DE102016217493A1 - Method for changing the direction of a DC voltage for generating pulse width modulated signals in a vehicle electrical system - Google Patents

Abstract

A method is described for the reversal of a DC voltage for generating at least two pulse-width-modulated and respectively multi-phase voltage signals in a vehicle electrical system. It is a nominal modulation (m) specified. An overall carrier signal shift (V) is determined for the desired degree of modulation on the basis of a predetermined mapping which allocates associated current loads (B) for different modulation levels to different total carrier signal shifts, in which the current load of a DC link is minimal; A first and a second pulse pattern are respectively generated with the desired duty cycle (m). The pulse patterns are offset from one another by the total carrier signal shift (V). The two reference voltage vectors, from which the two pulse patterns are calculated, are angle-corrected by the geometric offset of the two winding systems and the total carrier signal shift. A first and a second pulse-width-modulated voltage signal (SIG1, SIG2) in accordance with the first and the second pulse pattern are output.

Description

Electric drive machines in automotive engineering are usually driven by voltage-fed pulse inverters. As a result, three-phase machines can be used, wherein the DC voltage, that is, the supply voltage of the vehicle electrical system, is converted by pulse inverters in a (three-phase) AC signal. This makes it possible to set a desired effective voltage level for each phase or to define a power signal (such as a sinusoidal signal), wherein the pulse width modulated signal between two levels (positive and negative pole of the energy storage of the on-board network) can switch. The duty cycle indicates the ratio of the on state of the corresponding switch of the pulse inverter in relation to the period. In order to set a substantially sinusoidal phase current in three-phase machines, the clock frequency should be much greater than the electrical frequency (i.e., the fundamental frequency of the three-phase current) of the electric motor. The reason for this is that the pulse change signal is a rectangular signal which approximately simulates the sine signal of a three-phase current signal.

A component of pulse inverters or in general of inverters is an intermediate circuit which provides energy for controlling the drive machine according to the switching frequency of the inverter. The intermediate circuit comprises in particular one or more energy stores in the form of a capacitor (and / or a coil), wherein substantially the entire alternating component of the pulse width modulated signal is obtained from the intermediate circuit. The control results in a relatively large load on the DC link, depending on the degree of modulation. For better understanding, the intermediate circuit or its energy store (in particular the capacitor or the capacitors) is referred to as a "DC link capacitor".

In the following, a possibility is to be shown with which the load of the DC link capacitor can be reduced.

This object is achieved by the method according to claim 1. Further possibilities, alternatives, advantages and characteristics emerge from the subclaims as well as from the description and the figures.

Disclosure of the invention

It is proposed to use in a vehicle electrical system two mutually time-staggered pulse pattern for driving. From the pulse patterns voltage signals are generated, which can be used to control a load. It has been recognized that the loading of a common link capacitor, which supports the generation of the voltage signals, can be reduced by an appropriate overall shift between the pulse patterns. The term "displacement" should be considered as a time lag.

The load can be reduced in this way compared to a pulse width modulation with the same power, which uses only one (1) pulse pattern. As a result, the intermediate circuit capacitor can be designed for lower loads. On the one hand, this can save costs and, on the other hand, this can reduce the installation space required for the DC link capacitor. In addition, a higher reliability may result, since the DC link capacitor is less loaded.

A method is proposed for the reversal of a DC voltage for generating at least two pulse width modulated voltage signals in a vehicle electrical system. The pulse width modulated voltage signals are each multi-phase, in particular three-phase, whereby a higher number of phases can be used. The pulse width modulated voltage signals each have a voltage that changes between two values according to the pulse pattern. One of these values is zero, the other value may, for example, correspond to the DC voltage which is subjected to the alternating direction, for example the positive potential of the input voltage. In particular, an (approximate) sinusoidal signal is generated by the pulse width modulation, which can be supplied, for example, to an electrical machine. The voltage signals are, in particular, three-phase signals whose fundamental frequency component (based on the signal power) amounts to at least 30%, 50% or 70% and preferably at least 80% or 90% of the total signal power and particularly preferably at least 98% or 99%. The phase offset between the individual phases of each of the voltage signals is preferably the same, for example 120 ° in a three-phase system.

In order to regulate the power of the voltage signals, a desired modulation degree m is specified. This can in particular be predetermined by a control unit, wherein the voltage signals are converted in a downstream inverter circuit, in particular in an inverter circuit. With the pulse patterns are (several, in particular at least two) driven bridges of an inverter circuit. Each pulse pattern drives its own bridge (an inverter circuit).

It is determined for this nominal modulation degree m a (temporal) total carrier signal shift. Here, the term " Carrier signal shift ", also referred to as CSS (" Carrier Signal Shift "). The total displacement is determined in such a way that a minimum current load for the intermediate circuit capacitor results for the desired modulation degree m. The total carrier signal shift (or the total CSS) is the time offset between the first and the second voltage signal or between the first and the second pulse pattern. Within a pointer model of the complex AC calculation, the first and the second voltage signal together correspond to a setpoint pointer, which represents a (sinusoidal) three-phase current signal. The sum of the voltage signals thus results in a multi-phase, approximated sine signal; the desired pointer results from the sum of the two hands that represent the voltage signals. A shift is referred to here in particular as a delay. As mentioned, the shift refers to the dimension time.

