DE3243759A1 - Method and device for forming the angle and/or angular velocity of a converter-powered drive - Google Patents

Abstract

In this method for determining the angle (.) or the angular velocity ( omega ) of a converter-powered drive, a digital pulse generator (1) is provided which upon rotation of the drive transmits generator pulses (P(t)) to a pulse counter (3). The pulse counter determines the number of the generator pulses (n) arriving in a prescribed scanning interval. Using extrapolation methods, the angle during the period (tr) between the end of the scanning interval and the last preceding generator pulse is additionally formed and added as correction quantity to the counting information (Z0) of the pulse counter (3), which is weighted in accordance with the angular or phase increment (.i) of the pulse generator (1). The result of the very precise measurement of angle or angular velocity thereby achieved is a good true-running characteristic of the drive even at low rotational speeds. <IMAGE>

Description

Method and device for forming angles

and / or angular speed of a converter-fed drive The invention relates to a method and apparatus for forming Angle and / or angular speed of a converter-fed drive according to the preamble of claim 1. The rotational variable "angle" can be the translational variable "position" and the rotary variable "angular velocity" replaced by the translational variable "speed".

Such a method is from T. Ohmea, T. Matsuda, K. Kamiyama, M. Tachikawa, "A microprocessor-controlled high-accuracy wide-range speed regulator for motor drives ", IEEE-IECI Proceedings, Nov. 1981, pages 381 to 386.

For measuring the angular velocity from the pulses of a pulse generator two methods are generally known.

A first method is based on the counting of the one Pulse generators generate pulses within fixed measuring times. The counting result is proportional to the mean speed of the drive during the measuring time, however, very large relative measurement errors occur at lower speeds. Drives that only work with this type of measurement process are in the lower speed range very poor concentricity.

A second well-known method is based on the measurement of the Time between two pulses coming from the pulse generator. The measuring rate is proportional to angular velocity. The dead time of the measuring system increases with decreasing speed, which leads to unsatisfactory control behavior of drives at low speeds, who use this measurement method.

In a third method according to the literature reference mentioned at the beginning the encoder pulses are synchronized to the encoder pulses during a measuring time is counted. This method has the same disadvantages at lower speeds like the second method on.

On this basis, the invention is based on the object of a method to form the angle and / or angular velocity of a converter-fed Specify drive of the type mentioned, which points in all operating points of the Drive a satisfactory control behavior, in particular a good concentricity behavior enables.

This object is achieved by the features characterized in claim 1 solved.

The advantages that can be achieved with the invention are, in particular, that the sampling times from outside can be given. as Sampling times can, for example, use the ignition times of the converter will. The values for the angle and / or the angular velocity is always formed with great accuracy, in particular also at low drive speed.

This enables a good concentricity of the drive and a satisfactory one Control behavior.

Advantageous embodiments of the invention and a device for Implementation of the method according to the invention are characterized in the subclaims.

Embodiments of the invention are shown below with reference to the drawings explained.

They show: FIG. 1 a basic circuit diagram of a drive current converter feed, 2 shows a linear extrapolation for the formation of that which occurs in a sampling interval Angle, Fig. 3 a linear extrapolation for the formation of the angle with a strong reduction the angular velocity, Fig. 4 a pulse counting and evaluating device for the formation of angle or position and / or angular velocity or speed, FIG. 5 a variant for the formation of the angular velocity.

In Fig. 1 is a basic circuit diagram of a drive converter feed shown with angle or angular velocity detection. A DC machine M is fed with a direct voltage Udc via a converter Str. Input side the converter Str is supplied with an alternating voltage Uac.

The DC machine M is via a shaft with a pulse generator 1 connected. When turning the machine M resp.

the disk of the pulse generator 1 are two pulse chains offset by 900 PA (t), PB (t) issued by the pulse generator 1 to a pulse counting and evaluation device IP. The pulse generator 1 has a fixed number of angle increments t 1, the middle one Angular increment is denoted by ti.

The pulse counting and evaluation device IP evaluates the incoming Pulse chains and outputs a corresponding angle (or position) # or # (k) and a corresponding angular velocity (or velocity) CJ or CJ (k) a regulating and control device RS.

The regulating and control device RS receives further regulating or measured variables X from the converter Str as well as command values W and gives corresponding ignition pulses Z to the converter Str.

The pulse counting and evaluation device IP itself is described below as well as those for the precise determination of the angle (or position) and / or angular velocity (respectively.

Speed).

