DE4413153A1 - Torque control method for an asynchronous machine - Google Patents

Description

The invention relates to a method for torque ment control of an asynchronous machine according to the preamble of claim 1 and can be both general in industry driven as especially in the case of electrical rails vehicles are used.

Target of almost all driving regulations is the only torque management actual value M of the induction machine quickly his desired torque value M _nom possible to track. The torque per pair of poles p of the induction machine can e.g. B. from the space pointer amounts of total flow || and rotor flux ||, as well as from the flux angle ϑ enclosed by these space pointers

can be calculated (see also etz Archiv Vol. 11 (1989), H. 1, pages 11 to 16), wherein

= Total flow space pointer
= Rotor flow space pointer
L _σ = leakage inductance.

The space pointer sizes used in the following can be after known rules from the corresponding three strand sizes be calculated (see also IEEE Transactions on power electronics, Vol. 3, No. 4, October 1988, pages 420 to 429). No practical ones exist for river chains usable transmitter, so that this is usually through a machine model can be calculated from measurable quantities. The rotor flow space pointer moves in stationary operation with almost constant angular velocity on one Circular path.

The desired torque is as distortion-free as possible reached, for example, when the total flow space pointer also with constant angular velocity on one Circular path is guided and with the rotor flow space pointer Flow angle ϑ spanned. This is on the premise fast switching power semiconductor almost ideally reali adjustable because the pulse period

of the feeding inverter is then small against the ro Gate scatter time constant of the machine and also in the whole Speed range small against the stator period. In glide chung (2) applies

f _T = switching frequency of the power semiconductors
T _σ = rotor scatter time constant
T _s = stator period
R _r = rotor resistance.

To a good approximation, this allows the simplifying assumption that the pulse-frequency scanned system is quasi-continuous behaves fairly and according to differentials

d / dt ⇔ Δ / Tp (3)

to change into differences. In the further derivations, Therefore, for all sizes, there is initially a continuous time delay runs are accepted.

The rotor flux of a squirrel cage asynchronous machine can change only slowly so that fast torque changes in principle only by changing the total flow amount or the flux angle between total flux and rotor flux can be aimed.

The power section of the drive is already stationary To optimally utilize operation, the induction machine must required, the greatest possible torque always with minimal Stator current at maximum stator voltage amount, so generate the maximum possible total flow amount accordingly. However, the maximum total flow amount must be taken into account the saturation of the stator iron to the rated flux ₀ be limited.

This leaves only the possibility of dynamically adjusting the torque via the flow angle. The stator angular frequency ω _S , with which the electrical quantities oscillate in the stator windings of the induction machine, corresponds in stationary operation to the sum of the electrically effective angular velocity ω of the rotor compared to the stator and the rotor angular frequency ω _r (short rotor frequency) with which the electrical Sizes swing in the rotor. For dynamic torque adjustment, the stator circuit frequency of the three-phase voltage system (stator frequency) fed into the stator windings must also contain a dynamic component proportional to the rate of change of the flux angle ϑ. The frequency of the stator is obtained according to the equation

where _S = stationary stator frequency.

The electrically effective angular velocity ω of the rotor opposite the stand (hereinafter referred to as electrical Speed)) is the product of the pole pair number ρ and the mechanical angular velocity (circle frequency) Ω of the rotor according to

ω = ρ · Ω (5)

given. The rotor frequency is resistive if the rotor is known position and the easy-to-calculate square of the rotor flow amount || ² clearly by the torque according to

certainly.

Target of all known torque controls for induction measurement It is clear that the stand given in equation (4) rate so that the torque is at its setpoint follows with the best possible dynamics and target and actual torque match stationary.

To achieve this task, pulse changers are becoming increasingly common three-phase drives are used. you will be often with the so-called "field-oriented regulation" fitted. However, this control concept has the following Disadvantage:

