DE19805207C2 - Procedure for determining a direction of movement - Google Patents

Description

The invention relates to a method for determining a direction of movement, in particular re an incremental position transmitter, according to the preamble of the independent An saying.

Incremental position transmitters are e.g. B. used for path or angle measurement. The position transmitter can be designed as a linear scale or as a rotatable scale and each carry a track with different markings with an even division. The markings are in the event of a relative displacement between a sensor and the Scanned electromechanically, photoelectrically or magnetically, each markie tion, which is assigned a certain length, is converted into an electrical pulse. The impulses are summed up, the count being a numerical representation of the from The distance traveled is the yardstick.

In addition to determining the position, it also provides information about the direction of movement needed, it is known to use two similar tracks that ge by half a division are mutually offset. Such a position transmitter is, for. B. from DE-A1 43 26 640 be knows. There is a coding disc with incremental division is described, which two by half a grating has staggered tracks. Both order and off However, evaluation is expensive because sensors and evaluation devices for two are independent traces must be provided. DE-A1 195 45 999 is a position transmitter described, in which only signals of a single track are evaluated and in which the direction of movement of the position transmitter is determined by the time sequence the state changes of two comparators are evaluated and with direction information be provided. However, if the comparators do not respond to signals at the same time, or in the case of a transition from one marking to another, in which more than If a comparator changes state, intermediate pulses are observed  incorrect interpretation of the direction of movement and / or the position of the position lead.

The invention has for its object a method for determining the movement to indicate the direction of a position transmitter with at least one track, the coding of which we has at least two states and the faulty intermediate pulses, which a move reversal of the position transmitter can only be faked, reliably recognized and eliminated can.

The object is solved by the features of the independent claim. Further and advantageous embodiments are to the further claims and the description remove.

The invention is based on an incremental position transmitter, which acts as a scanning medium at least one scale with an essentially periodic, but asymmetrical Pattern. The pattern consists of a series of similar, repeating patterns Sequences with at least three distinguishable markings. The markie Each influences a test medium in a clearly different way. Each Marking influences a sensor assigned to the position transmitter or the marking and leads to a specific output signal of the sensor for each marking, in particular with optical markings at different signal amplitudes. The exit Signals from the sensor are in a measuring and evaluation scarf assigned to the scale forwarded and compared there with threshold values, which threshold values in Signal transitions of the sensor output caused by adjacent markings signals are set, with a transition from one marking to the next an overlap causes a threshold value, and being from sensor output signals and temporal Sequence of sensor output signals generated when the scale is scanned direction of movement of the scale is determined.

The invention consists in that when a threshold value is exceeded, a time value is taken is triggered that an average time interval from adjacent overlap It is determined that two successive sweeps, wel  che follow each other faster than the mean time interval, either which are eliminated if they are crossed over before and after Sensor output signals are measured which correspond to unequal markings, or be assigned a change in direction of movement if the sensor output signals correspond to the same markings. Suitable marks are grayscale, in particular with the steps light, gray, dark. However, colored markings and / or a larger number of gradations and / or colors can be provided or not markings.

It is advantageous, in addition to the sensor output, during the scanning of the scale assign signals to a currently scanned marker at least one further signal NEN, which is used to clearly assign a sensor output signal to one of these sensors mark causing the output signal is provided. A very cheap training extension of the method is that sensor output signals individual markings after scanning the scale and assign a motion curve of the Scale is created, especially with a microcomputer. This is particularly easy and inexpensive.

The invention is explained with reference to exemplary embodiments shown in the figures tert. Show it

Fig. 1 shows an encoder as a grid scale according to the invention,

Fig. 2 is a series of gray scale with measured values in a moving direction,

Fig. 3 is a series of gray scale with measured values with reversal of movement,

Fig. 4 is a series of gray scale with measured values with Bewegungsoszillationen,

Fig. 5 is a movement trajectory according to the method of the invention.

The invention is described using exemplary embodiments in which three of each other the distinguishable markings are present in a sequence, which sequence is connected repeated several times along the position transmitter. Markers are preferred gene optically distinguishable, these reflecting or absorbing light differently rend or colored markings.  

The position transmitter can be a linear scale, in which the marking sequences along the Longitudinal dimension of the scale are arranged or a coding disc, at the markie approximately sequences are arranged in the circumferential direction of the disc.

