controlled with a controller in order to modify an impedance at the input of the rectifier.

Furthermore, from the DE 11 2014 006 828 T5 discloses an active rectifier for a rotary electric machine, in particular for an alternator. The active rectifier uses power MOSFETs (Metal Oxide Semiconductor Field Effect Transistors) to rectify AC currents to DC currents for charging batteries on vehicles.

It is the object of the present invention to reduce a power loss of the converter module when operating an inductive converter module with controlled rectification.

This object is solved by the subject matter of the independent patent claims. Further possible configurations of the invention are disclosed in the dependent claims, the description and the figures.

The invention is based on the finding that in conventional inductive converters with controlled rectification, an inductive time delay between a voltage and a current in the converter for controlled rectification is not taken into account. Instead, the active or controlled rectification takes place in conventional converter modules, as described above, for example only as a function of a change in an input impedance or a load change on the converter module.

This time delay is caused in particular by the electrical properties of a coil arrangement that is used for intermediate energy storage in the converter. A pulsed coil voltage, ie an electrical alternating voltage applied to the coil arrangement, is transmitted through the coil arrangement to a rectifier circuit for controlled rectification, delayed in relation to a coil current flowing through the coil arrangement. If the rectifier circuit is operated according to a conventional switching method, it can happen that an electrical voltage is already present in the rectifier circuit although a respective semiconductor switch of the rectifier circuit is not yet switched through, ie is electrically conductive. An electric current then flows, for example, along a so-called body diode, which is intrinsically provided in a structure of semiconductor switches. This body diode has a much higher electrical resistance than a semiconductor switch when switched on. The idea now is to determine the time shift caused by the inductance of the converter module and to take it into account in the controlled rectification.

For this purpose, the invention provides a method for operating a converter module, which provides a coil arrangement for conversion for intermediate energy storage and a switch arrangement for controlled rectification. The converter module is thus designed as an inductive converter with at least one electrical inductance. The coil arrangement can be, for example, an individual inductor or a transformer. Accordingly, the converter module can be designed, for example, as a rectifier or as a DC voltage converter. The switch arrangement is the aforementioned rectifier circuit. For the controlled rectification, this can comprise at least one, that is to say one or more, half-liter switches in a suitable electrical circuit. A semiconductor switch may be, for example, a MOSFET, an IGBT (Insulated Gate Bipolar Transistor), a thyristor, or any other predetermined electronic switch.

In order to implement the controlled rectification, the switch arrangement is operated, for example, in a switching mode according to a predetermined switching pattern. A pulse duty factor or a duty cycle can be specified by the respective switching pattern. For exactly one switching cycle of the respective switch, the pulse duty factor specifies a ratio of an on period compared to an off period of the switch arrangement. The switching pattern thus specifies a switching time for switching over between an on and an off state of the switch arrangement. The switched-on or switched-through switching state means a switching state of the switch arrangement in which an essentially unhindered electric current flow through the switch is possible. In contrast to this, the switched-off or deactivated switching state means that the switch arrangement or the respective switch provides a high electrical resistance, so that a current flow in the rectifier circuit is only possible via the aforementioned body diode, for example.

Based on the predetermined switching pattern, the invention now provides that a switching time for switching through the switch arrangement is adapted according to a predetermined prognosis algorithm. In comparison to the predetermined or original switching pattern, a new switching time should therefore be determined. For this purpose, according to the prognosis algorithm, a time profile of a channel voltage that can be provided at the switch arrangement is predicted for a respective half-wave of a pulsed coil voltage that is applied to the coil arrangement. In the present case, “providable” means that the channel voltage can be applied to or tapped from the switch arrangement or the respective semiconductor switch. To predict the course of the channel voltage, a course of a coil current over time, which is provided by the coil arrangement as a function of the coil voltage applied, is calculated or simulated. Depending on the calculated coil current curve, the aforementioned switching time for switching through the switch arrangement is then determined. Finally, the switch arrangement is driven to switch through depending on the determined switching time.

In other words, the aim is to predict the channel voltage for the rectifier circuit by drawing conclusions about the time profile of the channel voltage based on the simulated coil current profile. This results in the advantage that in the controlled rectification, the inductive delay that results from an inductive converter is taken into account. In particular, the rectifier circuit is switched, for example, precisely when the channel voltage is actually present at the switch arrangement. When a channel voltage is applied, a channel current can flow through a channel of the switch provided by the switch arrangement or the respective semiconductor switch, and not through the aforementioned body diode. This results overall in fewer conduction losses when operating the rectifier circuit, and the efficiency of the converter module can be improved.

As used herein, “channel” means a portion of a semiconductor switch configured to conduct current. The channel voltage can thus be, for example, a collector-emitter voltage or a drain-source voltage, depending on a configuration of the respective semiconductor switch. This also applies analogously to the channel current. In the present case, the term “half-wave” means a curve profile of the coil voltage provided as an alternating voltage in half a period between two adjacent zero crossings. The half-wave can also be referred to as half-oscillation.

To implement the switching pattern, the switch arrangement can be controlled, for example, by a control device with a PWM signal. The control device can be, for example, a gate driver for controlling a gate connection or a base connection of the respective semiconductor switch.

As described above, the converter module can be embodied, for example, as a rectifier or DC/DC converter. In the configuration as a rectifier, the pulsed coil voltage applied to the coil arrangement can be provided, for example, directly in the form of an electrical AC voltage as the input voltage of the converter module. In the design as a DC/DC converter, the pulsed coil voltage can be generated, for example, by means of an inverter circuit. In particular, the converter module can be operated bidirectionally. This means that the converter module can be used, for example, to convert electrical energy from an input connection to an output connection or vice versa, depending on the operating mode. A so-called up-converter operation (boost converter mode) or a down-converter operation (back converter mode) can be provided as the operating mode, for example. For example, when using a converter module in an on-board network of a motor vehicle, electrical energy can be transmitted from the high-voltage on-board network to the low-voltage on-board network and vice versa.

For example, in a boost converter operation as exemplified above, the channel voltage leads the coil current due to the electrical properties of the coil assembly. In this case, the switching point in time for switching the switch arrangement on can thus be selected earlier, for example, than a switching point in time that would result according to the original switching pattern. In contrast, the channel voltage lags the inductor current in the buck converter operation as previously described. A switching time for switching through the switch arrangement can thus be selected later than a switching time that is predetermined by the original switching pattern.

In the following, embodiments of the invention are described which result in additional advantages.