• 1. The implementation of the control in a coordinate system oriented at the rotor flux requires two very complex coordinate transformations, since the required measured variables are always available in static coordinates. The target string voltages for the pulse width modulation (PWM) that drives the supplying pulse-controlled inverter must also be in static coordinates. For the coordinate transformation into the chosen rotor flux-fixed reference system, the rotor rotation angle ε, which is usually determined as an integral of the rotor speed relative to the stator, and the position angle ε _{r of} the rotor flux space pointer relative to the rotor ε _r = ∫ω _r dt must be determined very precisely. Integration errors, particularly at high speeds, easily lead to unstable behavior. Furthermore, the trigonometric functions sin and cos must be calculated with sufficient accuracy from the transformation angle, for which purpose an extensive Sin / Cos table must generally be stored in the data memory.
• 2. Good torque dynamics are only with sufficient Re gel reserve achieved for the stator voltage amplitude, d. H. the stator voltage amplitude cannot already be maximum in stationary operation. The stationary Utilization of the drive is therefore not optimal.
• 3. The field weakening operation due to the technically be in principle limited the stator voltage amplitude Operating point-dependent setting of the total flow is carried out controlled, the Steuerge principles based on elaborate calculations ren. They are therefore not usually closed form and are only on microcontrollers quite complex, e.g. B. by means of curves, approximate feasible.
• 4. So far there is no ge in coordinates fixed to the rotor flux suitable strategy for dynamic torque adjustment without reserve of the stator voltage amplitude  known, so that especially in the field weakening range unsatisfactory torque dynamics achieved becomes.

The in etz archive vol. 11,. . . bypassed procedures because of the representation in resting coordinates Transfor mentioned above under number 1 as a disadvantage mations in rotor flux-proof coordinates.

However, the following disadvantages remain:
The setting of the total flow amount listed above under number 3 is carried out in the field weakening range by a PI controller. This eliminates the calculation of the control characteristics, but good torque dynamics are also achieved with this method only with a large control reserve of the stator voltage amplitude.

Another disadvantage is that the transient response of the control is largely determined by the integral part of the torque regulator. This is unavoidable in the strategy chosen there, since the integral part of the torque controller in stationary operation must always correspond to the stationary stator frequency _S according to equation (4). It takes on large values at high speed, so that the integral component cannot be limited. With dynamic changes in the operating point, undesirable oscillations occur.

Furthermore, the voltage drop across the stator resistance R _{s is} not explicitly taken into account, so that it must also be implicitly included in the integral parts of the torque and total flux amount regulator.

The invention has for its object a method for Torque control of an asynchronous machine of the beginning named type to indicate a highly dynamic torque  control in the entire speed range of a highly utilized Guaranteed three-phase drive.

This task is done in conjunction with the characteristics of Preamble according to the invention by the in the characteristic of Features specified claim 1 solved.

The advantages that can be achieved with the invention are in particular another in that due to the indirect torque control an optimal sta normal use of the machine and a very good rule of money namik (torque dynamics) can be achieved without doing so complex arithmetic operations are necessary. Despite the ver no dynamic reserves of change judge control (|| - = 1 = const., see equation (16)) Torque without settling (overshoot) stimulated.

The absence of the dynamic control reserve of the stands voltage amplitude (no voltage reserve, i.e. always maximum possible stator voltage amplitude) allows the Induction machine already in stationary operation with maxi to feed the painterly stator voltage amplitude. This will the required torque with minimal stator current testifies and the power section optimally without loss of dynamics exploited.

It is noteworthy that the stator current amount during the no significant increase compared to dynamic processes has stationary operation. This means that the Power section despite the good dynamics regarding the current does not have to be oversized.

Based on Ro's indirect torque control Gate frequency control can be used to make full use of the An  protective measures necessary to prevent tipping and electricity boundary can be realized in a simple manner.

Advantageous embodiments of the invention are in the Un claims marked.

The invention is described below with reference to the drawing illustrated embodiments explained. It shows:

FIG. 1, the selected coordinate transformation on When playing the stator voltage,

Fig. 2 is a block diagram of the torque and total torque flußbetragsregelung with indirect control by the rotor frequency control,

Figure 3 shows a torque control with precise Ständerfre frequency precontrol for optimal Drehmomentdyna mik.,

Fig. 4 shows the trajectory of the Gesamtflußraumzeigers in the field weakening when changing from a sta tionary operating point in a new rem with height torque,

Fig. 5 shows a basic structure of the torque control by subordinate Gesamtflußbetragsregelung,

Fig. 6 is a torque control with precise Vorsteue tion tudenverstellung of Gesamtflußbetrags and Spannungsampli,

Fig. 7 shows a torque control with dynamic field weakening with a limited voltage angle.

The space vector of the stator voltage to be applied to the stator terminals of the induction machine on average per pulse period results directly from the stator equation of the machine model with the condition made in equations (2, 3). The stator voltage space vector is obtained as the sum of the voltage drop across the stator resistance and the magnetization voltage, which corresponds to the temporal change in the total flux space vector _{µ in} the stator-coordinate system

The stator current space pointer can be in analog or digital ler shape can be easily measured. The whole flux space pointers are obtained using the stator equation according to the machine model

Its phase position Xµ corresponds directly to that for the Transformation into coordinates that are fixed to the overall flow Transformation angle for which according to equation (8) the following simple connection applies

The coordinate transformation selected here is shown for clarification in FIG. 1 using the example of the stator voltage:

α, β = stationary coordinates x, y = total flow-oriented, with rotating coordinate axes,
δ _u = tension angle between and rotated by + 90 °.