Each voltage signal (or each winding system, which is supplied with the relevant voltage signal) is assigned its own desired voltage vector. The pulse pattern is calculated separately from the voltage phasors for each winding system.

The first and second pulse patterns are generated from angle corrected voltage vectors. The angle corrected voltage phasors are corrected for geometric offset (i.e., the geometric angle between winding systems). Furthermore, the angle corrected voltage phasors are corrected for carrier signal shift.

To form the overall CSS based on the desired modulation depth, a given mapping can be used. This mapping assigns different total CSS to associated power loads. The figure assigns the various total CSS in particular to associated current loads for different degrees of modulation. The image can thus be three-dimensional. If the degree of modulation is predetermined and corresponds to the desired degree of modulation m, then the mapping (based on the desired degree of modulation) maps different total CSS to corresponding current loads. It can be the minimum current load (for the desired modulation depth m, which is given) are determined, as well as the associated overall CSS. The voltage signals are generated according to the determined total CSS (for which the current load is minimal and for which the desired modulation degree m holds). The generated pulse patterns are then output, in particular to inverter circuits. The mapping can in principle assign value pairs, which respectively represent a desired duty cycle and a total shift, to a current load.

If the mapping is parametrized in advance according to the desired modulation degree m, which is predetermined, then the mapping can be reduced to a map or function associating different overall CSS with current loads, provided that the degree of modulation matches the given one Nominal modulation degree m corresponds.

One possibility of implementation is to start from a given basic degree of modulation and for this purpose determine the minimum of the current load (and the associated carrier signal shift or CSS). The basic modulation level can be predefined from the outset, for example when implementing the method in a control unit. On the basis of this, a difference between the basic modulation degree and the (current, possibly variable) desired modulation degree can be formed. A function allocates different differences between the basic modulation level and the desired modulation level to associated variable carrier signal shifts (i.e., to associated variable CSS values). The variable carrier signal shift resulting in the desired modulation level is added to a constant carrier signal shift (i.e., to a constant CSS value). The constant carrier signal shift is the one in which the current load is minimal from the basic modulation level. The constant carrier signal shift and the variable carrier signal shift add up to the total carrier signal shift. The voltage signals are pulse width modulated according to this total carrier signal shift. The possibilities for calculating the total carrier signal shift will be explained in more detail below.

There are generated (at least) a first and a second pulse pattern as two (or more than two) same pulse pattern, each having the desired modulation depth. The pulse patterns are generated such that they are offset from each other by the total Tastsignalverschiebung. This can be done, for example, by two (or more than two) coupled signal generators, or by a signal generator whose output is offset according to the total carrier signal shift and which controls the pulse width modulation directly on the one hand and on the other, for example. One bridge of the inverter circuit can be controlled directly and another bridge can be triggered in a delayed manner according to the total carrier signal shift. Alternatively, one bridge is driven by one signal generator and another bridge is driven by another signal generator driven. A microprocessor can be provided which has a plurality of (physical or logical) outputs or output channels, wherein a first pulse pattern and (at least) a second pulse pattern (in accordance with the total carrier signal shift) are output at the plurality of outputs. A running on the microprocessor computer program controls the outputs as mentioned.

The pulse patterns are used in particular for controlling inverter circuits. These convert the pulse patterns and generate voltage signals. These voltage signals are adapted to be applied to a load, in particular to an electrical machine. The voltage signals have levels that are adapted to the rated voltage of the electrical machine. The voltage signals are delivered as power signals, i. by impressing currents whose magnitude is adapted to a rated current of the load.

Based on the pulse pattern, a first pulse-width-modulated voltage signal (multiphase) is delivered or generated, and a second, pulse-width-modulated voltage signal (multiphase) is output or generated in accordance with the second pulse pattern. An inverter circuit having a plurality of bridges converts the pulse patterns (as drive signals) into the voltage signals (as power output signals).

In the implementation of the pulse patterns, i. used in the delivery of the voltage signals a plurality of bridges (or bridge circuits), which belong in particular to an inverter. The bridges each have a DC terminal (i.e., a positive and a negative terminal). The DC voltage connections of the bridges are connected in parallel. Parallel to this, the DC link capacitor is connected. This supplies the bridges with alternating current. The bridges each have arms in a number corresponding to the number of phases of a bridge. Each arm has switching elements, in particular two serial (controllable) switching elements (a switching element as a low-side switch and a switching element as a high-side switch). In a so-called continuous pulse width modulation, all switches are switched within each bridge. In so-called non-continuous pulse width modulation, all switches within each bridge are switched except for the switches of one arm.

Therefore, according to a pulse modulation method, switching elements of all the arms of a plurality of controllable bridges can be switched (within one phase of the relevant arm). Alternatively, (within one phase of the respective arm) the switching elements of all the arms of a plurality of controllable bridges-except for switching elements of one arm of the respective bridge-can be switched in accordance with a pulse modulation method. The first and the second pulse pattern can be generated as a continuous or non-continuous pulse width modulated DC voltage.

