Classifications

• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B19/00—Programme-control systems
  • G05B19/02—Programme-control systems electric
  • G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
  • G05B19/414—Structure of the control system, e.g. common controller or multiprocessor systems, interface to servo, programmable interface controller
  • G05B19/4142—Structure of the control system, e.g. common controller or multiprocessor systems, interface to servo, programmable interface controller characterised by the use of a microprocessor
• • G—PHYSICS
  • G05—CONTROLLING; REGULATING
  • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
  • G05B19/00—Programme-control systems
  • G05B19/02—Programme-control systems electric
  • G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
  • G05B19/416—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by control of velocity, acceleration or deceleration
  • G05B19/4163—Adaptive control of feed or cutting velocity

Definitions

• This invention relates to an adaptive pulsing system for controlling a motor which moves a device to a desired position.
• the invention is useful wherever precise position control is desired, and is particularly useful if it is required that the final position always be approached from the same direction. This requirement is found in systems for which the position information is derived from measurement of lead screw rotation, since slack in the driving mechanism can lead to an error in measurement.
• Application S/N 173,274 discloses a method of making a terminal approach to a destination by sending to the motor a series of short pulses of varying width. After sending an initial pulse to the motor having a predetermined -width (duration) , a short delay is observed; and then the position of the driven element is compared to its previous, or "target,” position. If the driven element * has not been moved past the target position, the motor pulse width in increased by a predetermined increment; and the longer pulse is used to
• variable duration pulse -width i.e. , a variable length delay after turning the motor on
• a multiple-axis control system ordinarily is implemented by "time-sharing" th central processor among the axes, which are independently approaching destination positions. If one axis were to appropriate the processor for an indefinite period, while another axis was making a high-speed approach, there -would b s substantial likelihood that the latter axis would overshoo its mark. Another conflict could occur if two or more axes ⁇ were in the final adaptive approach procedure, and if each were to add its pulse-on delay time to the others' pulse-off delay time.
• the pulse width for each axis could increase -without limit, without altering the "duty cycle," i.e. , the ratio of "on” time to the total on “on” time plus “off” time While the system of S/N 173,274 could be redesigned to permi time-sharing, the changes would be complex and costly.
• control * apparatus and method of the present invention has proved to be even more accurate than the prior systems, and to be more rapid in approaching the destination position, because (a) the pulsing zone provides "coarse” and “fine” adaptive responses based on remaining distance and (b) "slipping back" of the driven element is more effectively countered.
• Another benefit of the present system is greater adaptability to the frictional differences which may be encountered in different positions of the driven element. In other words, there is greater certainty that motion will occur at some point during each pulsing cycle.
• the present invention is a synchronous pulsing system, which varies the energy exerted during a predetermined motor "on” period.
• the motor "on” portion of each cycle is a predetermined “window,” or period, of time; and the motor “off” portion of each cycle is also a predetermined period of time, the two periods having a ratio which depends on the type of machine being operated.
• a "nudging” technique is used during final approach to move the driven element by providing initially a lesser amount of energy during the period of one "on” window, and gradually increasing the amount of energy during each successive "on” window, until forward movement is detected. Then the cycle begins again with the lesser amount of energy (which has a value designed to ensure against overshoot) .
• the duration of the actuation period ("on" window) is constant, but the driving energy is varied within the "on” window.
• This variation preferably is accomplished by varying the duty cycle of the motor.
• the preferred way of causing the duty cycle variation is to provide motor driving pulses whose width is varied to alter the ratio of "on” and "off” ti e within the "on” window.
• the motor driving pulses within the window are started at a low duty cycle, and their duty cycle is increased incrementally from "on" window to "on” window until forward motion is detected, after which the sequence is repeated.
• Figure 1 is a diagrammatic showing of a control system for a multi-axis machine
• Figure 2 is a diagrammatic showing of the control circuitry for one axis of the system of Figure 1 ;
• Figure 3 is a flow, or logic, diagram which summarizes the operation of the final stage of the positioning system under the control of the microprocessor; and Figure 4 is a pulse diagram showing the sequence and duration of driving pulses in the system.