The markings form an asymmetrical pattern. This can be a direction of movement tion can be easily determined since the change from one marking to the other due to the Asymmetry of the pattern when moving the scale is directional. The markie stanchions can be scanned either in reflection or in transmitted light. In Re flexion, a compact measurement setup is possible, while in the transmitted light method a small one gere adjustment accuracy of scale and optical transmitters and receivers is sufficient than in the reflection process, in which, for. B. an undesirable vibration of the scale Change the reflection conditions in an uncontrolled manner and falsify the measurement.

In a particularly simple embodiment, there are distinguishable markings in one Sequence of three shades of gray light, gray, dark, with such a sequence on the scale repeated several times. Grayscale are markings with under in the reflection process different reflectivities, which is large, medium and small for light, gray, dark. in the Transmission methods differ in the transmissivities of the markings and are for light, gray and dark, large, medium and low.

In a favorable arrangement, transmitter light is used as a test beam with an optical fiber on egg NEN position transmitter, e.g. B. a coding disc, sent and the on it Scale reflected or transmitted light captured with a detector array and analyzed. The reflection method uses optical fibers to emit the test beam and for receiving the test beam preferably arranged closely adjacent to one another. in the Transmission methods are preferably directly opposed to the sender and receiver fibers about. A detector arrangement, not shown, has at least one photodetector and an associated conventional amplification and evaluation circuit. In one Evaluation circuit becomes a position value and / or an angle value by counting Detector pulses, in particular comparator pulses, calculated by moving the Position transmitter caused. Usually the impulses are counted and  according to the pattern and / or the division of the scale into a position value or Converted angle of rotation value. Details of such a favorable arrangement are not further illustrated.

Fig. 1 is a schematic representation showing a scale 2 in the form of an encoder disk 1 having gray levels are net angeord in the marks in the circumferential direction on the edge of the disc 1. In order to determine an angle of rotation of the disk 1 , a test beam is sent onto the markings and reflected and / or transmitted radiation is detected in a detector arrangement. A transmitter sends e.g. B. by an optical fiber, a test beam on the scanning medium, and a photodetector receives the signals of the test beam, which are influenced by the scanned markings M1, M2, M3. Different light intensities are collected for under different gray levels M1, M2, M3 and preferably implemented in the detector arrangement in the usual way in voltage signals as output signal I of the photodetector. This output signal is passed on to comparators K1, K2.

In the case of colored markings, it is advantageous if these are formed by so-called primary colors, the primary color being understood to be a color which cannot be produced by a mixture of other primary colors, ie cannot be generated by a linear combination of the frequencies of the primary colors. It is useful if the basic colors of the pattern of a sequence have the same intensity. If the intensities are different, a mixture of color and intensity leads to ambiguous results and increases the evaluation effort. If red and green are provided as primary colors, yellow cannot be a primary color, since a mixture of red and green results in yellow. With n = 3 primary colors, 2 ⁿ = 8 colors can be generated. There must be as many threshold values as there are basic colors in a sequence. If, as with gray levels, intensities are used as a characteristic quantity for the scanning medium, n - 1 threshold values for n intensity levels are necessary. In the case of three gray levels, two comparators must therefore be provided. Are other, in particular linearly independent base variables used, such as B. primary colors red, green, blue, a threshold value and / or a comparator must be provided for each basic size.

In FIG. 2, the conditions upon movement of the scale 2 are explained relatively closer to a test beam. The top line shows several sequences of markings M1, M2, M3 in the form of light, gray, and dark gray shades, as can be observed when scanning the scale 2 in one direction. In the line below, the output signals I of the photodetector corresponding to the markings M1, M2, M3 are shown. Among them are the output pulses output by the respective comparators K1 and K2. In the line below, changes in the state of the comparators K1 and K2 with associated time values t1-t10 are listed, wherein when threshold values S1, S2 are passed over, a time value is taken and the time values t1-t10 are recorded. When changing from marker M1 to M2, signal I rises as well as when changing from M2 to M3. The response threshold S1 of the first comparator K1, which is set to a signal value between the signals of the markings M1 and M2, is swept over. The response threshold S2 of the second comparator K2 is set to a signal value between the signal values of the markings M2 and M3. For n markings within a sequence, n - 1 comparators with correspondingly set response thresholds S1, S2 are provided. The comparators K1, K2 emit a pulse at the comparator output from a first change of state of a comparator when its threshold value S1, S2 is exceeded to a second change of state when the comparator falls below its threshold value S1, S2.