According to one embodiment, the first voltage signal is applied to a first winding system and the second voltage signal is applied to a second winding system of the same electric machine, or to an electrodynamic system with the first and the second winding system, these two transmitting torque (directly or indirectly ) are interconnected. The two winding systems can be located in two separate electrical machines, which can be mechanically coupled. The two winding systems can be directly connected to each other, for example via a shaft, or can be connected to each other indirectly or switchably, for example via a transmission or via a coupling. A relevant aspect is that each winding system can by definition independently convert electrical to mechanical energy.

The winding systems are each multi-phase. The winding systems can be geometrically offset from each other. As the "same electric machine" is also an arrangement of a plurality of stators or rotors to understand that have the winding systems, and act on the same shaft or are connected by this. As part of the output of the voltage signals individual windings of a first and a second winding system of the same electric machine (or the same electrodynamic system) are supplied with the voltage signal.

The voltage signals can be delivered to single windings. Thus, the method includes providing single windings. In particular, first and second winding systems are supplied which comprise the individual windings. The individual windings correspond to the individual phases of the winding systems. The individual windings of each winding system are evenly distributed over the same angle. The individual windings are each supplied via a separate, uniquely assigned arm of the bridge.

The calculation of the total carrier signal shift between the voltage signals or between the pulse patterns can require a high level of computation, in particular if the nominal modulation level changes rapidly. Therefore, part of the determination of the total carrier signal shift may already be involved in the implementation or development of an implementation of the method (for example, of a controller or of a controller) Program codes within the controller).

It can be provided that a minimum for a current load of a DC link capacitor within a predetermined function is determined. For a given basic modulation level, for example 1, this predetermined function assigns different current loadings to different fundamental carrier signal shifts. The basic modulation level can also have any further values, for example also 0, 0.5, 0.75 or 0.25. The carrier signal shift which corresponds to the minimum (the current load) is determined on the basis of the predetermined function. It is assumed that the degree of modulation corresponds to the basic degree of modulation.

In particular, the given function assigns associated current loads to CSS values, provided that the degree of modulation corresponds to the basic modulation depth. It is determined that carrier signal shift, which leads to the minimum current load of the DC link capacitor. This is done on the basis of the given function, wherein this carrier signal shift is determined as a constant displacement K. This is part of the total carrier signal shift of the overall carrier signal shift. In particular, the total carrier signal shift is the sum of the constant carrier signal shift and a variable carrier signal shift. The determination of the minimum and the associated carrier signal shift described in this paragraph on the basis of the associated function can in particular also be carried out before commissioning, during assembly, during delivery or during the programming of a control device or an inverter. Instead of determining the minimum of the current load and the determination of the carrier signal shift and the constant carrier signal shift can be specified by the above-mentioned figure, which is applied for a modulation degree corresponding to the basic modulation level, and also the function described here can be used or others , empirical results that apply to the given basic modulation level.

The determination of the constant carrier signal shift can be considered as a first step, wherein in a further step in a differential manner the current (i.e., nominal) modulation degree is taken into account in the calculation of the total carrier signal shift. In particular, in a further step, a variable carrier signal shift can be determined, which adds to the constant carrier signal shift, so as to form the total carrier signal shift. This second step is preferably carried out during operation, since during operation the desired modulation degree can change or because the desired modulation degree is specified during operation.

m:

Modulationsgrad

sqrt(3):

Faktor für dreiphasige Systeme

Û_ph:

Spitzenspannung einer Phase des Spannungssignals

U_DC:

Zwischenkreisspannung bzw. Spannung an der Gleichspannungsseite des Wechselrichters

The degree of modulation indicates the ratio of the magnitude of the voltage signal to an intermediate circuit voltage and to a voltage on a DC side of the inverter, respectively. In particular, the degree of modulation (m) indicates the ratio of the magnitude of an output voltage pointer (of the voltage signal or of the winding system) to the intermediate circuit voltage. The degree of modulation can be defined as follows: m = √ 3 · Û _ph / U _DC With

m:

modulation rate

sqrt (3):

Factor for three-phase systems

Û_ph:

Peak voltage of one phase of the voltage signal

U_DC:

DC link voltage or voltage on the DC side of the inverter

The method can thus further provide for determining a variable carrier signal shift. This is part of the overall carrier signal shift. Together with the constant carrier signal shift K, which is also part of the overall carrier signal shift, the sum of these carrier signal shift values forms the overall carrier signal shift. In this case, a difference between the basic degree of modulation and the desired degree of modulation is assumed. The difference reflects how much the actual modulation level deviates from the basic modulation level set in advance. According to this deviation, expressed by the difference, the variable carrier signal shift is determined, which can be added to the constant carrier signal shift. Here, a further function is assumed, which assigns a respective carrier signal shift in a differential manner to different degrees of modulation. This further function assigns different variable carrier signal shifts to different differences between the basic modulation degree and the desired modulation degree m. The function thus maps the difference to the variable carrier signal shifts. In particular, this further (or differential) function maps the differences (that is, the deviations to the basic modulation degree) to variable carrier signal shifts or carrier signal shift values for which the current load is minimal (added to the constant carrier signal shift). This additional function is used during operation, ie with current target modulation level, as opposed to the first mentioned function, which is executed or used in the beginning (eg during the implementation).