• FIG. 1 shows the general system, which preferably is microcomputer controlled.
• a central processor (CPU) 12 is programmed to provide the desired sequencing of events and is in electronic communication, via bus 14 (which includes data, address and control signals), with a read-only (or program) memory 16, a read/write memory 18, an interface/control 20 for the first axis, and an interface/control 22 for the second axis.
• the CPU 12 also is connected by a bus 24 to a keyboard 26 and to displays 28.
• the read/write memory stores program variables and sequences of positions; and the keyboard is used for entry of destination positions.
• the fundamental frequency utilized by the clock of CPU 12 may be established by a crystal 30.
• Each axis of the system has a variable duty-cycle circuit -which controls motor speed, and two signal inputs which control direction and dynamic braking.
• the two signal inputs from the CPU 12 to the first axis are leads Al and Bl ; and the two signal inputs from the CPU 12 to the second axis are leads A2 and B2.
• the signal inputs from the CPU on these lines establish the ratio of the on window period to the off period during the final approach phase.
• Figure 2 provides a diagram of the control circuit for each axis of the system.
• the end result is controlled motion and precise destination-positioning of a driven element symbolically shown at 32, which may represent any of numerous elements which require. precise positioning.
• a driven element symbolically shown at 32, which may represent any of numerous elements which require. precise positioning.
• the intended uses of the present invention are positioning of the lens and film planes of cameras, of the chases of step-and- repeat machines, or of the tables of drilling machines or milling machines. While the present invention is particularly useful in multi-axis systems, it is also applicable to single-axis systems. Although most uses will favor final destination approach, or "settling in,” from one direction only; it is entirely practicable to provide final approach from both directions, as is desirable in step-and- repeat machines.
• a lead screw 34 which may be rotated by an electric motor 36 operatively connected to the lead screw by a gear belt 38.
• an electric motor and specifically a permanent-magnet DC motor, is preferred.
• the position of the driven element may be conveniently sensed for feedback to the control electronics by a shaft encoder 40 (preferably an incremental encoder) , which may be operatively connected to the lead screw 34 by a gear belt 42.
• the position signals from the shaft encoder 40 are conveyed by lines 44 and 46 to a position decoding circuit 48, which accumulates a count representing the position of the driven element 32, and which is in communication with ' CPU 12 via bus 14.
• the conceptual basis of the control portion of each axis of the system is a combination of (a) a predetermined time ratio of on time to off time -wit (b) means for varying the energy during the on time. This energy variation is used in the final stage of the settling- in motion to provide a gradually increasing "nudging" effort which begins its cycle from lower to higher energy after eac detected forward motion.
• the variation of energy within the period, or "window,” of on time may be accomplished in various ways.
• the preferred approach, which is disclosed in this application, is variation of the duty cycle by varying the pulse widths o on and off pulses during the constant periods of the successive on windows. (Because the on window time remains constant, increasing the on pulse -width is accompanied by a decrease of the off pulse width, and vice versa. ) Instead o duty cycle variation to vary the effective energy during the on -window period, the voltage level could be varied from one on -window to the next. But such a system would be more complicated and more difficult to control precisely.
• the control portion of the system includes four control lines from the CPU 12. In addition to the A and B lines, previously mentioned, a reset line 50 and a clock line 52 ar required.
• the reset line 50 is connected to a counter and t a flip-flop 54.
• the counter is an 8-bit counter provided by combining two 4-bit counters 56 and 58, each communicating with CPU 12 via bus 14.
• the reset line 50 is connected to the preset enable inputs of each of the counters 56 and 58 , and to the reset input of flip-flop 54.
• the clock line 52 is connected to the clock inputs of each of the counters 56 and 58.
• the carry in of counter 56 is grounded, and the carry out of counter 56 is connected to the carry in of counter 58.
• the direction control lines A and B are connected to AND gates 60 and 52, respectively.
• the carry out of the 8-bit counter is an input via line 64 to an OR gate 66, whose output is connected by line 68 to the clock input of flip- flop 54.
• the Q output of flip-flop 54 is tied to its Data (D) input.