In the following specific embodiment, the aforementioned prognosis algorithm is first discussed in more detail. It is provided that a mean value of an input voltage, an output voltage and an output current of the converter module and a duty cycle of the switch arrangement predetermined by the switching pattern are determined or measured as input parameters for the prognosis algorithm. Depending on the input parameters, an initial value of the coil current is then determined in the prognosis algorithm and the initial value is set as a setpoint value for the coil current. Since the input parameters for the prognosis algorithm depend on the respective operating mode of the converter module, the initial value results in particular taking into account the operating mode of the converter module.

Furthermore, a predetermined calculation routine is carried out, in which the course of the coil current over time is determined on the basis of the initial value. In the calculation routine, the half-wave of the pulsed coil voltage is divided into at least two phases, ie two or more phases. The respective phase can, for example, represent a predetermined transmission behavior of the components in the converter. An associated energy storage state of the converter module is calculated for each of the phases and then a value profile of the coil current associated with the respective energy storage state is determined. The course of the value of the coil current in the respective phase is thus determined between a start time and an end time of the respective phase. The course of values can be present, for example, as a pair of values, for example as the start and end value of the coil current at the start time and end time of the phase. Alternatively, the course of values of the coil current can be, for example, an actual course of the coil current over time, which can be calculated, for example, by using a predetermined time function.

Finally, according to the prognosis algorithm, the switching point in time for the switch arrangement is determined as the point in time at which the coil current assumes a value over time that corresponds to the specified setpoint value. This is because the aforementioned channel voltage correlates with the coil current, depending on an operating mode of the converter module, so that conclusions can be drawn about the channel voltage from the coil current.

Accordingly, the aforementioned initial value can be, for example, an initial value of the coil current in a first phase. When passing through a half-wave of the coil voltage, i.e. when passing through the following phase(s), the coil current will reach the initial value again. The time at which the coil current has risen or fallen back to the initial value should now be selected as the switching time.

According to a further embodiment, a ripple current or current ripple is determined from the input values for the half-wave in order to determine the initial value of the coil current. The ripple current is a current which occurs at the output of the switch arrangement during operation and can therefore be referred to as the switch arrangement output current. In this case, the ripple current indicates an oscillation behavior of the switch arrangement output current around the mean value of the output current of the converter module. The current ripple is therefore an alternating component which the output current of the rectifier circuit has. This occurs in a known manner during rectification and is also referred to as superimposed alternating current. Since the structure of the converter module is known per se, the initial value can be deduced from the current ripple using known circuit analysis methods.

According to a further embodiment, in connection with the aforementioned calculation routine, the phases are defined as a function of a transmission behavior of the coil arrangement for a respective value range of the coil voltage within the half-wave. The transmission behavior indicates, for example, whether or not power is currently being transmitted by means of the coil arrangement. The transmission behavior and thus the number of phases can result, for example, as a function of the aforementioned design of the converter module. When an AC voltage is applied directly as the input voltage to a rectifier, the transmission behavior in a half-wave can be divided into a so-called freewheeling and a so-called power transfer, for example. Thus, the half-wave can be divided into two phases, for example. A transition point between the two phases forms a maximum value of the half-wave in terms of absolute value, ie a maximum value for a positive coil voltage and a minimum value for a negative coil voltage.

In an embodiment as a DC/DC converter, according to a further embodiment, the coil voltage applied to the coil arrangement is provided by means of an inverter circuit of the converter module. The phases for dividing the half-wave are selected as a function of a switching configuration of the inverter circuit.

To provide the coil voltage, the inverter circuit can include, for example, a number of semiconductor switches in a suitable electrical configuration. For example, the semiconductor switches can be connected to one another in a bridge circuit, in particular as a so-called B4 full bridge. By suitably operating the semiconductor switches in switching operation, an electrical AC voltage can be generated from an input DC voltage as a coil voltage for applying the coil arrangement. For this purpose, the semiconductor switches are operated in the switching mode and are switched to the switched-on or switched-off state according to a respectively predetermined switching pattern. For example, the same switching logic can be used for the switching patterns of the inverter circuit and the rectifier circuit, so that the switching patterns essentially correspond.

The switching configuration now describes how the semiconductor switches of the inverter circuit are switched in relation to one another in a predetermined time interval. A phase corresponds to the time interval in which such a switching configuration is maintained. If the switching configuration changes, this is to be understood as a transition to the next phase. In the inverter circuit, for example, four such switching configurations can be provided to generate the half-wave of the AC voltage, so that the half-wave is divided into four phases.

According to a further embodiment, a predetermined switching delay time of the inverter circuit is taken into account in connection with the inverter circuit when determining the respective switching time for the rectifier circuit. The switching delay time can be, for example, a predefined dead time of the semiconductor switches of the inverter circuit. This can be specified by a manufacturer in a data sheet, for example. This dead time is specified, for example, in order to delay the switch-on or switch-off times of the semiconductor switches of the inverter circuit, ie to offset them in time in order to avoid an electrical short circuit in the inverter circuit. This is because different energy storage states can arise within the converter module within the dead time, so that the calculation of the coil current can change.

According to a further embodiment, it is provided that a value is selected as the desired value depending on the operating mode of the converter module which is predetermined or known by the input values and in which the channel voltage is present at the switch arrangement. The point is that the channel voltage has a value greater than or less than 0. In particular, the target value can be selected in such a way that the channel voltage exceeds a predetermined limit. An optimal conductivity of the switch arrangement can thus be ensured.

According to a further embodiment, it is provided that a rectifier circuit with at least two semiconductor switches is provided as the switch arrangement. The rectifier circuit can thus be designed, for example, as a synchronous rectifier with two semiconductor switches connected in a half-bridge circuit (B2 bridge). The at least two semiconductor switches are operated according to the aforementioned switching pattern. In this case, the switching patterns for the two semiconductor switches are applied to the semiconductor switches according to a predetermined switching logic with a predetermined switching offset, ie a time delay relative to one another. In this case, the switching offset represents the aforementioned pulse duty factor, ie the duty cycle of the semiconductor switches. According to the prognosis algorithm, a switching time for each of the semiconductor switches is adjusted.

In order to implement the respective switching pattern for the semiconductor switches of the rectifier circuit and the inverter circuit, two logic signals offset in time by the pulse duty factor can be used, for example. The respective switching position, that is to say the respective switching state of the semiconductor switches, then results according to the predetermined switching logic by applying a logical operator to one or both of the logic signals. The switching patterns for the semiconductor switches of the rectifier circuit can result, for example, from a predetermined logical combination of the two logic signals.