The further function preferably does not generally map the degree of modulation or differences thereof to carrier signal shift values, but depicts them on the assumption that the current load is minimal. In particular, the further function may be an approximation function, for example a linear function or a function which assigns modulation grades (or differences between the basic and desired modulation depth) at least in sections according to a linear function. In particular, the function can assign these modulation degrees or differences to constant carrier signal shifts (CSS) or to variable carrier signal shifts in a first interval of modulation degrees or of differences in modulation depths and in a second interval of modulation degrees or modulation degree differences according to a linear mapping Assign differences to different carrier signal shifts for which the current load is minimal. An interval like the first interval can occur several times, in particular before and after the interval in which there is a linear function.

The function may be defined by interval boundaries between which a constant displacement value exists, which is also part of the definition, and further interval boundaries between which there is a linear dependence, the linear dependence being defined by a displacement constant (corresponding to a constant carrier signal shift) and a slope that reflects the change in the shift depending on the degree of modulation (with minimized current load).

The first and the second winding system (and if present: also one or more other winding systems) can be present as a star or delta connection and thus supplied with the voltage signal. In particular, the winding systems are supplied as separate systems in star or delta connection. The winding systems supplied with the voltage signals preferably all have the same configuration (i.e., in star or delta connection).

The first and the second winding system (and, if present, also one or more further winding systems) may each have a star point. The neutral points of the winding systems are in particular not connected to each other.

The winding systems can be supplied independently of each other. That is, the first and second winding systems are supplied as separate systems.

The winding systems or their individual windings can, as mentioned, be supplied via a first bridge and a second bridge (or via at least one further bridge). In particular, the first winding system is supplied by switching a first bridge having controllable switching elements. The second winding system is powered by switching a second bridge, which also has controllable switching elements. The first and second bridges are preferably powered by the same voltage source. To this voltage source of the DC link capacitor is connected in parallel. In other words, the first and the second bridge (and possibly further bridges) each have a DC voltage input, wherein the voltage inputs of the two bridges are connected in parallel and further an intermediate circuit capacitor is connected in parallel to these voltage inputs. More generally, the first voltage signal may be delivered by switching the first bridge, and the second voltage signal may be delivered by switching the second bridge. Here, the voltage signals can be delivered to a load, or in particular (as mentioned above) to winding systems, in particular of electrical machines, such as three-phase machines.

The first and the second bridge can each be designed as a B6C bridge. A B6C bridge incorporates a DC input and, for each phase, an arm containing two series-connected switching elements. The connection point between the switching elements of each arm forms a phase connection. Since one arm is provided for each phase, there is one connection point per arm and thus a phase connection. Preferably, a B6C bridge having six switching elements is used as the bridge. These are distributed in three arms, each arm having two series-connected switching elements (a high-side switch and a low-side switch), and the connection point between the switching elements represents a phase connection.

The term "bridge" is an abbreviation for bridge circuit, in particular for a circuit in which switching elements are connected in bridge circuit. As a bridge circuit is in particular a full bridge circuit into consideration. The full bridge circuit is designed in particular as a B6C bridge. The full-bridge circuit has a shunt branch per phase, which has two series-connected semiconductor switches. Phases of a load or the windings (of the winding systems) are each connected to a connection point, over which said two semiconductor switches are connected to each other in series. The two semiconductor switches can be referred to as high-side switches and as low-side switches. With a phase number of three, three shunt branches result, each having two switches. This is called the B6C bridge (the number 6 indicates the number of switches). With a phase number of n, n shunt branches each having two switches result, so that a total of 2 × n switches are used. H-bridge circuits are also contemplated to form the full bridge circuit. Low-side switches are (externally controllable) switching elements that are directly connected to a negative or ground supply potential. High-side switches are (externally controllable) switching elements, which are directly connected to a positive supply potential. The switching elements are in particular controllable from the outside, preferably in the form of electronic switching elements. The switching elements may be semiconductor switching elements, in particular transistors such as bipolar transistors or field effect transistors, for example IGBTs or MOSFETs.

In particular, the method provides for supplying an electrical machine, for example by the voltage signals (that is, at least the first and the second voltage signal) being delivered to the electric machine. As mentioned, the electric machine may be a three-phase machine, in particular a synchronous or asynchronous machine. The electric machine has at least two winding systems, which may be offset in their angular position to each other, or are not offset from each other (geometrically). Preferably, one phase of a winding system is located midway between two phases of the second winding system. In other words, the phases of the winding systems are distributed equally in the circumferential direction. The winding systems act on the same shaft. The winding systems can be arranged on the same laminated core. Furthermore, the winding systems can be arranged on different laminated cores, as long as the winding systems act on the same shaft. The winding systems can act on different locations of the same shaft. An electrical machine is therefore also understood to mean a device which has a plurality of stators (for the different winding systems and / or a plurality of rotors, but which act on the same shaft.) The winding systems can therefore have different points of action, which are in particular offset axially relative to one another, if these act on the same shaft or on the same rotor.

