DE3332567C2 - Arrangement for controlling an asynchronous machine fed by fast electrical actuators in the field weakening - Google Patents

Abstract

The rotor EMF vector, the determination of which is described in the main application, is used for indirect flow measurement and flow control in the field weakening range. A vector phase-locked loop controlled by this rotor EMF supplies the frequency and direction angle information required for this. For this vector phase-locked loop, additions and expansions are given which enable precise and fast speed control even at the lowest speeds in the vicinity of standstill.

Description

This object is achieved by the characterizing features of claim 1.

Due to the measures specified in claims 5 and 6, the range of high control quality can also be achieved in the creep speed range can be extended in the execution after the main avoidance only controlled operation is possible.

In the present application, only control arrangements with a voltage-impressing actuator, in labeled PWR (pulse inverter) in Figures 3 and 4. The transfer to stream-indenting Actuators is easily possible.

Figure 1 shows an exemplary embodiment of the phase-locked loop as it is expedient in the current state of the art and tailored to the present application. Input variables are the components of the rotor EMF, e _2tt and e _lß, formed according to the equations in the preamble of claim 2 .

As output values, the circuit supplies the instantaneous values of y ,, the angle between the most distant and the coordinate system oriented towards the rotor emf and thus the runner flux (with the designations a, β or x, y), the time derivative of this angle, such as in the main application with ω, whose magnitude and sign, and also the components of the direction vector cos y _{ and sin y ,. In block 11, a voltage corresponding to the yi error is formed. Block 12 is the follow-up controller, a PI controller whose output ω- is determined. From this, in a known manner, with modules of analog technology | a> \ | and sign ω, formed. The amount is converted into a pulse train via a voltage-frequency converter 14. An up-> n <i down-counter module 16 turns it into a digital (binary) value for y, controlled by sign <y. The functions cos y \ and sin y \ are assigned to this via digital read-only memories.
The multipliers 111 and 112 in block 11 are then expediently multiplying digital-to-analog converter modules. The resolution can be ώ = 9 bits , for example.

The main advantage of using this technique compared to the main application is that it is no longer necessary to switch the measuring range to determine the rotor emf, that the adjustment and setting are significantly simplified and that, in addition to y \ , ω \ is also available for signal processing. Figure 2 shows the approximate determination of the rotor flux.

According to the invention, use is made of the fact that the amount of the rotor emf is the product of the Amount of | <u, | and the magnitude of the runner chaining flow, and it becomes the quotient

used as a substitute for the flow (B i 1 d 2a, blocks 24 and 25). B i 1 d 2a also shows the assignment of the flux default value to the speed as an example, block 21, the flux controller, block 22, whose output signal after limitation, block 23, represents the set or default value for the magnetizing current component, / f _y . Set or default values are always marked with an asterisk in the following.

B i 1 d 2b shows one way of determining the EMK amount. The »-./^- components are in themselves known way converted into 3-phase (a, i, c) components and these with a three-phase bridge circuit rectified.

If one would use the rotor EMF in the field weakening range directly, z. B. to a constant amount, this would have the disadvantage that the loop gain ω \ would be proportional. The flux control can therefore also be understood as an EMF control with adaptation of the loop gain by means of <y. With this procedure, the non-linear magnetization characteristic and the influence of the heating of the winding on the rotor time constant are practically irrelevant. The possible simpler design of obtaining an approximate dynamic image for the flux from the default value or the actual value of the magnetizing current component via a first-order delay element is also always better than a corresponding differentiating element in the default value of the magnetization component. The voltage reserve is limited in the field weakening range and it is therefore not certain that the specified values for the two current components can always be achieved. In the work "State of the art in regulated three-phase drives" by R. Jötten, VDE-Fachberichte 1978, it is explained why an overdistribution must also be possible for the magnetizing component and that it is beneficial to have a separate limitation in the specification. The flux regulator enables this over-adjustment if the limitation already shown in B i 1 d 2a for the magnetizing current component is higher than its nominal value i _lyN .

B i 1 d 3 shows a speed control with flux control in the field weakening range and a decoupled control of the current components. In B i 1 d 3, block 36 represents the rotor EMF determination known from the main application. Block 32 contains the above-described phase-locked loop controlled by the rotor EMF. The complex 31 outlined by a dash-dotted line represents the flux control. A> _m denotes the measured speed, block 311 the characteristic curve assignment of the flux setpoint to the measured speed, block 318 the formation of the amount of the rotor emf vector. The dividing stage 317 uses the value for] ω _χ | taken from block 32 from this the equivalent controlled variable for the flow, denoted by | ψι \, compare B i 1 d 2. The setpoint for the magnetizing current component / f _r is the limited sum of the flux controller output signal and a constant base value /, _fV behind block 313. The complexes 30, speed control, and 33, decoupled control of the Current components are known per se, as are the coordinate converters 34 and 35.