According to a further embodiment, the switching time determined by means of the prognosis algorithm is compared with a switching time provided according to the switching pattern in a controller routine to adapt the switching time for the switch arrangement and a difference between the two switching times is regulated or set to zero. For example, a conventional regulation of the converter module for regulating the output voltage can be expanded to include a further regulator routine for adjusting the switching time. A linear controller, for example, can be used to set or regulate the difference. A linear controller can be designed as a PI or a PID controller, for example.

The invention also relates to a converter device with the converter module according to the invention or with one of the aforementioned possible embodiments of the converter module. To operate the converter module, the converter device also includes a control device that is designed to operate the converter module by executing a method as described above.

The invention also relates to a motor vehicle with a converter module as described above. The converter device is used to convert electrical energy between a high-voltage electrical system and a low-voltage electrical system of the motor vehicle. The converter module of the converter device can thus in particular be designed to be bidirectional and have different converter modes.

Further features of the invention can result from the claims, the figures and the description of the figures. The features and feature combinations mentioned above in the description and the features and feature combinations shown below in the description of the figures and/or in the figures alone can be used not only in the combination specified in each case, but also in other combinations or on their own, without going beyond the scope of the invention to leave.

• 1 ein schematisches Prinzipschaltbild eines Wandlermoduls, bei welchem ein Schaltzeitpunkt zum Durchschalten einer Schalteranordnung mittels eines Prognosealgorithmus anpassbar ist, und
• 2 eine schematische Darstellung einer Regelung zum Einregeln des Schaltzeitpunkts durch Nutzen des Prognosealgorithmus.

The drawing shows in:

• 1 a schematic basic circuit diagram of a converter module, in which a switching time for switching through a switch arrangement can be adapted by means of a prognosis algorithm, and
• 2 a schematic representation of a control for adjusting the switching time by using the prognosis algorithm.

In the figures, identical and functionally identical elements are provided with the same reference symbols.

1 shows a schematic basic circuit diagram, ie a circuit diagram of a converter module, which is designed here as a converter K or DC voltage converter. The converter K is shown by way of example as a so-called phase-shifted full-bridge converter (PSFB, full-bridge forward converter with phase shift). Such a phase-shifted full-bridge converter can be used, for example, in a Motor vehicle are used to convert electrical energy between a high-voltage electrical system and a low-voltage electrical system.

For this purpose, the converter K has an input connection A1 to which the high-voltage vehicle electrical system can be connected. In addition, the converter K has an output connection A2 to which the low-voltage vehicle electrical system can be connected. The high-voltage vehicle electrical system can provide the converter K with an input voltage Vin via the input connection A1. The input voltage Vin is a direct electrical voltage and may have a setpoint or nominal value of 400 V, for example. The input voltage Vin can be converted into the output voltage Vo by means of the converter K. Of course, depending on the converter topology of the converter K, conversion in the opposite direction is also possible. The output voltage Vo is also a DC voltage which, depending on a converter topology, can have a higher, lower or inverse voltage level to the input voltage Vin. In the present case, for example, a desired value of 12 V can be provided for the output voltage Vo. The output voltage Vo can be provided via the output connection A2 to the low-voltage vehicle electrical system to supply one or more electromechanical consumers of the motor vehicle.

The converter K according to 1 is designed as an inductive converter. The converter K accordingly has a coil arrangement for the energy transmission or the voltage conversion. The coil arrangement is shown here as a transformer T, through which a galvanic isolation is implemented. Due to the galvanic isolation, systems with different reference potentials GND1, GND2 can be connected, for example, as is the case here. Because of the transformer T, the converter K thus has a primary side P, which is set to the reference potential GND1 of the high-voltage vehicle electrical system, and a secondary side S, which is set to the reference potential GND2 of the low-voltage vehicle electrical system.

For energy transmission, the transformer T has a primary winding or primary coil on the primary side P and two secondary windings or secondary coils with a center tap on the secondary side S. An electrical property of the transformer T, ie an effect on the present converter topology, is in accordance with 1 represented by an electrical inductance, namely the leakage inductance Lk. With the leakage inductance Lk, for example, the so-called leakage flux of the transformer T is simulated or simulated during operation. An electric current that occurs during operation of the transformer T due to the leakage inductance Lk on the primary side P is in 1 represented by the coil current ilk.

For the operation of the transformer T, the input connection A1 is connected to the transformer T on the primary side P via an electrical input capacitance Cin and an inverter circuit W. The input capacitance Cin is in the form of a so-called intermediate circuit capacitor for temporarily storing the input voltage Vin. The inverter circuit W includes four switches or switching elements Q1, Q2, Q3, Q4, which are interconnected in a full bridge circuit. An electrical resistance exhibited by the primary side P is represented by the primary resistance Rp.

On the secondary side S, the secondary windings of the transformer T are connected via a switch arrangement and a filter circuit F to the output terminal A2. In the present case, the switch arrangement is designed as a rectifier circuit G for the controlled rectification of an AC voltage transmitted by the transformer. For this purpose, the rectifier circuit G according to 1 two switching elements Q5 and Q6 connected as a midpoint rectifier and connected to the secondary windings. The filter circuit F is connected to the rectifier circuit G via the center tap of the secondary windings. In the present case, the filter circuit F is formed by an LC element. In this case, the filter circuit includes a filter inductance Lf, at which a filter current iL can be tapped. In addition, the filter circuit F includes a filter capacitance Co, at which a filter capacitance current ic can be tapped. An internal resistance or parasitic resistance Rc of the filter capacitance Co is in 1 symbolized by the resistance Rc. The output connection A2 is thus implemented by the center tap and a reference potential tap at which the reference potential GND2 is provided. The output voltage Vo and the output current lo can be tapped or provided at the output connection.

As in 1 shown, the switching elements Q1-Q6 of the inverter circuit W and the rectifier circuit G are designed as semiconductor switches, in the present example as MOSFETs. Of course, any other type of semiconductor switch can be used as an alternative. In 1 the switching elements are supplemented by their respective intrinsic capacitances C1-C6 and their respective intrinsic body diodes D1-D6 to represent their electrical properties.