The following simple transformation equation applies:

in which

esα, esβ = coordinates of in the α / β system esx, esy = coordinates of in the x / y system.

Differentiation of the total flux space vector calculated in equation (8) and insertion of the result into the stator equation (7) leads with _μ = ω _s to the following equation for the stator voltage to be set:

in which
e _Rx , e _Ry = coordinates of in the x / y system
e _{Ψ x} , e _{Ψ y} = coordinates of in the x / y system.

Equation (11) contains no trigonometric functions, so that no sin / cos table is required. The division of the stator voltage shown in FIG. 1 into a component e _sx parallel to the total flow _{space vector} and an orthogonal component e _sy can be seen directly therein.

The component of the stator voltage oriented parallel to the total flow area pointer changes or corrects the total flow amount. Their amount e _sx is referred to below as the flow correction factor.

The amount referred to as the stator frequency factor e _sy of the component of the stator voltage that is orthogonal to the total flow space vector is proportional to the stator frequency and thus enables the torque adjustment.

The simplest implementation of a torque and total flow amount control is obtained if the flow correction factor e _sx as a manipulated variable of a total flow amount controller with PI characteristics according to the following equation (12) and the stator frequency factor e _sy by a PI torque controller according to the following equation (13 ) can be set. The following applies:

and

in which

V _Ψ = P gain of the total flow controller
τ _Ψ = reset time of the total flow controller
Ψ _should reference flux =
V _M = P gain of the torque controller
τ _M = reset time of the torque controller

As already mentioned, the I component is the torque controller Required if stator frequency precontrol is missing This is because he then enters the stationary stator frequency poses.

In Fig. 2 is a block diagram of the torque and Ge total flow control with indirect torque control by rotor frequency control, as it is used for torque control in the voltage range (amplitude adjustment). There are two computers 1, 2 can be seen, said machine 1 _to M and || ² receives and Rotorkreisfre quenzsollwert ω _Rsoll according to the equation

calculated and computer 2 receives the torque M and || ² received and the rotor angular frequency ω _r according to the equation

calculated. A subtractor 4 forms ω _rsoll -ω _r , where ω _{rsoll is} previously limited to a positive or negative peak value _r or by means of a limiter 3 . The difference determined by the subtractor 4 arrives at a rotor frequency controller 5 (preferably a PI controller in the arrangement according to FIG. 2), the output signal of which corresponds to the stator frequency frequency factor e _sy and is fed to a coordinate transformer 6 .

A subtracter 7 forms the difference Ψ _{is to σ} - ||, wherein Ψ _{is to σ} corresponds to the Bemessungsflußsollwert, and this difference to a Gesamtflußbetragsregler 8 (. In Anord voltage according to Figure 2 is preferably a PI-controller) to the output signal of the Flußkorrekturfaktor e _sx corresponds and is also supplied to the coordinate transformer 6 . The coordinate transformer 6 on the output side of the voltage space pointer s can be removed. In order to avoid the influence of fluctuations in the DC voltage, a multiplier 9 forms the product of s and

where E _{d corresponds to} the DC link voltage and E _{d0 corresponds to} the rated DC link voltage and k _u is defined as the voltage factor. On the output side, the multiplier 9 outputs the space vector of the inverter control in the stationary α / β coordinate system to a pulse width modulator 10 for controlling the power semiconductors of an inverter. The switching frequency of the power semiconductors is 1 kHz or above.

It is essential in the control according to FIG. 2 that the torque setpoint and actual value are first converted into the corresponding rotor frequencies according to equation (6) and that the torque by the controller 5 corresponds indirectly via the rotor frequency

is regulated. This is advantageous since anti-tipping protection and current limitation, as will be described in the following, can easily be implemented by suitable limiting functions (e.g. equation (30a, 30b) of the rotor frequency _setpoint ω _rsoll .

The manipulated variables of the controller for the total flow amount and rotor frequency directly generate the coordinate e _{sx of} the stator voltage space vector parallel to the total flow space vector (= flux correction factor) and the orthogonal coordinate e _sy (= stator frequency factor). They implicitly contain the voltage drop across the stator resistance R _S. The coordinate transformation then carried out in accordance with equation (9), which only corresponds to the consideration of the current phase position of the total flow space vector, provides the stator voltage space vector in stationary coordinates. Finally, the stator voltage space pointer still has the reciprocal of the voltage factor

designated ratio multiplied from the measured DC link voltage E _d and rated _{DC link} voltage E _d0 .