34 converts the voltage specification into three-phase voltage coordinates, 35 the measured rotary current values into the components / Ί _v and i _{t r} , the actual values for the current control. The current control receives the setpoint values in a manner known per se from the speed control 30 (/ f,) or the flux control 31 (/ f _r ). For the regulation in the J

For the flux, the field weakening range is the substitute measured variable formed from the amount of the rotor EMF.

det. j

One advantage of the decoupled control is that it j Very good use of control limits. The indispensable advantage of using the rotor j

IiM K-formation according to the main application isl that slets information obtained from measurements on the machine for the orientation of the currents and voltages is available, even if because of the limited voltage reserve the actual values of the controlled variables, current components and flux, compared to the default values static or have dynamic errors.

The question then arises as to how the assignment of the Flow preset values as a function of the "speed <y," should be designed so that it is also in the field weakening range good regulation dynamics are maintained.

¹ According to the invention, this is done by assigning the flux to the speed at each point of the field weakening range so that with the available voltage reserve in the intermediate circuit, a specified maximum value of the torque-generating current component; * _v can be achieved without the need to change the flux is.

This is achieved by a characteristic curve according to claim 3.

In order to utilize the converter, it is particularly advantageous if it can work in the so-called block voltage mode without intermediate clocks. Then, however, only the rotational frequency of the stator voltage is available, and via the same with a limited resolution, the instantaneous value of the spatial angle of the stator voltage vector is available as a manipulated variable. In a further variant, use is now made according to the invention of the knowledge that in the field weakening range at the voltage limit the internal torque is a function of the angle between the rotor EMK Vckicr and the strength savings vector, hereinafter referred to as δ . The following applies with a good approximation for the internal (electromagnetic) torque:

,,,, = λ-κ \ ₂ · - · II -sin 2 J (with the number of pole pairs 1).
2 L ₁ L ω _χ J

The other symbols have the following meanings:

I »ι Lv = - U _d

K ,, = MAL _x L ₂ - M ² )

According to the invention is therefore for a method for torque control in the field weakening range in block voltage operation, the z. B. can be part of a method for speed or frequency control, the limited to ± 45 ° difference angle between the rotor EM K vector determined according to the main application and claim 1 and the specification of the stator voltage vector δ in B i 1 d 4 used as a manipulated variable for the required torque current component and the instantaneous direction of the stator voltage vector and thus the switching state of the voltage-impressing inverter are specified so that the directional angle of the stator voltage increases

Y ,, = Yt + <$ + 180 °

results. A closer examination of the dynamic conditions shows that a PID controller is required for this is.

Figure 4 shows an embodiment in which the individual blocks mean the following:

Blocks 35,36,32,318 and 317 are identical to those in B i 1 d 3, with the functions of coordinate conversion, rotor EMF calculation, vector phase-locked loop, generation of the EMF magnitude and generation of a substitute measured variable for the flux magnitude, in this order. A limited setpoint value i * _x for the torque-generating current I _tx , the main controlled variable, is made available from the specified torque value m * by dividing it by the flux. The magnetizing component i _iy adjusts itself freely. The output of the PID controller 42 corresponds to the angle δ = arctg u _ly / u _x and is limited in 42 b. After the analog / digital conversion in 43, the direction angle of the stator voltage vector is formed in the adder 44 in accordance with the above relationship.

The value for y comes in digital form from the vector phase-locked loop 32 already described above, shown in Figure 1. The conversion of the angle information into the inverter switching state is not shown in detail, since it is described in the literature and is relevant Is known to those skilled in the art

The main application explains that with values of <ai close to zero, special measures are necessary

because the rotor emf is approaching zero here. They are necessary in order to be able to start from a standstill, to brake to a standstill and to be able to reverse through the standstill point. This was done there by switching to a different control method at very low speeds. The vector phase-locked loop used in this additional application now makes it possible, with some additions according to the invention, to solve this problem even better. Furthermore, in the case of speed control, they also enable regulated operation with creep speeds close to Nail. In the case of speed control with an analog speed measurement value, Fig. 5 the necessary addition. The basic idea is to switch to the method described in DE-PS 12 50 914 in the slow speed range and a> \ as the sum of the measured speed <y ", or n" · ω ' _α with n _p pole pairs, and the desired slip circle frequency ω * to be specified. As stated in the main application, the latter must be made proportional to the default value for the torque-bearing current component. The controller 12 in B i 1 d 1, which is specified in normal operation ω _{χ, is switched to kieinem |} © ι] blocked according to the invention, ie its output value z. B. by means of a field effect transistor or an analog switch construction

Stein blocked, ie set to zero, and also switched from flux control to control of the magnetizing current component to the nominal value. In B i 1 d 5, 53 represents the switching element. In B i I d 3, the addition is clarified by adding the additional input signals <y ″, and / f. _v and a signal acting on the flow regulator are added in dashed lines. The latter blocks the flow regulator, ie sets its output to zero with the help of an analog switch module, and the constant input value / | _; . ^ in Figure 3, complex 31, determines the setpoint Zf ₁ ..