To operate the converter K, the switching elements Q1-Q6 can be operated in a known manner according to a predetermined switching logic in a respectively predetermined switching pattern. The switching pattern allows the switching elements Q1 - Q6 to be switched between on and off switching states. In the switched-on switching state, the respective switching element Q1-Q6 is switched through and has a low electrical resistance R1-R6. In particular, the respective switch resistance R1-R6 is very much lower than a diode resistance or forward resistance that the body diode D1-D6 of the respective switch Q1-Q6 has. This enables a current to flow through a respective channel (channel current) of the switch Q1 - Q6. In the present case, channel means a conductive section between a drain and source connection of the respective switch Q1-Q6. When switched off, the respective switching element Q1 - Q6 has a high electrical resistance, so that the flow of current through the channel is blocked. The current can then flow away via the respective body diode D1-D6, for example.

The respective switching pattern for the switches Q1 - Q6 can be generated, for example, by two predetermined logic signals. The logic signals have the same profile and are phase-shifted relative to one another by a predetermined time offset. This time offset is referred to as the duty cycle or duty cycle. The respective switching pattern can now result, for example, by applying a logical operator to one of the logic signals or by logically linking the two logic signals. The duty cycle is thus also set for the respective switching pattern of the switches Q1-Q6. In the case of the inverter circuit W, in particular a predefined dead time Td can be taken into account for determining the switching pattern from the logic signals. The dead time Td is a time offset specified by the manufacturer, for example, which should be observed when switching the switches Q1 - Q4 in order to avoid a short circuit. The respective switching pattern can be provided to the respective switching element Q1-Q6 by means of a control device, such as a so-called gate driver.

By suitable switching according to the respective switching pattern, an electrical AC voltage can now be generated with the inverter circuit W from the input voltage Vin. This AC voltage is applied to the transformer T for energy transmission, which is why the AC voltage is also referred to as the coil voltage. Depending on the aforementioned leakage inductance Lk, the coil current ilk now sets itself on the primary side P. Electrical energy or power is transmitted to the secondary side S by magnetic coupling of the transformer T, depending on the coil current ilk and the coil voltage. On the secondary side S, the switches Q5 and Q6 of the rectifier circuit G are also operated in the respective switching pattern. As a result, an AC voltage induced by the power transmission from the primary side P to the secondary side S can be rectified into an electrical DC voltage. The rectified voltage is then smoothed or filtered by means of the filter arrangement F and is made available as an output voltage Vo at the output terminal A2.

In conventional converters K, the switching pattern for switching the switches Q1-Q6 is generally set as a function of an impedance of the converter K at the input or a load change at the output. The aim is to keep the output voltage Vo constant. In this conventional operation, however, the electrical property of the coil arrangement, that is to say here of the transformer T, is not taken into account. Because of the leakage inductance Lk, the coil voltage and the coil current are made available with a time delay, ie out of phase. As a result, it can happen that a channel voltage Vs is already present on the secondary side S at the respective switch Q5 and/or Q6, although the respective switch is not yet switched through. The current flows through the associated body diode D5, D6, and since this has a higher electrical resistance than the switched-through switch Q5, Q6, additional power loss is generated.

In order to avoid such losses when operating the converter K, as in 2 shown, a predetermined prognosis algorithm F is carried out in order to adapt a switching time for switching through the switch arrangement in the respective switching pattern. According to the prognosis algorithm F, a time course of the channel voltage Vs is predicted or forecast for a respective half-cycle of the coil voltage with which the coil arrangement, ie the transformer T is acted upon. For the prediction, a time profile of the coil current ilk is calculated or, in particular, estimated by mathematical calculations. The switching time is then determined as a function of the coil current profile ilk and the rectifier circuit G is controlled to switch through. A half-wave is to be understood as a curve of the pulsed coil voltage between two zero crossings. A point in time at which the coil voltage has passed through zero is therefore always selected as the beginning of the calculation or the starting point in time for the prognosis algorithm F.

Mean values of the input voltage Vin, the output voltage Vo and the output current Io of the converter K are determined as input parameters H for the prognosis algorithm F. In addition, the duty cycle d is provided as a further input parameter H. The prognosis algorithm gives A as the output parameter 2 for example a time difference Δt. The time difference Δt represents the time shift between the switching time according to the conventional switching pattern and the adjusted switching time tE, which was calculated according to the forecasting algorithm F. In order to set or regulate the respective switching times tE for the switches Q5 and Q6, as in 2 shown, a scheme can be used. For the control, the time difference Δt is compared with a setpoint difference Δtsoll. The setpoint difference Δtsoll is, in particular, zero. The aim of the regulation is therefore to regulate the time difference Δt to zero. Depending on the comparison, a resulting system deviation is provided to a rectifier controller CΔt, which adjusts the switching time tE for the rectifier circuit G or its switches Q5 and Q6 as a control variable depending on the system deviation. The rectification controller CΔt can be, for example, a linear controller or a linear or continuous controller, such as a PI or PID controller.

The following is a code example for the prediction algorithm F:

In the prognosis algorithm F, an initial value of the coil current ilk as a function of the input parameters H is first determined. This initial value is then set as a target value for the coil current ilk. The initial value is determined by calculating a ripple current Ir. The ripple current Ir is an estimated curve calculated from the input parameters H for the current curve of the filter current iL, which is provided by the filter inductance Lf on the secondary side S for the respective half-cycle of the coil voltage. For example, the ripple current Ir can be calculated as follows: $$I\mathrm{right}=\frac{\frac{Ts}{2}\left(VO-\frac{1}{n}Vin+lO\left(Rsek+\frac{Rp\mathrm{right}im}{n^2}\right)\right)\left(\mathrm{i.e}L-\mathrm{i.e}+\frac{2T\mathrm{i.e}}{Ts}\right)}{\left(Lƒ+\frac{1}{n^2}Lk\right)}$$

Here, Ts is a duration of a complete switching period of the respective logic signal used to determine the respective switching pattern for the switches Q1 - Q6. n is a turns ratio of the primary and secondary windings of the transformer T and dL is a so-called duty cycle loss. The duty cycle loss dL describes a duty cycle loss that results in comparison to the original duty cycle d. Rprim and Rsek are resistance values obtained for the primary side P and the secondary side S at the beginning of the measurement, where Rprim = Rp + Rdsp and Rsek = Rs + Rdss. Rdsp is an electrical resistance exhibited by the switches Q1-Q4 in the on-resistance state, and Rss is an electrical resistance exhibited by the switches Q5 and Q6 in the on-resistance state.