The rated _{DC link} voltage E _d0 is _chosen to avoid overdriving of the inverter, insofar as these are impermissible, so that the inverter delivers the rated voltage ê₀ of the fed induction machine at maximum sine output.

The room pointer of the inverter control calculated in this way

is then passed to the pulse width modulator (PWM). Pulse Width modulation methods are state of the art and who not dealt with further here.

A significant improvement in the torque control behavior compared to the basic version according to FIG. 2 is achieved if the rotor frequency control described above is supplemented by a suitable precontrol of the stationary stator frequency _s .

In Fig. 3 this torque control with precise stator frequency pre-control is shown for optimum torque dynamics. The basic arrangement is as shown under Fig. 2 ge. The additional components of the control are explained below.

An adder 11 sums _to ω and ω, and supplies the stationary stator circuit frequency reference thus formed _{S to} a multiplier 12 to. The multiplier 12 calculates the orthogonal to the total flow space stationary component _{Ψ y of} the magnetization voltage as the product of _{s soll} and Ψ _{soll 0} and feeds it to an adder 13 , which forms the stationary stator frequency factor _sy from e _Ry and _{Ψ y} and one between the rotor frequency controller 5 and Coordinate transformer 6 arranged adder 16 passes. To form e _Ry , a coordinate transformer 14 is provided which calculates the components i _sx and i _{sy of} the stator current space vector in the x / y system. A multiplier 15 forms the product e _Ry = i _sy · R _s . Between total flux regulator 8 and coordinate transformer 6 , an adder 17 is arranged, to which e _Rx = Ψχ and eΨxD are supplied, a multiplier 18 forming the product e _Rx = i _sx · R _s .

It is essential in the stator frequency precontrol according to FIG. 3 that it is based on simple analytical equations, which can also be exactly solved in real-time conditions on signal processors. First, the setpoint of the stationary stator frequency _{s soll is determined in} accordance with equation (4) as the sum of the electrical speed ω and the rotor frequency _setpoint ω _rsoll calculated from the torque setpoint in accordance with equation (6).

By multiplying the stationary stator frequency setpoint by the rated total _{flux setpoint} Ψ _{µsoll 0} it is obtained with equation (11) directly the component _{Ψ y of} the magnetizing voltage space vector which is to be set in the stationary operation with constant total flow amount _{and is} orthogonal to the total flow space vector.

Further, to the gonal to the total flux ortho share e _Ry of the voltage drop across the resistor Ständerwi added and one obtains the steady-state value of the orthogonal to the total flux content of the _sy STAs derspannung (= stationary stator frequency factor). The coordinate transformation of the stator current space pointer carried out for this purpose corresponds only to a multiplication by the conjugate complex total flow space pointer and subsequent division by the total flow amount ||.

These operations can be done on a signal processor too accurate under real-time conditions with little effort be carried out.  

The rotor frequency controller now only delivers as a manipulated variable nor the dynamic required to change the flow angle Proportion of the stator frequency (= dynamic stator fre quenz factor share). This disappears in the stationary loading driven and the I component of the rotor frequency controller can theo are not necessary because the stator frequency is stationary is piloted. In practice, however, he corrects pre tax errors caused by mismatch of model parameters or measurement errors can be caused.

However, these are so small that the I component is very small ne values can be limited and no longer influence has the dynamic torque control behavior. The torque ment can be started very quickly and without overshoot be regulated.

It is consequent to also pre-control the amount e _Rx of the component of the voltage drop across the stator resistance that is parallel to the total flow of the room pointer, because the total flow amount regulator is then also relieved in a stationary manner and can be designed as a simple P-controller. It only intervenes if the total flow amount is to be changed.

With an unlimited amount of the stator voltage space pointer could the machine in the entire operating area with the Be measurement flow ₀ are operated. With the above be written rotor frequency and total flux amount control could then, because of the arbitrarily high control reserve Stator voltage amplitude, regardless of the speed very good torque dynamics can always be achieved.

In practice, however, this is not the case since the Bemes stator voltage ê₀ must not be exceeded can.  

The equations (11, 16) make it clear that the design flow with full control of the inverter (|| = 1) even if the voltage drop e _R at the stator resistance is neglected, only up to a design frequency

can be maintained. The rated frequency is determined by the rated flux ₀ and the rated voltage ê₀ of the machine as well as the voltage factor k _u .