The electrical machine is powered by delivering the first voltage signal to the first winding system and the second voltage signal to the second winding system. It can also be provided more than two winding systems, preferably all winding systems to each other (in the direction of rotation) are offset and the phases of all winding system are equally distributed (seen in the circumferential direction). It can also be provided that the winding systems are not offset from one another. In this case, each winding system receives an associated voltage signal, with the voltage signals different carrier systems being carrier-to-each other shifted (i.e., there is a time shift between the individual voltage signals), in particular such that a minimum of the load of the intermediate capacitor results.

The required or provided apparent power of the electric machine can be divided into active and reactive power. The DC link capacitor provides the reactive power, which leads to a current load. The current load leads to a thermal load via the ohmic resistance in the DC link or in the component (for example in the DC link capacitor).

The carrier signal shift (CSS) may be positive in the sense of delaying or may be negative in the sense of premature. Furthermore, the first pulse pattern can be shifted relative to the second pulse pattern, the second pulse pattern can be shifted relative to the first pulse pattern, or both pulse patterns can be shifted relative to a temporal clock reference point. In the latter case, the total carrier signal shift results from the sum of the (partial) carrier signal shift of the first and the second pulse pattern.

The figure is obtained by calculating the (normalized) alternating current component of both voltage signals or pulse patterns at different degrees of modulation and different carrier signal shifts. To compute the map for a given degree of modulation (or to calculate the function), calculate the AC component for different carrier signal offsets. To obtain the more general mapping, this calculation is performed for different degrees of modulation. The resulting family of curves gives the mapping for different carrier signal shifts V and different modulation degrees m. Alternatively, the data can be determined empirically using a (virtual) model or measured using a (real) model.

The 1 and 2a - 3b serve to explain the method described here, while the 4a and b represent a circuit suitable for implementing the method.

In the 1 a first multi-phase voltage signal SIG1 and a second multi-phase voltage signal SIG2 are shown, which are mutually temporally shifted by the carrier signal shift V. The level is shown on the y-axis and the x-axis corresponds to the time axis t. The levels (ie the peak-to-peak voltage or the effective voltage) of the voltage signals are the same, the shift from SIG2 to SIG1 is for better illustration only. Each of the voltage signals SIG1 and SIG2 comprises several phases. The phases of the first voltage signal SIG1 are the phases U1, V1 and W1, while the phases of the second voltage signal SIG2 are the phases U2, V2 and W2. These are in each case three-phase voltage signals, for example, to control two three-phase winding systems of an electrical machine.

The purely exemplary constant displacement K (which is only a part of the total carrier signal shift between SIG1 and SIG2) in FIG 1 between two reference times of SIG1 and SIG2 (represented by vertical bars) is 36% of the total period T of the voltage signals. This corresponds to the constant (temporal) carrier signal shift K. Further, two variable carrier signal shifts D1 and D2 are shown. This serves to explain that the variable carrier signal shift may be in one piece, or may consist of two parts, as shown in the part D1 and D2. The carrier signal shifts D1 and D2 add up.

The illustrated variable carrier signal shift D1 is + 5% of the total period T. The illustrated variable carrier signal shift D2 is -6% of the total period T. The variable carrier signal shift D1 refers to a temporal (variable) shift of the first signal SIG1 relative to a reference instant t = 0, which can be considered the beginning of the period shown. The carrier signal shift D1 is positive and thus corresponds to a delay. The variable carrier signal shift D2 relates to a temporal (variable) shift of the second signal SIG2. The carrier signal shift D2 is negative and thus corresponds to a time preference or a premature.

The total carrier signal shift (CSS) V results from the sum of the constant displacement K and the variable (partial) displacements D1 and D2, taking into account the relevant signs: V = K - D1 + D2 and results in the illustrated example to 36% - 5% + (-6%) = 25%. The two variable (sub) carrier signal shifts (CSS) can be set independently.

At a constant switching frequency of the two winding systems, the maximum total carrier signal shift can be achieved if the partial carrier signal shift (of the individual voltage signals) are maximized, that is, if, for example, both pulse patterns (in the time axis) are pushed outward to the edge. However, this is a boundary consideration: Most of the maximum possible total CSS is not needed and for reasons such as the sampling time of the phase current, it may be useful to weight the sub-carrier signal shift differently. Thus, a weighting of the sub-carrier signal shifts (i.e., the carrier signal shifts of the voltage signals) may be made. The weighting may be configured to yield a desired sampling instant, such as a sampling instant associated with a minimum harmonic content of the phase current, or associated with a minimum load on the intermediate circuit.

The total carrier signal shift V corresponds to the time interval between the center of gravity of the pulses of the signal SIG1 and the center of gravity of the pulses of the signal SIG2. The center of gravity of the pulse pattern is its axis of symmetry (see dot-dash line). These times are shown with the thicker of the two dash-dotted lines. The centroids of the signals, as would occur with a total carrier displacement of V = K (i.e., without the variable displacement D or D1 and D2, respectively) are shown with the thinner of the two dash-dotted lines. The thicker dot-dashed lines thus show the total carrier signal shift (incl. Variable displacement), the thinner dot-dashed lines show the (basic) carrier signal displacements without variable displacement and the difference of the two shows the variable displacements D1, D2.