The initial value can be calculated from the ripple current Ir as follows, for example: $$ilk=\frac{1}{n}\left(IO-\frac{I\mathrm{right}}{2}\right)$$

The initial value thus depends on the mean value of the output current Io and the calculated current ripple Ir, and is transformed to the primary side via the turns ratio n. In particular, the initial value results depending on an operating mode in which the converter K is operated. The operating mode can be, for example, a step-up converter mode (boost converter mode) or a step-down converter mode (buck converter mode) and describes an energy transmission direction in the converter K. The operating mode is known in particular by the input parameters H, i.e. a ratio of the input voltage Vin to the output voltage Vo known.

The desired value of the coil current ilk is selected in particular in such a way that the channel voltage Vs is present at the respective switch Q5, Q6, depending on the operating mode, when the coil current ilk reaches the desired value. By "presence" it is meant that the channel voltage has a value greater than or less than zero.

After the calculation of the ripple current Ir, a calculation routine is carried out in the prognosis algorithm F. In the calculation routine, the course of the coil current ilk over time is determined on the basis of the initial value. For this purpose, the half-wave of the coil voltage under consideration is subdivided into individual phases. The phases represent a predetermined transmission behavior of the components in the converter K. Since the converter K in the present case uses the inverter circuit W to generate the coil voltage as AC voltage, the phases are selected according to a respective switching configuration of the switches Q1-Q4. A phase is thus to be understood as a period in which a predetermined switching configuration of the switches Q1 - Q4 is maintained, while a phase transition between two phases is to be understood as a change in the switching configuration. Since four switching configurations of the switches Q1-Q4 are necessary to generate the half-wave for the coil voltage, the half-wave is divided into four phases p1, p2, p3 and p4.

A respective energy storage state of the converter K is then determined for each of the phases p1-p4. The energy storage state relates to an energy storage behavior that the energy-storing components of the converter K have in the respective phase p1-p4. These include, for example, the intrinsic MOSFET capacitances C1 - C6 and the leakage inductance. It is therefore a question of determining whether and how the energy-storing components in the respective phase p1 - p4 affect the coil current ilk. The particular energy storage state may depend on possible diode conduction. "Diode conduction processes" means which of the body diodes D1 - D6 of the converter K is currently conducting current or has current flowing through it.

An assigned value profile of the coil current ilk can then be estimated from the energy storage states. Depending on the phase, the course of values can be present, for example, as a pair of values at the start time and end time of the phase or as a course over time by applying a predetermined time function. In the present case, the last two phases p3 and p4 are particularly relevant for determining the switching time tE for switching on the switches Q5 and Q6. Thus, in these phases, for example, the time profile of the coil current ilk can be determined, while in phases p1 and p2 it is sufficient to know the start and end values at the start time and end time of the respective phase.

Based on the course of the coil current ilk determined in this way, the switching time tE can finally be determined as the point in time at which the coil current ilk assumes a value over time that corresponds to the specified setpoint value, ie the initial value. This exploits the fact that the channel voltage Vs, which can be tapped off at the respective switch Q5, Q6, correlates with the coil current ilk through the known transfer behavior of the transformer T. As a result, the course of the channel voltage Vs can be deduced from the course of the coil current ilk.

The prognosis algorithm F can, for example, be carried out for each individual half cycle of the coil voltage. In particular when a calculation time of the prognosis algorithm F is shorter, that is to say the duration of a half-wave curve. Alternatively, for example, the prognosis algorithm F can only be executed for every xth half-wave of the coil voltage profile, for example for every second, or every fifth, or every tenth half-wave. Overall, it is only necessary to ensure that the calculation time is kept shorter than the dynamics of the converter K when it is operated in a load system, such as the vehicle electrical system mentioned at the outset. The dynamic range in this area of application is typically around 20 kHz, so the calculation should take no more than 50 µs. In tests, for example, calculation times of 2 - 3 µs were achieved.

Calculation steps for calculating the coil current curve ilk in the individual phases p1-p4 are described in more detail below by way of example. In the present exemplary embodiment, state space calculations, numerical calculations using Newton's method and an initial value set and final value set calculation using the Laplace transformation are used to calculate the coil current profile ilk. To calculate the coil current profile ilk, the transmission system provided by the converter K can be described, for example, by a state space model according to the following state space equations: $$\begin{array}{l}{\text{A}}_{\text{x}}s\text{x}={\text{A}}_{\text{x}}\text{x}+{\text{B}}_{\text{and}}\text{and}\\ \begin{array}{cc}\begin{array}{l}\text{A}={\text{A}}_{\text{x}}{}^{-1}{\text{A}}_{\text{x}}\text{B}={\text{A}}_{\text{x}}{}^{-1}{\text{B}}_{\text{and}}\\ \text{sx}-x_0=\text{axe}+\text{Bu}\\ \text{y}=\text{Cx}+\text{You}\end{array}& \end{array}\end{array}$$

In the first phase p1, the switches Q1, Q5 and Q6 are closed, ie switched through, and are therefore represented by their respective switch resistances R1, R4 and R6. The switches Q2, Q3, and Q4 are open, ie switched off. In the first phase p1 their influence can be seen through their respective body diodes D2, D3 and D4 and their respective capacitances C2, C3 and C4. The first phase p1 thus includes a dead time Td of the inverter circuit W and can be referred to as a dead time phase.

The state space description according to the above state space equations for the first phase p1 is as follows: $$\begin{array}{l}{\text{A}}_{\dot{\text{x}}}=\left[\begin{array}{cccc}0& -\left(R\mathrm{i.e}sp+Rcp\right)\mathrm{<figure-callout\; id="0"\; label="C\; p"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">C</figure-callout>}p& 0& 0\\ -Rc\mathrm{<figure-callout\; id="0"\; label="p\; C\; p"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">p</figure-callout>}Cp& 0& -Rc\mathrm{<figure-callout\; id="0"\; label="p\; C\; p"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">p</figure-callout>}Cp& 0\\ 0& RcpCp& -RcpCp& -Lk\\ \mathrm{<figure-callout\; id="0"\; label="C\; c\; p"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">C</figure-callout>}cp& 0& -Ccp& 0\end{array}\right]\\ {\text{A}}_{\text{x}}=\left[\begin{array}{cccc}0& 1& 0& R\mathrm{i.e}s\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">p</figure-callout>}\\ 1& 0& 1& 0\\ 0& -1& 1& Rp+\frac{n^2}{4}R\mathrm{i.e}ss\\ 0& 0& 0& -1\end{array}\right]\text{Bu}=\left[\begin{array}{ll}-1\hfill & 0\hfill \\ -1\hfill & 0\hfill \\ 0\hfill & \frac{n}{4}R\mathrm{i.e}ss\hfill \\ 0\hfill & 0\hfill \end{array}\right]\\ \text{x}=\left[\mathrm{and}2\mathrm{and}3\mathrm{and}4ilk\right]\text{and}={\left[ViniL\right]}^{\text{T}}\\ {\text{x}}_0={\left[VinVin0\frac{1}{n}\left(iL\right)\right]}^{\text{T}}\end{array}$$