Still higher stator frequencies than the rated fre quenz ω₀ set, the total flow must correspond to the so-called field weakening number

be weakened so that the stand then required voltage amplitude just with fully controlled change judge is reached. This will power the Drive optimally used in stationary operation.

The torque control described below by un overall flow rate control (dynamic field wea chung) assumes that the machine is in the field weak area always with maximum constant stator chip voltage amplitude is fed. For the stator voltage room pointer then applies

Its amount corresponds to the rated value and it is only possible to adjust the tension angle δ _u shown in FIG. 1 between the space vector of the stator voltage and the total flow space vector rotated by + 90 °. The tension angle

is directly due to the ratio of the stator voltage coordinates given naten.

The inverter is already full in stationary operation controlled, there is in principle no dynamic rule serve the stator voltage amplitude and the Path velocity of the total flow space pointer cannot be increased more.

In order to increase the torque quickly under these conditions, the flow angle ϑ between the total flow space pointer and the rotor flow space pointer must be increased quickly in a different way according to equation ( 1 ). The large rate of change of the flow angle required for this can only be achieved by a suitable shortening of the total flow path curve at a constant path speed of the total flow space pointer.

To illustrate this process, FIG. 4 shows an example of the trajectory of the total flow space pointer in the field weakened when changing from a stationary operating point to a new one with higher torque, whereby

_C = stationary field weakening number at point C,
_A = stationary field weakening number at point A,
_µC = total flow _{space pointer} at point C,
γ _min = _minimum field weakening number Δ = total flow area pointer increase relative to point A

The extent of the new stationary Ge reached in point C. velvet river circuit is due to the increased torque and  the higher stator frequency required is less than at the old circular path left at point A.

The transition between the operating points A, C is carried out optimally quickly if the total flow space pointer at point A is brought to a negative value on the linear connection between A and C chosen as the shortest possible dynamic curve by suddenly reducing the voltage angle δ _u . The reduction in the voltage angle is achieved according to equation (20) by reducing the portion e _{sx of} the stator voltage that is parallel to the total flow of the space vector.

During the total flow space pointer through the shortcut path runs, the rotor flow space pointer remains approximately open his with almost unchanged angular velocity traversed stationary trajectory. The Flow angle ϑ essentially due to the path shortening increases and the torque increases. With decreasing Torque control deviation becomes the tension angle again enlarged. Has reached the torque - for example already in point B - its setpoint, so the stator chip voltage amplitude suddenly reduced so that the torque ment does not overshoot.

While the total flow space pointer then continues to point C, the modulation and the total flow amount are adjusted so that the torque matches the setpoint. When the total flow area pointer finally reaches its new stationary circular path in point C, the total flow amount is no longer changed. The parallel to the total flow space component e _{sx of} the stator voltage space generator becomes very small and the stator voltage space pointer is again almost perpendicular to the total flow space pointer according to equations (19, 20). The voltage angle δ _u again reaches approximately zero, because δ _u assumes a small positive end value in order to cover the voltage drop across the stator resistance.

The turning angle δ _uA at point A determines how quickly the torque increases. It can theoretically assume values up to 90 ° in accordance with equation (20), so that the total flow space pointer would then be led directly from the point A in FIG. 4 through the origin. The machine would be de-energized at this moment and there would be no more torque.

This effect is avoided if the amount of the tension angle is limited to approximately δ _u = 60 °. The limitation of the voltage angle can easily be done according to equation (20) by keeping the condition

will be realized. The limit is chosen so that the Average rate of increase in torque maximum is.

Fig. 5 shows the basic structure of the torque control by subordinate total flow amount control. A subtra here 30 forms the signal MM _should and passes this difference to a torque controller 31 . Another subtractor 32 forms from the total _{flow setpoint} Ψ _µsoll at the output of the controller 31 and the amount of the total flow space pointer | µ | the difference Ψ _µsoll - || to be _{fed to} a total _{flow controller} 33 . On the output side, the controller 33 outputs the voltage angle δ limited to a positive and negative maximum value -u, -u with the aid of a limiter 34 to a coordinate transformer 35 . The output signal

the coordinate transformer 35 is fed to a machine model 36 , which forms the torque M and the total flow space pointer = || e ^jx _-Ψ . is fed to the coordinate transformer 35 and an absolute value generator 37 , where in the latter the size || forms and passes to the subtractor 32 .