The constant carrier signal shift K preferably corresponds to the minimum of the current load at a modulation degree m = 1 1 However, it does not assume a desired degree of modulation m = 1 but, for example, m = 0.5. This difference between the basic modulation degree m = 1, at which the minimum of the current load at a carrier signal shift (CSS) of 36% (= K) are present, and the actual desired modulation depth of, for example m = 0.5 is reflected in the variable carrier signal shift, ie in the sum of D1 and D2. The variable displacement D1, D2 again indicates whether and how strongly the total carrier signal shift, which is associated with the minimum of the current load at nominal duty cycle (here m = 0.5), differs from the total carrier signal shift that corresponds to the minimum of Current load at basic duty cycle (here m = 1) is assigned. The variable carrier signal shift can be considered as a correction of the constant carrier signal shift, which results from the fact that the desired modulation degree differs from the basic modulation degree, that the position of the minimum shifts with deviation from the basic modulation degree. The 2a - 3b serve for a more detailed explanation in particular of this fact.

It should be noted that the reference V refers exclusively to the total carrier signal shift (CSS), while the designations U1, U2, V1, V2, W1, W2 are designations of different phase signals. All phase designations comprise a digit in the last position, while the reference symbol of the total carrier signal shift (reference symbol V) does not comprise a digit.

The signals SIG1 and SIG2 correspond to a continuous pulse width modulation, in which all phases U1-W1 and U2-W2 have a level change within the period. A non-continuous pulse width modulation would have a phase in a winding system which has no level change within the period, ie a phase is not switched within the period. The non-switching of a phase alternates evenly between the individual phases of a winding system within one electrical period. Both - 3b Both winding systems were each driven using the same PWM method. It would also be possible to use different PWM methods for the different winding systems. The pulse pattern for the first voltage signal may thus differ from the pulse pattern provided for the second voltage signal.

The 2a FIG. 12 illustrates the mapping between a total carrier signal shift V, a modulation degree m, and the resulting current loadings B for the DC link capacitor with constant phase shift.

The 2a refers to the current load of a link capacitor, which emits several pulse width modulated signals to loads. It can be seen that the loading depends both on the total carrier signal shift V and on the degree of modulation m. The degree of modulation m corresponds to the magnitude of the phase voltage U_ph in relation to the intermediate circuit voltage U_DC according to the formula m = sqrt (3) · Û_ph / U_DC. The figure results from the calculation of the current load or from the calculation of the alternating current component as a function of the degree of modulation m and the total CSS V with continuous pulse width modulation.

The 2 B shows an approximation of the course of the minimum of the current load (as in 2a shown) depending on the degree of modulation m. The current load B of the 2a is in ampere and refers to a phase current of 80 A. The phase shift between phase current and phase voltage in the 2a and b is 40 °. The two winding systems were not geometrically shifted. In a three-phase system with identical power as in the case shown, ie a phase current of 160 A (= 2 x 80A in double-star configuration in which each winding system has its own star point) would be the DC link load equivalent to the load as in 2a at a total carrier signal shift V = 0%.

The 2 B is an approximation of the function of the total carrier signal shift V, for which the current load is minimal, depending on different degrees of modulation. The approximation is based on consecutive constant and linear sections. The approximation in 2 B is a (partial) linear approximation, there are basically other approximations or functions possible.

m_Soll:

Soll-Modulationsgrad

m_Grund:

Grund-Modulationsgrad.

Referring to the 1 can be initially assumed by a basic modulation level of 1. This value is as in 2 B represented outside achievable values (due to dead times and minimum pulse durations in the real inverter), so that the variable carrier signal shift D is not zero. Nevertheless, a basic modulation degree m of 1 may be the basis of a calculation. It is determined how much the basic modulation degree m differs from a desired modulation degree, for example, of a desired modulation degree of m = 0.5. This is expressed by a difference of degrees of modulation (basic modulation degree m = 1 less nominal modulation degree of m = 0.5). In order to take into account the dependence of the carrier signal shift V with minimum current load on the degree of modulation, the carrier signal shift constant at 36%, which adds a (negative) variable carrier signal shift D at the basic modulation degree m = 1. In 1 this is in two parts, but the sum is 11%. This corresponds to the difference of the displacement (with minimum current load) at basic modulation degree (m = 1 -> V = 36%) and the shift at target modulation degree (m = 0.5 -> V = 25%), ie the difference between 36 % and 25%. It may be provided a function f, which assigns the difference (Diff) between the basic and desired modulation degree of a variable displacement D. This results in the consideration of the change from V starting from the basic modulation degree with increasing distance of the desired modulation degree from the basic modulation degree. In other words, the variable displacement D depends on the desired degree of modulation to: D (m_Soll) = V (m_base) - V (m_soll).

M_set:

Target modulation degree

m_Grund:

Basic modulation depth.

The total carrier signal shift V is: V = K + f (diff) with: constant displacement K = V (m_base) and f: Diff -> D

In the 2 B f is a decreasing function and in particular concerns values of D less than zero.