Rcp is a primary resistance of the switch capacitances Cp. It is assumed that all switches Q1 - Q6 have the same switch capacitance. For the first phase p1, the initial states are given in the vector x ₀ . Capacitors C2 and C3 are charged at the beginning of phase 1, while capacitor C4 is fully discharged. The initial value of the coil current ilk through the leakage inductance Lk thus corresponds to the maximum value of the filter current iL=Io+Ir/2 reduced by the turns ratio n.

In order to determine the associated initial value uQi (t0), with i ∈ {2, 3, 4} for the channel voltage Vs of the respective switches Q2, Q3 and Q4, the possible diode conduction processes are next determined in the prognosis algorithm F on the basis of the state space descriptions. Diode D2 conducts when uQ2 (t0) = u2 + (s u2 - u2,0) CpRcp < Vd, where Vd is the forward voltage of diode D2. The same applies to the diodes D3 and D4. The following initial values uQi (t0) result for the respective switch voltage uQi by applying the Laplace final value theorem calculation: $$\begin{array}{l}\mathrm{and}Q2\left(t0\right)=Vin-\frac{1}{2}Rcpilk\\ \mathrm{and}Q3\left(t0\right)=Vin-ilkR\mathrm{i.e}sp\\ \mathrm{and}Q4\left(t0\right)=\frac{1}{2}Rcpilk\end{array}$$

In the event that one of these voltages is less than the negative forward voltage Vd, the associated diodes D2, D3, D4 conduct. Since the on-state resistance Rdsp of the diode D3 is negligibly small, the diode D3 will generally not conduct in phase 1, so that only the capacitance C3 is relevant. For a negative coil current ilk (ilk<0), which typically occurs in the step-up operation of the converter K, the diode D4 conducts. In step-down converter operation, ie when the coil current ilk is positive (ilk>0), in particular with a high output current lo, the diode D2 conducts.

Accordingly, in the prognosis algorithm F, it is first calculated which diode is currently conducting in order to then select the state-space system to be calculated for phase p1 from the state-space description. The point in time at which the next diode conducts is then calculated for the selected state-space system using the specified switch voltage equations uQi, which are specified in the time domain. If the calculated time is outside the dead time Td for the switches Q1 - Q4, an end time of the dead time Td is selected as the next time for the calculation. Otherwise, the final values for phase p1 are calculated for the shortest time resulting from the switch voltage equations uQi, and the next state space system is then selected.

Usually, the representation in the time domain of the states x or the switch voltages uQi corresponds to that of a decaying oscillation around an offset (decaying offsetted oscillation), as shown below: $$\begin{array}{l}x\left(t\right)=A{e}^{-tB}\left(cOs\left(tC\right)+\text{sin}\left(tC\right)D\right)+E\\ =A{e}^{-tB}\sqrt{D^2+1\text{sin}\left(tC+\text{arctan}\left(\frac{1}{D}\right)\right)+E}\end{array}$$

Here, the coefficients A, B, C, D and E are known system-characteristic constants or depend on the initial values u according to the state-space description. Since the above equation for x(t) = 0 cannot be determined analytically for certain t, a numerical calculation using Newton's method can be chosen to solve it.

A first estimate for a possible approximate solution can look like this after applying a simplified trigonometric addition theorem: $$\begin{array}{l}{\mathrm{<figure-callout\; id="0"\; label="t"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">t</figure-callout>}}_0=\frac{1}{C}\left(\text{arcsin}\left(\frac{-E}{A\sqrt{\left(D^2+1\right)}}\right)-\text{arctan}\left(\frac{1}{D}\right)\right)\\ x\left(t_0\right)=A{e}^{-t_{0}B}\left(\left(-B+CD\right)cOs(t_{0}C\right)\\ +\left(-DB-C\right)\text{sin}\left(t_{0}C)\right)\\ x\left(t_0\right)=A{e}^{-t_{0}B}\left(cOs\left(t_{0}C\right)+sin(t_{0}C\right)+E\\ ={\mathrm{<figure-callout\; id="0"\; label="t"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">t</figure-callout>}}_0-\frac{x\left(t_0\right)}{x\left(t_0\right)}\end{array}$$

The approximate solution can be used because the decay factor B has a comparatively long time constant, but the calculations here relate to dead time effects. Typical dead times are generally below 100 ns.

After an end time for the diode conduction process of the selected diode or an end time for a period in which none of the diodes conducts has been determined, the end values for the first phase p1 result as follows: $$\text{x}={\left(s{\text{I}}_n-A\right)}^{-1}\left(\text{Bu}+{\text{x}}_0\right)$$

In principle, only a final value of the coil current ilk is of interest, since this should be known as precisely as possible in order to determine the switching time tE. However, since the system can change within the dead time Td due to different diode conduction processes, the states of charge or voltages at the capacitances are also determined in order to know them for the subsequent phases.

wobei m₂ ein systemabhängiger Steigungskoeffizient der zweiten Phase p2 ist. Durch eine iterative Berechnung, kann der Endwert des Spulenstroms ilk festgelegt werden. Eine erste Schätzung für den Endwert ilke des Spulenstroms ilk kann durch ilke = ilk (m₂t + 1) bestimmt werden. Anschließend kann die Ableitung des Spulenstroms ilk für einen vorbestimmten Strom von ilk (m₂t + 1) m₂ berechnet werden. Dabei kann angenommen werden, dass die berechnete Ableitung für den Langzeitprozess in Phase p2 konstant ist. Daraus ergibt sich die Mittelwert-Ableitung für den Anfangswert und den Endwert des Spulenstroms ilk in Phase p2 wie folgt: $$ilk\left(t=\left(1-d\right)\frac{Ts}{1}-Td\right)=1+\frac{m_2}{2}\left(2+m_{2}t\right)t$$