The superimposed torque controller 31 with PI characteristic tracks the torque according to its setpoint by specifying the radius of the total flow trajectory with the subordinate total flow amount controller 33 with P characteristic. The manipulated variable of the total flow rate controller 33 corresponds to the voltage angle δ _u defined in equation (20), which is limited in accordance with equation ( 21 ). According to equation (19), the stator voltage space vector is clearly given by the voltage angle δ _u and the angular position X _{Ψ of} the total flow space vector. The machine model 36 provides all the sizes required for the control.

With the basic structure described above without dyna The control reserve of the stator voltage amplitude becomes a good torque pickup behavior achieved.

If the torque setpoint jumps from idling to at the given speed maximum achievable torque he increases, the torque follows the setpoint despite constant Stator voltage amplitude with good dynamics. The total River space pointer suddenly becomes almost li linear shortcut path. Due to the used li Near controller for torque and total flow amount remains the shortcut path is not linear, but nestles with a decreasing control deviation to the new stationary Circular path. Their length is therefore not minimal and that Rise time correspondingly somewhat longer than theoretically possible Lich.  

When the torque reaches its setpoint, it vibrates oh ne further measure because of the integral part of the Torque controller currently set total flow setpoint is still too small.

In order to prevent the torque from overshooting, the stator voltage amplitude and thus the Bahnge speed of the total flow space pointer must be suddenly reduced at the moment when the torque reaches its setpoint. How the stator voltage amplitude can be reduced is described in FIG. 6.

In Fig. 6 a torque control with exact before control of the total flow amount and voltage amplitudes adjustment is shown. The basic structure is as described under FIGS. 2 and 3. The signals e _Rx = Ψχ and _sy are fed to an absolute value generator 19 , which uses this to generate the stationary amount || of the stator voltage space pointer. A multiplier 20 generates the product from || and 1 / ku and feeds this to a reciprocal value generator 21 . The stationary field weakening digit which can be tapped at the output of the reciprocal value generator 21

is limited with the aid of a limiter 22 to values less than or equal to 1 and fed to a multiplier 23 (|| = stationary amount of the space vector of the inverter control). The multiplier forms the product of Ψ _{soll 0} and the stationary field weakening digit limited to 1 and feeds this to the subtractor 7 .

It has already been shown above that the dynamic behavior of a control system is considerably improved if the integral components of the controller are replaced by suitable precontrol systems. Therefore, the total flow amount set stationary in the basic structure by the I-channel of the torque controller is precisely precontrolled by means of the additional measures according to FIG. 6.

First, the amount || the inverter control calculated according to equation (16) in the event that the desired stationary stator frequency while maintaining of the design flow could be achieved. The reciprocal this amount of levy is then called a statio nary field weakening rate according to

set. This ensures that the required Stator frequency exactly with maximum stator voltage ampli tude is achieved. The subsequent limitation of the field Weak digit to values less than 1 prevents stan frequencies below the rated frequency with impermissible sig high total flow amount can be set.

The total flow rate controller 8 receives the total flow setpoint

Ψ _µset = Ψ _{µset 0} · (23)

It corresponds to the product of the total flow setpoint Ψ set ₀ set in the voltage setting range and the stationary field weakening factor calculated according to equation (22).

In the control described under Fig. 6, the over-oscillation of the torque after reaching the setpoint is avoided by suddenly reducing the stator voltage amplitude. Furthermore, the control structure presented is easy to implement on a signal processor even under real-time conditions because of the low cost.

In Fig. 7, a torque control with a dynamic field weakening is shown with a limited voltage angle. The basic structure is as shown in FIGS. 2, 3 and 6 on. In addition, a proportional element 24 is provided, which forms the product of eΨyD and V _{M Ψ} and feeds it to a subtractor 25 , V _{M Ψ representing} a P gain. The subtractor 25 forms a dynamic field weakening number γ _D by calculating the difference between 1.0 and the product of the proportional element 24 . A multiplier 26 forms the product of the dynamic field weakening digit γ _D and the stationary field weakening digit. This product is limited to 1 with the help of the limiter 22 and arrives as a field weakening number γ for multiplier 23 , which multiplies γ by Ψ _{should be 0} and supplies the flow setpoint thus formed to the subtractor 7 .

A multiplier 27 forms the product of e _sy and tan 60 ° and feeds it to a computing element 28 . This computing element 28 realizes the condition specified under equation (21) and receives the sum signal of the adder 17 on the input side and outputs e _sx to the coordinate transformer 6 on the output side.