By relative movement of the scale 2 relative to the test beam, the test beam sweeps markings M1, M2, M3, M1, M2, M3 and so on in one direction of movement. If the test beam changes from one marking to another within a sequence, the intensity of the detector signal rises accordingly and falls over two stages during the transition to the next sequence, which in turn begins with M1, then rises again and so on. The movement of the scale 2 relative to a test beam should expediently be constant or only slowly variable over at least several sequences of the markings M1, M2, M3.

When the output signal I rises during the transition from M1 to M2, a threshold value S1 of the comparator K1 is swept and a change in state of the first comparator K1 is triggered at a time t1. If a threshold value is exceeded, a time value is triggered. The rise or fall of the signal I does not follow abruptly, but has a finite slope. K1 outputs a signal at the output until it falls below S1 again. Exceeding the threshold S1 of the comparator is always interpreted as a transition from M1 to M2 or from M2 to M1. During the transition from M2 to M3, the signal I sweeps over the threshold value S2 of the second comparator K2, and the comparator K2 changes state at time t2. Exceeding the threshold value S2 is always interpreted as a transition from M2 to M3 or from M3 to M2, even if e.g. B. on the scale 2 from the marking M3 (dark) to the marking M1 (light). When changing from M3 to M1, when the test beam changes from M3 to M1, the decrease in the intensity of the signal I leads first to a change in state of the second comparator K2 at t3, then to a change in state of the first comparator K1 at t4. The time interval t4-t3 of these two sweeps of the threshold values S1 and S2 is significantly shorter than the time interval between the first two changes of state t2-t1 and t3-t2. The length of this time interval depends on the edge steepness of the output signal I and is not due to a marking M1, M2, M3 of scale 2 . With ideally steep signal edges, such a short interval would ideally not occur. In the measurement, however, two apparent transitions between two markings that follow one after the other are recognized. According to the method according to the invention, such a short interval is eliminated if it is not due to a reversal of movement. This makes it possible to determine a movement curve of the scale that corresponds to reality.

The slope of a real sensor signal I depends on several influencing variables. On The boundary between two markings is given to the photodetector partly because of the finite spatial size of the photosensitive surface, the convergence of the light beam and the size of the test beam, a light intensity that adjoins the two the markings are affected. This smearing of the received by the photodetector An optical signal is a major cause of sources of error in the determination the direction of movement of the scale. The steepness is partly due to the Size of the test beam, the detector area or optical entrance pupil for the detection send signal, the increment size and the relative speed between scale and Test beam affected.  

In Fig. 3 it is shown that a reversal of movement also leads to a short time sequence of sweeps of threshold values S1, S2, the time interval being significantly shorter than the average time interval of several successive sweeps of threshold values S1, s2, which Transitions from one marking to neighboring markings are caused. In this case, a short interval represents exactly the desired information about a movement reversal and must not be eliminated. The representation of the markings and signals essentially corresponds to that in FIG. 2. In contrast to a movement of the scale 2 or the scanning of the markings M1, M2, M3, which is essentially uniform in one direction and described there, occurs here at the time t8 shows a reversal of the scale 2 movement, which is shown in the representation of the markings in the top line of FIG. 3 by a change in the marking sequence M3-M1-M3-M1-M2 ... instead of M3-M1-M2-M3 ... is indicated.

Information which at least the times t, at which the threshold is swept Values S1 and S2 occur, which triggers the state change of the comparators K1 and K2 are, as well as the comparator K1 addressed at the time of the sweep or K2 are continuously preferably sent to a data processing unit, in particular forwarded to a computer and / or microprocessor. The individual markings can NEN, for optical markings z. B. small dimensions in the order of one have around one hundred micrometers, so that a high number of Data to be processed in a microprocessor. A microprocessor is in this phase fully utilized and cannot carry out any other processing steps in the short term. The on taken data have no immediately readable direction information. This is obtained from the recorded data in subsequent evaluation steps.

Each marking M1, M2, M3 can be assigned a value in the form of a numerical value. Colors, optical polarization states, intensities, grayscale and other types of marking can thus advantageously be described and processed in an analogous manner. It is particularly advantageous to carry out and evaluate such an assignment using corresponding data processing programs. The grading of the value between adjacent markings suitably depends on the number of comparators involved in the transition from one marking to another. So z. B. involved in the transition from M3 to M1 in Fig. 2 or Fig. 3, two comparators K1 and K2, while only one comparator K1 is involved in the transition from M1 to M2. If three gray levels are used as the marking, the marking light can preferably be assigned a value 2, the marking gray a value 1 and the marking dark a value 0.