In Fig. 2 is a simple embodiment for a linear extrapolation for the formation of the angle occurring in a scanning interval. With K is the consecutive number of a sampling interval. The sampling interval with the number k has the duration tz (k), the sampling interval with the number k + 1 the duration tz (k + l) etc. The limits of the sampling interval are indicated by the pulses I (tz) marked, i.e. begins with each occurrence of a pulse I (tz) (= sampling time) a new sampling interval. In the example chosen, occur in the k-th sampling interval n (k) = 3 encoder pulses P (t) des with the drive via a shaft connected pulse generator 1, wherein for the extrapolation example according to FIG it is assumed that n (k)> O.

With each occurring encoder pulse P (t) a line increases # (t) (= counting information) in steps of an angle increment tp i of the pulse generator 1. To create an angle or position and / or angular velocity or

The speed of the drive is the angle # (k) at the top Interval limit of the sampling interval with the number k (= end of interval) is determined or the angle difference ## (k) = # (k) - # (k-l) of the angle between the upper and lower Interval limit (= between the beginning of the interval and the end of the interval).

There are several options for linear extrapolation to education of T (k) or ## (k). In the exemplary embodiment, the similarity of the triangles with the corner points A, B, C and C, D, E are used. For the angle difference ß ## (k) the following applies: = n (k). # i - # E (k-t) + # E (k) where T E (k) or # E (k-l) is the limited one is extranolated angle at the end of the interval or the beginning of the interval. The limited extrapolated Angles fE (k), #E (k-l) are obtained by limiting the unlimited extrapolated Angle # EO (k), # EO (k-l) to the minimum value from f EO (k) and t '. The angular velocity # (k) is calculated from the angle difference ## (k) as follows: # (k) = ## (k) / tz (k) To determine T E (k) = # EO (k) (in the example according to FIG. 2 there is no limit required on # i) the remaining time tr (k) to be measured either directly or indirectly needed. The remaining time tr (k) is the time between the upper interval limit of the sampling interval with the number k and the last previous encoder pulse P (t). It applies to the linear extrapolation for n (k)> 0: # EO (k) = n (k). #l 'tr (k) / tz (k) + tr (k)) = # E (k) where fEO (k) = distance DE, n (k). ## = section BC, tr (k) = section CD, tz (k) + tr (k-1) -tr (k) = segment AB.

The remaining time tr (k-1) and the extrapolated angle tE (k-1) of the previous one Sampling intervals with the number k-1 are available as stored values.

With this the angles T (k) and # # @ (k) can be calculated: f (k) = fo (k) + tE (k) where for the angle corresponding to the number of encoder pulses #o (k) applies: #o (k) = #o (k-l) + n (k). #i where the angle #o (k-l) is also used as stored value is available.

In Fig. 3 a further example with linear extrapolation is shown, with a strong angular velocity reduction at the beginning of the k-th interval is assumed so that the number of encoder pulses in the k-th sampling interval n (k) = O will. Since n (k) = O and the remaining time tr (k-1) may also be very short a modified linear extrapolation method is selected for this case, that first of all the unlimited extrapolated angle # EO (k) yields: #EO (k) = #E (k-l). tr (k) / tr (k-l).

This follows from the similarity of the triangles with the corner points F, G, H and F, I, K. The segment FG (= trCk-l)) is related to the segment FI (= tr (k)) as the segment GH (= fE (k-1)) to the segment IK (= #EO (k)). Since p #EO (k)> i becomes #EO (k) limited to #E (k) = # i. The thus determined angle # # (k) or the angular velocity # (k) = ## (k) / tz (k) represent much better approximations than when using the unlimited value # # EO (k) would be worked, resulting in better control behavior low speeds is reached.

In Figures 2 and 3 are simple examples with linear extrapolation shown. By extrapolation of a higher order, in particular of the second order and / or Switching on an acceleration value a of the drive can improve the Extrapolation can be achieved.

In Fig. 4 is a pulse counting and evaluating device IP for formation of angle or position # (k) and / or angular velocity or velocity # (k) shown. The pulse generator 1 with the angle increments? I generates two around 900 offset pulse trains PA (t) and PB (t), which are in a direction detection circuit 2 to pulses (encoder pulses) P (t) and a direction signal for the direction of rotation of the drive Sig (P) can be processed. In one of the direction detection circuit downstream pulse counter 3, which in the exemplary embodiment consists of an up-down counter whose counting direction is controlled by the direction signal Sig (P) (V / R input = Forward / backward input), they are entered via a clock input CL Pulse P (t) v are counted and are located as counting information Z (t) (= step-shaped Line).