With the dynamic field weakening controller 24/25/26 , the principle of dynamic field weakening described above is implemented. This controller weights the ge pre-controlled field weakening digit according to equation (22) with the dynamic field weakening digit

γ _D = 1 - V _M · e _{Ψ Ψ y D} (24)

The dynamic field weakening _figure is obtained directly from the _{manipulated variable} e _{Ψ yD of} the rotor frequency controller 5 . The manipulated variable is controlled by the additionally inserted proportio nalglied 24 with the P gain V _{M Ψ} attenuated, so that the stability of the torque control loop is ensured with dynamic field weakening.

The torque control described in FIG. 7 with voltage amplitude adjustment is characterized in that the voltage amplitude theoretically possible in the absence of dynamic control reserve of the stator torque control times are almost achieved.

The good torque dynamics in the voltage setting range are retained since the rotor frequency controller 5 can have an undiminished effect on the amplitude of the inverter drive. The torque overshoot can thus be practically completely avoided.

The additional P-element 24 takes into account that the mean stator frequency per pulse period in the voltage setting range can be influenced by the suddenly adjustable stator voltage amount, but in the field weakening range can only be influenced by the continuously adjustable total flow amount.

The total flow setpoint is finally obtained as a product from the total controlled according to equation (23) flow setpoint and the dynami calculated in equation (24) field weakening rate according to

Ψ _to Ψ = _{to 0} · γ _D (25)

With highly utilized rotary field drives, it is imperative Measures to limit current and prevent tipping the dynamic behavior of the drive is not may be affected.

The following are suitable measures for the current limits tion and the anti-tipper specified. It will be the approach here selected, both protective devices by suitable restrictions implementation of the rotor frequency setpoint. this leads to to solutions with little effort, even under real time conditions are processed on a signal processor can.

From the equivalent circuit diagram of the induction machine (see at for example IEEE Transactions. . ., Picture 4) the stand can current according to

be calculated. The following also applies in stationary operation the relationship between the rotor frequency and the flux angle

ω _r · tanθ = Tσ (27)

The rotor flow space pointer is then by the flow angle and the total flow space pointer according to

certainly.

With the equations (26... 28) the amount of the stand current space vector as a function of rotor frequency and Ro gate flow amount accordingly

being represented.

Equation (29) shows that a given limit || _{Max of} the stator current amount when limiting the rotor frequency ω _r

respectively.

is observed. The plus sign stands for the Be limit in engine operation and the minus sign accordingly for limitation in generator operation.

The anti-tipper must be designed so that a rule reversal of meaning is reliably prevented without the dynamic ver hold the drive impair. How he started thinks that the anti-tipper is limited by suitable limitation of the Rotor frequency realized.

To do this, the torque curve must first be a function the rotor frequency can be calculated. It is convenient the torque to the so-called design breakdown torque pro Pole pair

to obtain, the maximum with the rated flow stationary can be reached. In addition, frequencies become corresponding

n = ωT _σ (32)

related to the reciprocal of the rotor scatter time constant T _σ . By using the equations (27, 28, 31, 32) in equation (1), the related torque m is obtained when the voltage drop across the stator resistance R _{S is} neglected as a function of the related rotor frequency n _r at a related speed n₀

For the exact implementation of the anti-tipper, the ro tor frequencies must be known at which the motor tipping moment or the generator tipping moment occurs. For this purpose, the extreme points of equation (33) with known analytical methods, such as. B. determined the Cardanic formula. In this way, solution functions are obtained which contain the related tilt rotor frequencies _r for the motor area and _r for the generator area as a function of the related speed n.

If one takes into account that the torque-rotor frequency characteristics are very flat in the vicinity of the tipping point, the tipping protection can be implemented very easily. For this purpose, the related stationary rotor frequency n _r stationary in the motor area is limited to the constant upper limit value _rm , at which the limit characteristics as functions of n _r for operation with maximum stator current || Cut _max and maximum torque. In the generator area, the related stationary rotor frequency n _{r is} steadily limited to the likewise constant lower limit value _rm , at which the limit characteristics for operation at maximum speed n _Max and operation with minimum torque intersect.

The limitation of the stationary rotor frequency has none Influence on the dynamic operating behavior, since the Stell size of the rotor frequency controller is not limited. There stationary, but is zero, this in turn has no effect on the anti-tip.

With the tilt protection described above, the actual tilting moment of the machine can be used almost completely (| m _max | <0.99 m _tilt ). When attempting to dynamically regulate a target torque that is too large, the maximum achievable torque is initially set with unlimited dynamics. Only when the rotor flux drops does the limitation of the rotor frequency setpoint n _{rsetpoint begin} to the limit value _rm defined according to the previously described method and the torque drops to the stationary overturning moment.