• the Q output of flip-flop 54 is an input via line 70 to both AND gates 60 and 62, and also to OR gate 56.
• the two AND gates 60 and 62 provide controlling signals to the motor 36 through the intermediary of a suitable motor drive circuit, which operates according to the logic of this truth table:
• motor 36 will run in the forward direction if 24V is supplied to lead 72 via Darlington 74, and ground is supplied to lead 76 via Darlington 78.
• the motor will run in reverse when lead 76 receives 24V from Darlington 80 and lead 72 is grounded via Darlington 82.
• Dynamic braking is accomplished if both leads are shorted together to ground via Darlingtons 78 and 82. If all Darlingtons are off, the motor is in an off state.
• Transistors 84 through 102 provide the logical interfacing between the AND gates 60 and 62 and the
• Transistor 90 also turns on transistor 92.
• the output of AND gate 62 is 0, so that transistor 94 is off, causing transistors 96 and 98 to be on, and transistors 100, 102 and Darlington 82 to be off. Since both transistors 92 and 98 are on, Darlington 74 is turned on, supplying 24VDC to motor lead 72, thereby causing motor 36 to move in the forward direction.
• the following is a table of transistor states as a function of the states of the A and B lines.
• circuitry may be used in controlling the motion of motor 36, but the disclosed circuitry is particularly simple, cost-ef ective and rugged for fractional horsepower motors operating in the region of 24 volts.
• the circuitry which determines the amount of energy applied during the period of the on window also may be selected from a number of options.
• the arrangement disclosed is considered relatively cost-effective.
• the counters 56 and 58 are CMOS 4-bit presettable binary counters
• the flip-flop 54 is a D-type flip-flop.
• the purpose of the circuit is to provide a pulse train whose duty cycle is proportional to an 8-bit value provided by the CPU 12.
• the counter is ' run in its "up" mode so that, upon counting up to its full count of 255, it generates a carry- out pulse on its CO pin and starts counting up again from zero. In the configuration shown in the drawing, the count increments at the clock frequency.
• the Q output of flip-flop 54 is tied to the D input so that at each clock transition the Q output changes state.
• the gate 66 ORs together the CO output of the counter and the Q output of the flip-flop so that, once the flip-flop has set its Q output to 1, no further transition can occur on its CLK input until after a reset.
• the clock signal is generated by the computer for various purposes; it is 400 kHz as it leaves the CPU, and is divided down to give the lower frequency.
• the reset signal is generated once every 256 clock cycles (at half the display refresh rate); it resets the flip-flop and presets the counter to the 8-bit value present on the data bus. The computer sees this as "writing" the value to the counter, as if it were a location in memory.
• the A and B signals provide direction and on/off logic.
• a 1 When a signal is present on A or B and while a 1 is present at the Q output of the flip-flop, a 1 -will be present at the output of one of the two AND gates, and current will flow through the motor.
• the CPU 12 In establishing the overall relation of on window time to off time of the motor, i.e. , the percentage of total time during which the on window permits motor actuating energy to be exerted, the CPU 12 is controlling the true or false signal on either line A or line B to the motor control circuit. If the axis under control is moving in the forward direction, the line on 'which a true, or positive, signal appears during the on window period is line A.
• the motor is off for 128 clock periods, then on for 128 clock periods — a duty cycle of 50%.
• the motor is off for 192, on for 64, a 25% duty cycle.
• the electronic position control system of Figure 2 is, as previously stated, for one axis only. Each axis requires it own control system, i.e. , an additional motor, lead screw, encoder, position decoding circuit, reset line, motor drive circuit, and direction control lines.
• the clock signal may be common to all axes. If a multi-axis system is used, the CPU 12 coordinates actuation of the axes, and prevents any conflicting demands from occurring.
• the control system described above permits a high degree of flexibility in operating motor 36 to move the driven .element 32 to its desired, or destination, position. In general, it has proved desirable to initially cause rapid, continuous movement of the motor and driven element until the latter reaches a position within a certain distance from its destination.
• the driven element If it is then far enough from the destination for the fast speed phase to be used 'without likelihood of an overshoot, the driven element is moved forward at high speed until that condition is no longer true, whereupon it is braked to a very low speed.