The pulse counter 3 can alternatively contain two counters, wherein in a counter only the encoder pulses P (t) while the drive is running forwards and in the other counter only the encoder pulses while the drive is running backwards be enumerated. The counting information of this pulse counter 3 is read out when both counters can be read out at the same time.

A time counter 4 is at its reset input with each pulse P (t) R to a fixed value, here reset to the value 0. In the time counter 4 a pulse train with a fixed frequency f = constant is counted in via a clock input CL.

When an externally prescribed one occurs, the individual structural units each Uber-Ladeeingange L inputtable sampling time, here by the pulse I (tz) is marked, v is the counting information Z (t) from the pulse counter 3 read out and into a downstream first memory element 5 as counting information Zo (k) inscribed. The SD memory element 5 receives the necessary pulse I (tz) via its loading input L. At the same time, the previously in the first storage element 5 contained information Zo (k-1) in a downstream second memory element 6 transmitted, to which the pulse I (t2) is also fed via the charging input L.

The time counter 4 is also read out at the same sampling instant and its content corresponding to the remaining time tr (k) in a subsequent third Memory element 7 inscribed. For this purpose, the pulse I (tz) is at the charging input L of the Memory element 7. At the same time, the previously contained in the third memory element 7 is information as the remaining time tr (k-1) in a downstream fourth Transfer memory element 8, which for this purpose also receives the pulse I (tz) via its charging input L receives.

In a first subtracter connected downstream of the storage element 6 9 becomes the difference # Zo (k) = Zo (k) - Zo (k-1) between the contents Zo (k) of the first Memory member 5 and the content Zo (k-l) of the second memory member 6 is formed (= Difference between two consecutive items of counting information = n (k)).

This difference Zo (k) is in a first multiplier 10 with the mean angular increment Ti of pulse generator 1 is weighted and ## O (k) = # l . #Zo (k) obtained (see route BC in Fig. 2).

The circuit shown in the example for linear extrapolation with Limitation comprises the following members 11 to 21, namely a control member 11, a multiplier 12, a divider 13, a subtracter 14, an addition point 15, a divider 16, a multiplier 17, a selector 18, an addition point 19, a limiter 20 and a memory element 21.

In the circuit shown, there are two extrapolation devices as an example contain. The first extrapolation device comprises the members 12 to 15, the second extrapolation means comprises members 16 and 17. The controller the control element allows you to choose between the two extranolation devices 11, which controls the selection element 18 on the output side. In the example, the control takes place only as a function of the difference t Zo (k), in such a way that at # Zo (k) # 0 Via the selection element 18, the first extrapolation device (elements 12 to 15) and with # Zo (k) = 0 the second extrapolation device (components 16, 17) to be selected.

In Fig. 4 is a further possibility of controlling the control member 11 shows the input of the remaining time tr (k) determined by the time counter 4 (dashed lines Connection between memory element 7 and control element 11).

In the case of the first extrapolation device, the second Subtractor 14 formed the difference tz (k) - tr (k) and the addition point 15 supplied. This forms the sum tz (k) - tr (k) + tr (k-1) and carries this sum the divider 13 as a divisor.

The remaining time tr (k) is applied to the divider 13 as the dividend. The divider 13 forms the quotient tr (k) / (tz (k) + tr (k-1) - tr (k) (= ratio of the distance CD to route AB according to FIG. 2) and forwards this quotient to the multiplier 12 to. The multiplier 12 has the value ## O (k) = as a further input variable n. # i (route BC according to FIG. 1). On the output side, the multiplier gives 12 the unlimited extrapolated angle # EO (k) (= distance DE according to Fig. 2) to the selection member 18 on.

In the second extrapolation device, the divider 16 entered the remaining times tr (k) as a dividend and tr (k-1) as a divisor and that Ratio of two consecutive time counter samples tr (k) / tr (k-l) formed. The preceding sampling interval is extrapolated in this ratio Value fE (k-1) converted to the new value # EO (k) with the aid of multiplier 17 (see FIG. 3) and the selection member 18 supplied.

In Figures 2 and 3 it is assumed that a in the addition point 19 acceleration value to be added to the output value of the selection element 18 a of the drive is zero.

In the limiter 20 connected downstream of the addition point 19, the extrapolated Value # EO (k) limited to the range between + ti and - Yi. That is the value t E (k) available at the limiter output. The value f E (k) becomes that of the pulse I (tz) specified sampling times into the memory element connected downstream of the limiter 20 21 taken over.

To calculate the angular velocity # (k), the already calculated value ## O (k) is used.