The described method, which causes problems of accuracy at higher speeds and has been the reason for increased expenditure for the measurement base, is precisely in the slow speed range, where <y _{m is of} the same magnitude.

'< order of magnitude is like <y ₂ , experience has shown that this is completely unproblematic.

'If creep speeds are required in closed-loop operation, there is often a position control, and thus a position measurement. The measured value is often absolutely digitally coded or available as a counter reading behind an incremental pulse generator. In this case, you can at creep speed range to the known per se adding the angle y ", and y ₂ y to angle pass. This is done by modifying the vector phase-locked loop according to B i 1 d 6. In the creep speed range, the controller 12 is blocked from B i 1 d 1 and instead of its output signal ω \ proportional to the torque- generating setpoint current component if _x with the known proportionality factor (51, see. B i 1 d 5) is fed. The count of counter 16 now represents Y2. The rotor position angle y _m must now be added in an additional adder 62 in order to obtain Xi. y ", iiegi after the above either absolutely coded before, or with an incremental encoder it must be formed by a second counter module, 61 in Fig. 6.

The additional measures according to the invention described here make it possible to achieve an extremely high speed control range from a few per thousand of the basic speed to a multiple of the basic speed high accuracy and the best possible rise and settle times with speed control with moderate effort for the measurement base to be realized.

In addition 6 sheets of drawings

Claims (6)