Is the marker known, which is actually before and after a short interval lies, a change of movement can clearly be an apparent change of movement be distinguished. There is always a change of movement if it is a short one Interval adjacent marks are similar, i. H. same signals in the detector call. The neighboring marks before and after the short interval are not the short interval must be eliminated.

The markings can be recorded in parallel during the scanning, so that a current scanned marking is known at any time during the measurement and a de tector output signal is clearly assigned to an individual marker. This will preferably also detects an additional signal during data recording, with which the marking is identifiable. An advantageous embodiment is an assignment of output signals I to perform scanned markings only after the data has been recorded. This makes a very inexpensive solution possible because of expensive processor capacity and expenditure the sensors can be saved.

In FIG. 4 a situation is shown schematically, in which the scale 2 performs oscillation before a movement takes place in one direction. In this example, the scale jumps back and forth several times between the markings M1 and M2 (times t1-t5). In this case, threshold values S1 are scanned. The comparator K1 outputs a pulse from t1 to t2 as an output signal, between t3 and t4 and then between t5 and t6. The comparator K2 only outputs an output signal at t6, which is present until t8. Only from time t6 there is a movement in the direction of M3 and so on, as has already been done e.g. B. is described with reference to Fig. 1 and Fig. 2 with movement reversal.

The processing steps according to the method according to the invention are as follows. In the first Step a quantity of data is recorded and saved. The data include at least sweeps of the threshold values S1, S2 of each comparator K1, K2 with the respective because associated time values t. It can either be associated with the current marker M1, M2, M3 can be detected in parallel, so that the detector output signals at all times can be assigned to an individual marker. A cheap and cost-saving age Native to the simultaneous measurement of the increment color is the assignment to Mar in a second step after the data acquisition due to the chronological sequence to determine the state change of the comparators and the sweeps.

The assignment of output signals to an individual marking can be done according to the Determine data acquisition by number of consecutive overlaps of the threshold values of a comparator is considered. It is given as given take that a crossing of the threshold value of a comparator a change between two marks in one direction or the other, e.g. B. of gray after dark or from dark to gray for comparator K2. How many over is counted deletions one comparator delivers until the next comparator responds. If there is a close change of position of the first comparator K1, according to the interpretation of the measured values - in a direction of movement - either a transition from light to gray or a transition from gray to light. It is an even number of strokes and the first sweep of the closest comparator in time leads to a Marking with lower value, the starting marking is the lower value Color of the two possible markings of the first state in the case of three gray levels alternating The next comparator leads to a higher value In contrast, the starting marking is the higher marking of the two possible marks of the first swipe. If the number of overs counted is odd deletions, the situation is exactly the opposite. Knowing the order The direction of the markings of the stored sequences of sweeps is thus the movement of the scale. Starting from the first exit mark can a subsequent marking can be easily determined.  

The z. B. when changing from dark to light, short intervals are then eliminated and no longer lead to a falsification of the measurement. This is done in step 3 by comparing the time period of one or more time intervals between sweeps of threshold values S1, S2 with the time period of neighboring intervals. If the duration of an interval is significantly shorter than that of the neighboring intervals, and if the marking of the preceding and the following marking of the short interval is additionally different, then this short interval is recognized as an interval which cannot be assigned to any marking or reversal of movement and is eliminated by the two associated ones Overlaps of the threshold values S1 and S2 are merged to a common mean time, so that a single transition occurs and the threshold values S1 and S2 of the comparators K1 and K2 are simultaneously swept, expediently the mean time is within the previous short time interval, preferably in the middle of the interval.

Are the neighboring markings, which are measured before and after the short interval the, however, the same, there is a change in the direction of movement, so that the short In tervall must not be deleted in any case. The difference between the length of time a short interval and an interval between two come from a marker the sweeps depend on conditions comparable to those of the flank of the detector signals and must be adapted to the current test conditions sen or to determine. A short interval is significantly short if it is safe from the duration of the usual intervals can be distinguished. For a structure with optical fibers 700 µm in diameter and 1 mm long increments and a measuring speed A short increment of x mm / s is at most 30% of a normal increment. This difference can be reliably recognized.

The fourth step is based on those already determined in the previous steps Overlaps and exit markings determine the associated direction of movement. Since the order of the scanned markings of the scale is now known because of the asymmetrical pattern the direction of movement is known.  