The simplified mapping of modulation degrees m to V (with minimum current load) can be represented by a first interval m = 0 to m1, in which V = Vmin is constant, a second interval m = m1 to m2, in which V (in particular linear) increases with increasing m, and a third interval m greater than or equal to m2 where V = Vmax is constant. In 2 B m1 = 0.66, m2 = 0.92, Vmin = 25% and Vmax = 36%. Referring to 1 is initially assumed by m = 1 as a basic modulation level, ie of a carrier signal shift V (m = 1) = 36%. This corresponds to (in 1 ) the value K, ie the constant displacement. However, for example, since the target modulation level is 0.5, a (negative) variable carrier signal shift of -11% is added to the constant carrier signal shift (CSS) of 36% to account for the difference between the base and target modulation levels. This results in a shift of 25% at a desired modulation depth of 0.5. The variable shift of 11% results from the function that assigns differences between the fundamental and desired modulation levels to variable carrier signal shift D. In the 2 B If D is constant, then for differences Diff = 0 ... diff1, it decreases linearly with increasing difference for differences Diff = diff1 to diff2 and is constant equal to Db for differences Diff greater than or equal to diff2. The variable shift Da is in 2 B 0%, the variable shift Db is -11%. In 2 B Furthermore, Diff1 is equal to (1-0.92) and Diff2 is (1-0.66). The 1 stands for the basic modulation degree.

The 3a represents the same context as the 2a with the difference that the 3a refers to a non-continuous pulse width modulation. Otherwise, the basic conditions are the same. It can be seen that the minima of the total carrier signal shifts V otherwise depend on m than in 2a is shown.

The 3b shows an approximation of the curve of the total carrier signal shift V (with minimum current load) depending on the degree of modulation m. In the 2 B a rising function is shown, as opposed to falling function of the 3b , The simplified mapping from modulation level m to V (with minimum current load) 3b can be represented by a first interval m = 0 to m1 in which V = Vmax is constant, a second interval m = m1 to m2, in which V (in particular linear) decreases with increasing m, and a third interval m greater than m2 in which V = Vmin is constant. In 3b m1 = 0.66, m2 = 0.92, Vmax = 50% and Vmin = 36%. Here as well, in the calculation of V, starting from a desired degree of modulation, one can start from the V which corresponds to the basic modulation degree 1, ie from the constant carrier signal shift K = 36%. If the nominal degree of modulation actually lies at 0.8, then a function f can be used, which is the difference (here: 0.2) between the basic modulation degree (here: 1) and the desired modulation degree (here: 0.8) a variable carrier signal shift D (here: + 6%). The total carrier signal shift V, for which the current load is minimal, is then V = K + D, ie 42% = 36% + 6%. The function f can be constant up to a first difference diff1 (in particular zero), increase linearly between the first and a second, larger difference diff2, and be constant from the difference diff2, in particular with a value greater than the constant that is at differences less than diff1.

In the 4a is an inverter circuit 10 with a first bridge 12 and a second bridge 14 shown. The bridges are each B6C bridges which comprise controllable switching elements S _HU -S _LZ . The designation of the switching elements shows the affiliation with one of the phases U-Z with the last digit, and the affiliation with a part of the branch with the second digit, where H stands for high-side and L for low-side. The S in the first place stands for "switching element". It can be seen that the DC inputs of the two bridges 12 and 14 are connected in parallel with each other, wherein a capacitor C is also connected in parallel thereto.

The inverter 10 is an electrical machine with a first winding system U, V, W and a second winding system X, Y, Z downstream. At this flows the currents i _PU to i _PZ . The electric machine 20 thus comprises six phases U-Z, which are grouped into two winding systems, namely a first winding system U, V, W and a second winding system X, Y, Z. The first bridge circuit 12 supplies the first winding system U, V, W and the second bridge 14 supplies the second winding system X, Y, Z. It can be seen that the winding systems of the electric machine 20 each have a star point ST1, ST2. Star points ST1, ST2 of different winding systems UVW, XYZ of the electrical machine 20 are not connected. In the machine shown, the two winding systems are geometrically offset by 180 °.

The bridge circuits 12 and 14 are each designed as a B6C bridge. The inverter and thus the bridges 12 and 14 , including the DC link capacitor C, are from a voltage source 30 provided. The voltage source 30 is part of a battery, which also has an internal conductance, a decay constant shown as RC element and an inductance (summarized by the reference F). The (ideal) voltage source 30 and the elements labeled F represent a real battery 4a It is shown that a current i _C flows away from the intermediate circuit capacitor C, which coincides with that of the battery 30 resulting current to the current i _d results. This current i _d is shared between the two bridges 12 and 14 on.

The inverter 10 is with the two separate bridges 12 and 14 designed as a double inverter.

The 4b shows a comparable circuit with an inverter 110 , the two bridge circuits 112 . 114 includes. The winding systems 140 . 142 are electrically isolated and are supplied by different bridge circuits, which in turn are supplied together with DC voltage. The winding systems 140 . 142 are each star systems, the winding system 140 (three-phase) from the bridge 112 is fed (also three-phase), while the winding system 142 from the second bridge 114 (as well as three-phase as the winding system 142 ) is fed. The DC inputs of the bridges 112 . 114 be from a battery 130 Fed 'F'. reference numeral 130 denotes an ideal voltage source, while the elements represented by F 'represent an internal conductance, a decay constant and an inductance as inherent elements of a real battery. This corresponds to the representation of 4a , where are the elements 30 and 130 on the one hand and the elements F, F 'on the other hand. Parallel to the DC voltage inputs of the bridges 112 . 114 a DC link capacitor C 'is connected.