Patent claims: 1. Arrangement for carrying out the method for operating a fast electrical actuators with voltages and currents of variable amplitude and frequency fed asynchronous machine with a the default signals for the control of the steep links forming setpoint and actual value signals in vector form processing device, the motion EMF induced in the barrel winding in a stand-fixed Signals representing the reference system according to the equations e _la = L ₁ IM · [u _u - R ₁ ■ i _la - [UK _n ] · d / _{I (r} / di] e _lß = L ₁ ZM [U ₁ P - R ₁ H ₈ ~ 1 ^ K _n ] - di, j, / dr]
with
is K _n = L ₁ I [L _x ■ L ₁ - M ¹ ] are formed, characterized by a vector phase-locked loop (32) known per se, containing a frequency controller (12) _{and which forms the components of a direction vector COSy 1} and sin y _x as well as the rotational frequency ω _χ , by coordinate converters (34 , 35) for converting the controlled variables into the actual stator variables and possibly vice versa, to which the components of the directional vector (sin y _u cos y _x ) are fed, and by a flow control (31) that serves as a substitute actual value for the amount of the rotor flux linkage the quotient of the amount of the rotor EMF vector and the amount of the stator frequency ω _{χ is} supplied. The symbols used have the following meanings:
25th Indices: 1 = stator size;
2 = rotor size, a, ß: Coordinates in the fixed, right-angled reference coordinate system.
Quantities: / = current component;
u = stress component, e = EMF component,
ω, = instantaneous value of the differential angular speed between the rotor interlinking Cuss and reference axis fixed to the stand = temporal derivation of the differential angle,
Parameter: L = total inductance of a winding, for example
35 L ₂ = rotor winding inductance when the other windings are de-energized, R ₁ = stator winding resistance, taking into account that the values for calculating with the components are 2/3 of the phase quantities used in classical theory,
M = mutual inductance between stator and rotor winding 2. Arrangement according to claim 1 with an individual control of magnetizing and torque-generating current components via the voltage components in the axis directions and signal connections decoupling the two controls, furthermore an amount-limited controller output as the setpoint of the flux control (31) for the magnetizing current component and one via a characteristic curve of the measured speed associated flux setpoint, characterized in that the output variable <o \ is used for decoupling (Fig. 3). 3. Arrangement according to claim 2, characterized in that when the limit voltage of the intermediate circuit is known the characteristic according to the equation ,. if = Vl (C / "t,) /« J ^: - (ο · if,) ² is designed in such a way that a specified torque- _{generating? * x} can be achieved at every operating point without changing the field. c, = Hn ■ U ₁₁ (U /. DC link voltage of the pulse inverter) - ^J "" -γ- -H ₂ ) - cos pv (cos φ ^: nominal power factor of the machine) C ₃ = ä | / « _m , v (ωΓ stator frequency at <y _m =><a _miV , Ψ _1ν . = '/ ^y _2l ., _v and / _u . = -7f _v ) M, Lu L ₁ , R \, R ₁ according to claim 1. 4 Arrangement according to claim 1 with a voltage-impressing inverter, which is connected to the voltage The application relates to a control arrangement for carrying out a control method for asynchronous motors according to the preamble of claim 1. Such a control method is the subject of the main patent 3212 439 2-32. The control method is described, which - instead of the field as previously usual ^{: is based} on the rotor EMF, based on a co-ordinate system that is fixed to the stator. The rotor EMF is based on the measured quantities for stator current and stator voltage according to the formula specified in claim 1 "Sef lecture" Control methods for good dynamic performance induction motor drives based on current and voltage as measured quantities ", in accordance with the attached lecture manuscript the Interne-! ional sSifconducto? Power Converter Conference IEEE / IAS Orlando in May 1982, i.e. before the filing date, manuscript no. CH 1682-4 / 82 0000-0397 00.75 / 1982, had essentially the content of the main patent for ; nfle oiromKonipuiicnut vuijvoi-uvu i ^ - ~ Setpoint via a limiter from aem urenn.umcm default value after division by that from the rotor emf determined flux value, get its actual value from the current measurement via a coordinate converter, that the output variable of this PID controller is used as the default value for the lead angle * of the terminal voltage vector serves against the rotor EM K-vector and is limited to ± 45 °, and that finally the Direction _{angleTy o of} the terminal voltage vector, which determines the switching state of the voltage-impressing inverter, according to the relationship ¥ " = K ₁ + δ + 180 ° is determined by summation, the phase angle y of the motion EMF vector being available as the output variable of the phase-locked loop (32) (B 11 d 4). _n , 5. Arrangement according to claim 1 with an additional device for creep speeds with analog speed measurement and speed control, characterized by a device (53) that controls the control. (12 in Fig. I) in the vector phase control loop (32) blocks at low speeds and through a summing point (52) behind the controller, in which the stator frequency ω, as the sum of the measured mechanical speed « _m , a default value wf proportional to the nominal value of the torque-generating current component and, if applicable, the controller output signal (Fig. 5) 6 Arrangement according to claim 1 with an additional device for creeping numbers when there is one absolute or incremental, digital rotor position measurement, characterized by a device (53) which blocks the controller (12 in B i 1 d 1) in the vector phase control loop (32) at low speeds a summing point (52) arranged behind the controller (12) for adding one to the setpoint of the torque 2, ment-forming current component proportional default value «*, through a first, the summing point (52) via a voltage-frequency converter (14) downstream up-down counter (16) by a second up-down counter (61) connected downstream of the incremental encoder, and by an Add.erer (62) which the counts are added (B i I d 6). ^^ The invention relates to the development of the method described in the main application door operation of the machine in the field weakening range. As is known, this is understood to mean the increase in frequency and speed above the basic frequency or basic speed with an approximately constant terminal voltage and power. This means that the flow of the machine in this speed range must decrease approximately inversely proportionally with increasing speed. For this purpose, the setpoint or default value is assigned to the amount of the linkage flow of the rotor to the speed in the form of a characteristic curve. The U ₁ V) coordinate system is always oriented so that the ^ axis points in the direction of the flux vector. The Frfahrung has shown that the stability of the control loops at high speed SENR sensitive to errors in the orientation of this IXJ. · "Coordinate system is" and that therefore a controlled setting of the flux vector with dor rotor position as a measuring base is no longer feasible here A circuit for relay-oriented control for asynchronous motors is known from the "Control Engineering Practice and Process Computing Technology" issue 9, Se-U 217-221 (1973) and the "S.emens Zeitschrift" issue 10, pp. 761-764 (1971) "Control engineering practice" describes how the rotor nut of the asynchronous machine " _h : _{n =} -.... A _0n (ilpmmpnsnanmmeen and the terminal currents of the machine can be obtained, whereby the wind resistance and" the leakage inductance of the machine are to be taken into account . from the Siemens magazine the use of there as "vector filter" designated vector Phasenregelkre.sen is known. there, too, "•, the components of the direction vector sin φ and cw ψ ^d for the designated there as a vector rotator Koor 'Da NachteHdergenannten circuits is in that either Hallsohden in the machine or non-feedback integrators in the arithmetic circuit have to be used to form the flux vector existing ofiset voltage oe? Measuring elements or the operational amplifier unusable. . ... The aim of the present application is to provide an arrangement for carrying out the aforementioned Rule to create a procedure that provides an accurate, fast and stable control even with extreme field monitoring enables.