The value ## O (k) becomes in an addition point 23 that in a subtracter 22 calculated difference from two successive extrapolation values # E (k) - # E (k-l) added. The value # E (k) is at the output of the limiter 20 and the value fE (k-1) can be tapped off at the output of the memory element 21. So that is Dividend ## (k) for a divider 24 connected downstream of the addition point 23 certainly. The divisor fed to the divider 24 is the one specified externally Sampling interval duration tz (k). The angular velocity # (k) = ## (k) / tz (k) can be determined To determine the angle # (k), the count information is used Zo (k) of the first memory element 5 is used. Zo (k) is used in a multiplier 25 is weighted with the mean angular increment # i and # o (k) is obtained (see Fig. 2). The limited extrapolated value is added to the value T #o (k) in an addition point 26 Angle T E (k) added, with which t (k) is determined.

A variant for determining the angular velocity C) (k) is shown in FIG. The angle # (k) is, as described under FIG. 4, as Output signal of addition unit 26 obtained. In addition to the arrangement according to Fig. 4, a memory element 27 is provided, which to the sampling times takes over the angle t (k). For this purpose, the pulse I (tz) is a charging input L of the storage element 27 at. With the aid of a subtracter 28, the difference between two The following angle values # ## (k) = t (k) - (k-1) are determined and the divider 24 as Dividend supplied, where # # (k) at the output of the adder 26 and # (k-l) am Output of the memory element 27 can be tapped. The divisor is again the sampling interval duration tz (k). This enables the angular velocity # (k) to be calculated. In the Arrangement according to FIG. 5, the subtracter 22 and the addition point are omitted 23 of the arrangement according to FIG. 4.

Claims (16)

A n p r ü c h e 1. Process for forming an angle or position and / or Angular speed or speed of a fed from a converter electric drive in which a pulse generator is provided for angle or position measurement is, which emits encoder pulses to a pulse counter, the read counting information of the pulse counter is weighted according to the angle or position increments of the pulse generator is characterized by the following features: - the counting information (Zo) of the pulse generator is read out at any sampling times (I (tz)) that can be specified externally, - the time (tr) between a sampling time and the encoder pulse (P (t)) is measured by means of a time counter, - the measured values of the time counter become a Corresponding correction angle (<E) or correction position determined. 2. The method according to claim 1, characterized in that the period of time measured between a sampling time and the first subsequent encoder pulse will. 3. The method according to claim 1, characterized in that the period of time measured between a sampling time and the last previous encoder pulse will. 4. The method according to any one of the preceding claims, characterized in, that for the formation of angle (t or position of the correction angle (fE) or correction position to the weighted read out counting information ((#o) of the pulse counter is added. 5. The method according to any one of the preceding claims, characterized in, that for the formation of angular velocity (W) or velocity the change () of the angle () or position between two consecutive sampling times and this change is determined by the duration (tz) of the time interval between them Sampling times is divided. 6. The method according to any one of the preceding claims, characterized in, that to determine the change (##) of the angle (#) or position between two one another the difference between the correction angles tf E) or correction positions formed at these sampling times and for the weighted difference (t yO) of the counting information (Zo) of the pulse counter is added at these sampling times. 7. The method according to any one of the preceding claims, characterized in, that the correction angle (/ E) or correction position by means of an extrapolation method are formed. 8. The method according to claim 7, characterized in that controlled or a choice is made between different extrapolation methods. 9. The method according to claim 8, characterized in that the controller to select the extrapolation method from the difference in the counting information ( Z0) of the pulse counter depends on the two previous sampling times is. 10. The method according to claim 8, characterized net, that the control for the selection of the extrapolation method from the measurement value (tr) des Time counter is dependent. 11. The method according to any one of the preceding claims, characterized in, that a first extrapolation method is selected when the difference in the counting information (a Zo) is not equal to zero and a second extrapolation method is selected, when the difference in the counting information is zero. 12. The method according to any one of the preceding claims, characterized in that that the angular acceleration or linear acceleration of the drive as an additional value is introduced in the extrapolation. 13. The method according to any one of the preceding claims, characterized in that that the correction angle (E) or correction position to a predetermined or guided Limit value is limited. 14. The method according to claim 13, characterized in that the amount of the limit value corresponds exactly to an angle or position increment (fi) of the pulse generator. 15. Device for carrying out the method according to one of the preceding Claims, characterized in that the pulse counter (3) is an up / down counter contains whose counting direction by a direction detection circuit (2) as a function the direction of rotation of the drive is controlled. 16. Device for carrying out the method according to one of the preceding Claims, characterized in that the pulse counter (3) contains two counters, whereby in a counter only the encoder pulses (P (t)) while the drive is running forwards and in the other, only the encoder pulses are counted while the drive is running backwards will.