The bridges 112 . 114 each comprise three arms, which in turn each comprise two switches, namely a low-side and a high-side switch. The switches are labeled S, with their index indicating (first digit), whether it is a high side or low side switch. The second digit of the name of the switches indicates which phase the switch is concerned with. The last position of the designation of the switches indicates which of the two winding systems is driven, with a 1 for the first winding system 140 and a 2 for the second winding system 142 stands.

The winding systems 140 . 142 of the 4b are single windings. This also applies to the UVW and XYZ groups of the electrical machine 20 These belong to the same electrical machine. The second winding system 142 (please refer 4b ) is opposite to the first winding system 140 shifted by 0 °. The winding systems are in 4b in particular separate electrical machines that may be mechanically coupled. In the case of a positive mechanical coupling (or other coupling that does not permit translation or separation) of the two winding systems, the geometric offset would be constant. In the case of a force-transmitting coupling with a gearbox and / or a clutch, there is no constant geometric offset.

In 4a the winding groups are shifted by 180 °. In the usual double star motor, in which both winding systems are located on a common stator, the winding systems are usually offset by 30 ° or 180 ° or are not offset from each other at an angle, corresponding to an offset of 0 °.

The switching elements of the bridges are in particular semiconductor switches, for example transistors, in particular MOSFETs or IGBTS. The battery is in particular a Bordnetzakkumulator, such as lithium-based. The battery may have a rated voltage of 12 volts, 24 volts, and preferably 36 volts, 48 volts, or even 60-800 volts, especially 350-410 volts.

Claims (10)

A method for reversing a DC voltage for generating at least two pulse width modulated and respectively multi-phase voltage signals in a vehicle electrical system, the method comprising the steps of: - specifying a desired modulation degree (m); Determining, for the desired degree of modulation, a total carrier signal shift (V) for which a predetermined mapping which assigns associated current loads (B) for different modulation levels to different total carrier signal shifts (V) has a minimum of a current load of a DC link; - generating a first and a second pulse pattern having the desired modulation degree (m) and offset from each other by the total carrier signal shift (V); and - Delivering a first, pulse width modulated voltage signal (SIG1) according to the first pulse pattern; and - outputting a second, pulse width modulated voltage signal (SIG2) according to the second pulse pattern. The method of claim 1, further comprising: - determining a minimum for a current load of a DC link within a predetermined function, which assigns associated current loads for a given basic modulation level to different total carrier signal shifts (V); and - Determining that carrier signal shift that corresponds to the minimum for the given basic modulation level, based on the given function, as a constant carrier signal shift (K), which is a part of the total carrier signal shift (V). The method of claim 1 or 2, further comprising: Determining a variable carrier signal shift (D), which is a part of the total carrier signal shift (V) and for which the current load is minimal, starting from a difference between basic modulation degree and desired modulation degree m and starting from a different function, the different Assigning differences between the basic modulation degree and the desired modulation degree (m) to associated variable carrier signal shifts (D) for which the current load is minimal. Method according to one of the preceding claims, wherein the outputting of the voltage _signals comprises: switching of switching elements (S _HU -S _LZ ; S _HU1 -S _LW2 ) of all the arms of a plurality of controllable bridges ( 12 . 14 ; 112 . 114 ) or switching of switching elements of all arms of multiple controllable bridges to switching elements of an arm of the respective bridge according to a discontinuous pulse modulation method. The method of any one of the preceding claims, wherein dispensing the first and second voltage signals comprises: - Supplying individual windings (U-Z) of a first of a second winding system (UVW, XYZ) of the same electric machine. The method of claim 5, wherein supplying the single windings comprises: - Providing the first and second winding system as separate systems in star or delta connection. Method according to one of the preceding claims, wherein the phase voltages are modulated so that set sinusoidal phase currents. Method according to one of the preceding claims, wherein supplying the individual windings of the first winding system comprises: switching of controllable switching elements (S _HU -S _LW ; S _HU1 -S _LW1 ) of a first bridge ( 12 ; 112 ) according to the first pulse pattern and switching of controllable switching elements (S _HX -S _LZ , S _HU2 -S _LW2 ) of a second bridge ( 14 ; 114 ) according to the second pulse pattern wherein the first and the second bridge ( 12 . 14 ; 112 . 114 ) from the same voltage source ( 30 , F; 130 , F ') are supplied to which an intermediate circuit capacitor (C, C') of the intermediate circuit is connected in parallel. The method of claim 8, wherein the switching of the first and second bridges ( 12 . 14 ; 112 . 114 ) comprises: switching a first B6C bridge as the first bridge ( 12 ; 112 ) and a second B6C bridge as the second bridge ( 14 ; 114 ). Method according to one of the preceding claims, further comprising: supplying one or more electric machines ( 20 ; 140 . 142 ), which has at least two winding systems (UVW, XYZ, U1V1W1, U2V2W2), wherein the winding systems are offset in particular to each other in the angular position, and further wherein supplying is performed by delivering the first pulse width modulated voltage signal to the first of the winding systems (UVW; U1V1W1 ) and by outputting the second voltage signal to the second of the winding systems (XYZ; U2V2W2).

Images