Anschließend folgt die Berechnung der dritten Phase p3, die ähnlich zu der Berechnung in Phase p1 abläuft. Da in der Phase p3 die Kanalspannung Vs anliegen kann und die Phase p3 somit der Ursprung für den vorgenannten duty cycle loss dL sein kann, ergeben sich hier wesentliche Einflüsse für eine Dynamik des Systems.Compared to the first phase p1, the second phase p2 is a long-term process. In the inverter circuit W, the two upper switches Q1 and Q2 are closed, ie switched through, while the two lower switches Q3 and Q4 are open. The phase p2 is thus the so-called freewheeling phase (free wheeling mode) of the converter K. The coil current ilk will flow through the switches Q1 and Q2 in phase p2. This relationship is shown analytically in the following equation: $$\frac{\mathrm{i.e}}{\mathrm{i.e}t}ilk=-\frac{1}{Lk}\left(\frac{n^2}{2}R\mathrm{i.e}ss+2R\mathrm{i.e}sp+Rp\right)ilk=m_{2}ilk,$$

where m ₂ is a system dependent slope coefficient of the second phase p2. The end value of the coil current ilk can be determined by an iterative calculation. A first estimate for the final value ilke of the coil current ilk can be determined by ilke=ilk(m ₂ t+1). Then the derivative of the coil current ilk can be calculated for a predetermined current of ilk (m ₂ t + 1) m ₂ . It can be assumed that the calculated derivation for the long-term process in phase p2 is constant. out of it the mean value derivation for the initial value and the final value of the coil current ilk in phase p2 is as follows: $$ilk\left(t=\left(1-\mathrm{i.e}\right)\frac{Ts}{1}-T\mathrm{i.e}\right)=1+\frac{m_2}{2}\left(2+m_{2}t\right)t$$

This is followed by the calculation of the third phase p3, which is similar to the calculation in phase p1. Since the channel voltage Vs can be present in phase p3 and phase p3 can therefore be the origin of the aforementioned duty cycle loss dL, there are significant influences on the dynamics of the system here.

In the third phase p3, the switches Q4 and Q6 are closed, ie switched through, and are therefore represented by their respective switch resistors R4 and R6. The switches Q1, Q2, Q3, and Q5 are open, ie off. In phase p3 their influence can be seen through their respective body diodes D1, D2, D3 and D5 and their respective capacitances C1, C2, C3 and C5. The phase p3 can thus also be referred to as a dead time phase, analogously to the phase p1.

• Auffällig ist, dass eine Systemdynamik in Phase p3 gemäß der Zustandsbeschreibung stark von dem Anfangswert x₀(4) = ilk abhängt. Dieser Zusammenhang betätigt die Relevanz der Bestimmung des Spulenstromverlaufs.

In particular, the following state space description results in phase p3: $$\begin{array}{l}{\text{A}}_{\dot{\text{x}}}=\left[\begin{array}{cccc}-RcpCp& -Rc\mathrm{<figure-callout\; id="0"\; label="p\; C\; p"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">p</figure-callout>}Cp& 0& 0\\ Cp& -\mathrm{<figure-callout\; id="0"\; label="C\; p"\; filenames="DE102021108698A1\_0018.png"\; state="\{\{state\}\}">C</figure-callout>}p& 0& 0\\ 0& 0& \frac{2}{n}Cs& 0\\ 0& RcpCp& -\frac{n}{2}\left(Rcs+R\mathrm{i.e}ss\right)Ca& -Lk\end{array}\right]\\ {\text{A}}_{\text{x}}=\left[\begin{array}{cccc}1& 1& 0& 0\\ 0& 0& 0& 1\\ 0& 0& 0& 1\\ 0& -1& \frac{n}{2}& R\mathrm{i.e}sp+Rp\end{array}\right]\text{Bu}=\left[\begin{array}{ll}-1\hfill & 0\hfill \\ 0\hfill & 0\hfill \\ 0\hfill & \frac{-1}{n}\hfill \\ 0\hfill & \frac{n}{2}R\mathrm{i.e}ss\hfill \end{array}\right]\\ \text{x}={\left[\mathrm{and}2\mathrm{and}3\mathrm{and}4ilk\right]}^2\text{and}={\left[ViniL\right]}^{\text{T}}\\ {\text{x}}_0={\left[Vin00ilk\right]}^{\text{T}}\end{array}$$

• It is noticeable that a system dynamics in phase p3 according to the state description depends strongly on the initial value x ₀ (4) = ilk. This relationship confirms the relevance of determining the course of the coil current.

The following initial values uQi (t0) result for the respective switch voltage uQi in phase p3 by applying the Laplace end value theorem calculation: $$\begin{array}{l}\mathrm{and}Q1\left(t_0\right)=Vin+\frac{1}{2}ilkRcp\\ \mathrm{and}Q3\left(t_0\right)=-\frac{1}{2}ilkRcp)\\ \mathrm{and}Q5\left(t_0\right)=-\frac{RcpCp}{2Cs}\left(IL-nilk\right)\end{array}$$

In this case, as described above for phase p1, the respective diode conducts when the switch voltage uQi is less than the negative forward voltage Vd of the associated diode. The diode conduction processes can then again be determined depending on the selected state space system. In phase p3 it should be noted that a number of diodes can be conductive, in particular at the same time.

After the appropriate state space system has been selected depending on the diode conduction processes and the respective diode conduction time has been determined using Newton's method, the end values x in phase p3 are obtained, analogously to the equation for the end values x according to phase p1.

In phase p3, the switching time tE for turning on the switches Q5 and/or Q6 is brought forward if the diode D5 is not conducting in phase p3, ie, for example, if no diode conduction processes occur or diode D1 is conducting. An additional voltage uadd can then also occur within the dead time Td. This case can occur, for example, when the converter K is in step-up mode or when there is a low load, ie a low output current Io. The secondary voltage, ie the channel voltage Vs, then increases within the dead time Td. This can change the dynamics of the system. In this case, the secondary voltage Vs can be calculated as follows: $$Vs=\frac{\mathrm{and}3+RcpCp\frac{\mathrm{i.e}}{\mathrm{i.e}t}\mathrm{and}3-Lk\frac{\mathrm{i.e}}{\mathrm{i.e}t}ilk-\left(Rp+R\mathrm{i.e}sp\right)ilk}{n}$$

By integrating the voltage Vs over the duration of phase p3 and dividing the result by half the switching time Ts/2, an average value for the secondary voltage Vs can be determined. This average value then results in the additional voltage uadd that can occur in phase p3.