It should be added to the above explanations that it is generally the physical current and the current standardized to the rotor short-circuit current I _∞ . The rotor short-circuit current I _∞ = ₀ / L _δ would flow if the main inductance magnetized to the rated flux ₀, the rotor was short-circuited and the rotor current would only be limited by the rotor leakage inductance L _δ . This normalization of the current leads to simple equations.

Claims (10)

1. A method for controlling a torque powered by a pulse width modulated controlled inverter squirrel-cage induction machine, characterized in that torque value (M _soll) and actual torque value (M) corresponding to and in the corresponding rotor _circuit frequency _setpoint (ω _rsoll ) and the rotor _{circuit frequency} (ω _r ) are converted and that the torque via a rotor frequency controller and a total flux amount controller accordingly and is regulated, whereby
R _r = rotor resistance
= Rotor flow space pointer
esy = stator frequency factor
esx = flow correction factor
V _M , v _Ψ = P gains of the controller
τ _M , τ _Ψ = readjustment times of the controller
Ψ _soll = Gesamtflußsollwert
= Total flow space pointer. 2. The method according to claim 1, characterized in that to form the stator voltage space vector (s) from the stator frequency factor and the flux correction factor a coordinate transformation with respect to the stator frequency factor and the flux correction factor accordingly is carried out with
x _μ = phase position of the total flow space vector. 3. The method according to claim 2, characterized in that the stator voltage space pointer (s) to form the space pointer of the inverter control () with the reciprocal of the voltage factor is multiplied, whereby
E _D = _DC link voltage
E _d0 = rated _{DC link} voltage. 4. The method according to any one of claims 1 to 3, characterized in that from the rotor _circuit frequency _setpoint (ω _rsoll ) and the electrical speed (ω) of the stationary stator circuit _frequency setpoint ( _ssoll ) is formed that the water setpoint with the rated _{total flow} setpoint (Ψ _µsollo ) is multiplied so that the stationary component ( _{Ψ y} ) of the magnetizing voltage space pointer (Ψ) thus obtained, orthogonal to the total flow space vector (Ψ) is summed with the portion (e _Ry ) of the voltage drop (R) orthogonal to the total flow vector over the stator resistance, and that the so contained stationary stator frequency factor (sy) is added to the dynamic stator frequency frequency component (= e _{Ψ yD} ) formed by the rotor frequency controller. 5. The method according to claim 4, characterized in that a dynamic flow correction factor portion (e _{Ψ xD} ) formed by the total flow amount controller is added to the total flow space parallel portion of the voltage drop () over the stator resistance as a stationary flux correction factor (e _Rx ). 6. The method according to claims 4 and 5, characterized in that the stationary stator voltage space vector (||) from the stationary stator frequency factor ( _sy ) and the stationary flow correction factor (e _Rx ) is formed accordingly a stationary field weakening digit () is formed, that this field weakening digit is limited to values less than one and that the design flux setpoint (Ψ _sollo ) is multiplied by this limited field weakening digit, where ê₀ = rated stator voltage, = stationary space pointer of the inverter control. 7. The method according to claim 6, characterized in that the stationary field weakening digit () is multiplied by a dynamic field weakening digit γ _D = 1-V _{M Ψ} · e _{Ψ yD} , where V _{M Ψ} = P-gain. 8. The method according to claim 7, characterized in that the amount of the tension angle (δ _u ) between the stan derspannungsraumzeiger (-) and the Ge rotated by + 90 ° total flow space pointer () is limited to about 60 °. 9. The method according to any one of claims 1 to 8, characterized in that for limiting the current, the rotor frequency setpoint in motor operation and on is limited in generator operation, whereby
L _μ = magnetization inductance,
L _σ = leakage inductance, || _Max = specified limit of the stator current of the room pointer amount. 10. The method according to any one of claims 1 to 9, characterized in that for the anti-tip protection, the stationary rotor frequency (n _r ) based on the reciprocal of the rotor scattering time constant is limited in motor operation to the constant upper limit value ( _rm ) at which intersect the characteristic curves for operation with maximum stator current (|| _-Max ) and operation with maximum torque () and that the related stationary rotor frequency (n _r ), which is constant in relation to the reciprocal of the rotor scattering time, stationary in generator operation to the constant lower limit value ( _rm ) is limited at which the limit curves for operation at maximum speed (n _Max ) and operation with minimum torque () intersect.