In the fifth step, indices are assigned to the markings following an initial marking and a summary with time values stored in the first step. The indices correspond to the order of the markings, taking into account changes in the direction of movement of the scale 2 . Each pair of values from index and time value forms a point on a movement curve of scale 2 . Each transition from one marking to another triggers a counting pulse in the measuring and evaluation circuit. Accordingly, the movement of the scale 2 can be assigned to a path variable, in particular an angle or a path. Starting from an initial pulse value, the index of an initial marker, the number of output pulses from at least one comparator is counted and each time threshold values of the comparators K1, K2 are exceeded, the pulse number is incremented when the scale 2 moves forward and decremented when the sensor moves backward. The time value of the associated sweep is assigned to each pulse number and a path variable to be assigned to the pulse number is linked to the time value. The value pair of path size and time value form the movement curve.

Optionally, averaging of time values can take place in the sixth step. With that a Motion curve can be smoothed. At the same time, inaccuracies in the time value balanced. It is advisable to use a time value in the middle of an In as the time value tervalls to choose between two sweeps of threshold values. This is done for each index half the difference between the times of the over enclosing the interval deletions formed and added to the time value of the front pulse. It is very beneficial the procedure with a microcomputer and the processing steps with the help of one in it stored computer program to perform.

In Fig. 5, a movement curve obtained according to the method described is Darge provides. The movement of the scale 2 shows an increasing course over time up to approximately 160 ms, where a reversal of movement is found. The scale then moves slowly backwards between 160 ms and 200 ms.

Claims (9)

1. A method for determining the direction of movement of an encoder with incremental scale with repeating sequences of markings, each sequence having at least three markings acting differently on a scanning medium, a sensor receiving signals from the scanning medium, the output signals of which are assigned to a scale The measurement and evaluation circuit is forwarded and compared there with threshold values, which threshold values are set in signal transitions of the sensor output signals caused by adjacent markings, a transition from one marking to the next causing a threshold value to be crossed over, and with sensor output signals and a time sequence of one Scanning the scale generated sensor output signals, the direction of movement of the scale is determined, characterized in that when a threshold value (S1, S2) is exceeded, a time value is triggered that an average erer time interval from temporally adjacent sweeps is determined that two immediately successive sweeps, which follow each other faster than the mean time interval, are either eliminated, if sensor output signals (S) are measured during their previous and subsequent sweeping correspond to different markings (M1, M2, M3), or be assigned a change in direction of movement if the sensor output signals correspond to the same markings (M1, M2, M3). 2. The method according to claim 1, characterized in that during the scanning of the scale ( 2 ) in addition to the sensor output signals (I) a currently scanned mark (M1, M2, M3) at least one further Si signal is assigned, which for the unique assignment of a Sensor output signal (I) is provided for a marking (M1, M2, M3) causing this sensor output signal (I). 3. The method according to claim 1, characterized in that sensor output signals (I) are assigned individual markings (M1, M2, M3) after scanning the scale ( 2 ). 4. The method according to one or more of claims 1 to 3, characterized, that from the number of immediately successive sweeps a assignment of a sensor output signal (S) to a first threshold value (S1) keyed current marker (M1, M2, M3) is made by possible values of Values of two adjacent markings (M1, M2) based on the overlined first threshold value (S1) and the current marking (M1, M2, M3) with an even number of immediately consecutive sweeps towards the first threshold value (S1, S2) and a lower value of the following ones After the next threshold value (S2, S1) has been crossed, the lower one Value and at a higher value the following marking after the overlap The next threshold value (S2, S1) is assigned the higher value, ww rend the current marking with an odd number of strokes of the first threshold value (S1) and a lower value of the following marker (M1, M2, M3) the higher value after the next threshold has been crossed the following marking (M1, M2, M3) If the next threshold value is passed over, the lower value is assigned. 5. The method according to one or more of claims 1 to 4, characterized in that a movement curve of the scale ( 2 ) is created. 6. The method according to one or more of claims 1 to 5, characterized in that a movement curve of the scale ( 2 ) is created by starting from a predetermined or predetermined initial value of a marker (M1, M2, M3) the number of output pulses from comparators (K1, K2) is counted, each pulse a loading movement of the scale (2) is associated with each scan at Vorwärtsbewe supply of the scale (2) the pulse count is incremented and is cremented during reverse motion de, that each pulse number of the value of the associated Scanning is assigned and a pair of values is formed from the path variable to be assigned to the number of pulses and the associated time value. 7. The method according to one or more of claims 1 to 6, characterized, that a microcomputer for tracking and evaluating the scanning of the scale (2) is used. 8. The method according to one or more of claims 1 to 7, characterized, that colored markings are used. 9. The method according to one or more of claims 1 to 7, characterized, that colored markings and grayscale are used.