In the fourth phase p4 it depends on whether the diode D5 is still conducting. The fourth phase p4 is also referred to as the power transfer phase (power transition mode) of the converter K. If the diode D5 is still conducting at the end of phase p3 and thus in the fourth phase p4, the switching time tE for switching on the switches Q5 and/or Q6 must be delayed. In this case, the aforementioned duty cycle loss dL then also occurs. The diode conduction time of the switch Q5 thus has a direct impact on the switching delay and the duty cycle loss dL. The duty cycle loss dL can be determined by calculating an average coil current ilka, a current rise islp and the secondary voltage Vs as follows: $$\begin{array}{l}iL=IO-\frac{1}{2}I\mathrm{right}\\ ilka=\frac{1}{2}\left(\frac{1}{n}iL+ilk\right)islp=\frac{1}{n}iL-ilk\\ Vs=\left(\frac{1}{4}\left(iL+nilk\right)-\frac{IL}{2n}+iL\right)R\mathrm{i.e}ss-\frac{V\mathrm{i.e}}{2}\\ \Delta t=\frac{Lk}{Vin-ilka\left(2R\mathrm{i.e}sp+Rp\right)+Vs\frac{n}{2}}isip\\ \mathrm{i.e}L=2\frac{t}{Ts}\end{array}$$

Overall, the course of the coil current ilk can be calculated in advance using a few measured mean values and a mathematical model of the converter K. The coil current ilk can then be used to draw conclusions about the channel voltage Vs and the switching times td can be adjusted.

Reference List

A1

first connection

A2

second connection

co

filter capacity

C1 - C6

intrinsic capacity

cin

input capacitance

CΔt

rectification regulator

i.e

duty cycle

D1 - D6

intrinsic body diode

f

prediction algorithm

G

rectifier circuit

GND1, GND2

reference potential

H

input parameters

lo

output current

i.e

filter stream

ilk

coil current

iL

filter coil current

Irish

ripple current

K

converter

lf

filter inductance

Luke

leakage inductance

P

primary side

p1 - p4

phases

Q1 - Q6

switching elements

rc

internal resistance

S

secondary side

Δt

time difference

Δtset

target difference

te

switching time

T

transformer

Vo

output voltage

vintage

input voltage

W

inverter circuit

QUOTES INCLUDED IN DESCRIPTION

This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent Literature Cited

• WO 2013/075896 A2 [0004]
• DE 102015107960 A1 [0005]
• DE 112014006828 T5 [0006]

Claims (10)

Method for operating a converter module (K), which provides a coil arrangement (T) for intermediate energy storage and a switch arrangement (G) for controlled rectification, the switch arrangement (G) being operated according to a predetermined switching pattern, and for the predetermined switching pattern according to a predetermined prognosis algorithm (F) a switching time (tE) for switching through the switch arrangement (G) is adjusted, wherein According to the prognosis algorithm (F), for a respective half-wave of a coil voltage applied to the coil arrangement (T), a time profile of a channel voltage (Vs) that can be provided at the switch arrangement (G) is predicted by a time profile of a coil current (ilk), which is provided by the coil arrangement (T) as a function of the coil voltage applied, is calculated and the switching time (tE) is determined as a function of the coil current curve (ilk) and the switch arrangement (G) is actuated to switch through. procedure after claim 1 , wherein the input parameter (H) for the prognosis algorithm (F) is a mean value of an input voltage (Vin), an output voltage (Vo) and an output current (lo) of the converter module (K), and a duty cycle (d) of the switch arrangement predetermined by the switching pattern (G) are determined, and the following steps are carried out according to the prognosis algorithm (F): - determining an initial value of the coil current (ilk) as a function of the input parameters (H) and setting the initial value as a target value for the coil current (ilk), - Carrying out a calculation routine in which, starting from the initial value, the time profile of the coil current (ilk) is determined, with the half-wave being divided into at least two phases (p1, p2, p3, p4) in the calculation routine and for each of the phases (p1, p2 , p3, p4) a respective energy storage state of the converter module (K) is calculated and a value profile of the coil current assigned to the respective energy storage state ms (ilk) is determined, and - determining the switching time (tE) for the switch arrangement (G) as that time at which the coil current (ilk) assumes a value over time that corresponds to the specified setpoint value. procedure after claim 2 , wherein to determine the initial value of the coil current (ilk) from the input values for the half-wave, a ripple current (Ir), which is set as a switch arrangement output current at the switch arrangement (G), is determined, the ripple current (Ir) having an oscillation behavior of the Switch arrangement output current to the mean value of the output current (lo) of the converter module (K) indicates. Procedure according to one of claims 2 or 3 , wherein according to the calculation routine the phases (p1, p2, p3, p4) depending on a transfer behavior of the coil arrangement (T) for a respective value range of the coil voltage within the half-wave are determined. Procedure according to one of claims 2 until 4 , wherein the coil voltage applied to the coil arrangement (T) is provided by means of an inverter circuit (W) of the converter module (K), and the phases (p1, p2, p3, p4) for dividing the half-wave depending on a switching configuration of the Inverter circuit (W) can be selected. procedure after claim 5 , A predetermined switching delay time (Td) of the inverter circuit (W) being taken into account according to the prognosis algorithm (F) when determining the respective switching time (tE). Method according to one of the preceding claims, wherein a rectifier circuit (G) with at least two semiconductor switches (Q5, Q6) is provided as the switch arrangement (G), and the at least two semiconductor switches (Q5, Q6) are operated according to the switching pattern, the switching pattern for the semiconductor switches (Q5, Q6) are applied to the semiconductor switches (Q5, Q6) according to a predetermined switching logic with a predetermined switching offset to one another, the switching offset representing a pulse duty factor (d), and according to the prognosis algorithm (F) the switching time (tE) is adjusted for each of the semiconductor switches (Q5, Q6). Method according to one of the preceding claims, wherein to adapt the switching time (tE) in a controller routine, the switching time (tE) determined by means of the prognosis algorithm (F) is compared with a switching time provided according to the switching pattern and a difference (Δt) between the switching times is zero is regulated. Converter device with a converter module (K) according to one of the preceding claims and a control device which is designed to operate the converter module (K) by carrying out a method according to one of the preceding claims. Motor vehicle with a converter device claim 9 , wherein the converter device is designed for converting electrical energy between a high-voltage electrical system and a low-voltage electrical system of the motor vehicle.

Images