• a slower approach speed can easily be provided as an intermediate stage of continuous motion; but this stage has not proved to be necessary in test operations.
• the driven element is next driven at slower speed, using the fractional distance approach system of Application S/N 62,416, until the terminal neighborhood is reached.
• the terminal neighborhood is a distance of about .004 inch.
• the position of the driven element is first measured and stored in memory as a target position. If the distance to the .final destination is greater than 0.001 inch, an initial duty cycle of 60% is chosen. If the distance is less than 0.001 inch, the initial duty cycle is 40%, which under most circumstances is not quite sufficient to move the driven element in the prototype device.
• the motor is turned on in the forward direction for five milliseconds, then turned off. For the rest of a fifty millisecond period, the central processor attends to other tasks, including control of another axis (or axes). Then it returns to the pulsing procedure for the axis in question, and checks the current value of the position of the driven element.
• this position is forward of the target position, which is the last attained position, the initial value is again used for the next pulse, and the current position becomes the target. If no motion is sensed, the duty cycle is increased by 6.25%, whereas, if the driven element has slipped backwards, an increase of 12.5% is used. This cycle is repeated until forward motion occurs, increasing the energy in increments b increasing the duty cycle during the five millisecond "on" period. If the duty cycle has reached 100% without forward motion, a slipping back may have occurred, and a new target (the attained position) will be selected.
• the resolution of the position-sensing circuit is 0.0000625 inch.
• the computer has been able to position with no detectible error, or overshoot, on at least half of all trials .
• a clock frequency of 100 kHz and a reset frequency of 390 Hz is in use.
• the motors hum a the reset frequency when the duty cycle is below 90%.
• the initial choice was 1562 Hz, but this -was perceived as an unpleasant whine.
• a new frequency was therefore chosen to provide a more acceptable sound.
• Figure 3 discloses a logic flow chart used in controlling the experimental apparatus. As shown startin with- the entry symbol, the initial logic step in each axis- control sequence is an input/output block 110, which represents a subroutine from -which control is returned to th illustrated logic system when an absence of lead screw motio is indicated.
• control is relinquishe from the illustrated logic system to the time-sharing period utilized by the other axes. .'hen control is returned to the illustrated axis, the present position of the driven element is checked.
• decision block 114 it is determined whether the driven element has reached its destination position. If the answer is "yes,” the normal path is to "exit.” However, in some situations , it may be pre erred to follow the dashed line back to the top of the chart, which would have the effect of causing the apparatus to maintain its position in spite of any external forces tending to alter its position.
• the duty cycle sequencing will start at 60%, as shown by process block 126. If the position is in the inner zone, the duty cycle sequencing v;ill start at 40%, as shown by process block 128.
• the 40% duty cycle value 'which is both an approximate value and a somewhat arbitrary one, has been chosen to provide an amount of effort slightly below that normally required to cause forward movement of the driven e1ement.
• the duty cycle is set at the initial value (either 40% or 60% depending on position) , and the motor is turned on at clock 132.
• a fixed delay provides a period, or "window,” of on time. This delay is shown as approximately 5 ms , or 10% of the total on-plus-off period of approximately 50 ms , which is the period from one "triggering" to the next.
• the percentage of on time to total time in each on/off period is chosen to meet the requirements of the apparatus which is being positioned. But, a predetermined ratio is initially- established, and it remains the same.
• the motor turns off, as shown by block 136.
• "attention" of the CPU 12 is transferred to the other axis, or axes, of the apparatus.
• the current position of the driven element (on the axis shown in the figure) is repetitively determined.
• the logic flow moves to decision block 142, -which determines whether or not the driven element has moved past the target (which, as stated, preferably was set at the actual position reached by the driven element).
• the loop is repeated, beginning with a new target setting at block 120 (the new target being the new position attained by the driven element, forward of its previous position) . If decision block 142 indicates' that the driven element has not moved forward from the target position, a determination is made at decision block 144 whether it has slipped back from the target position.
• decision block 142 indicates' that the driven element has not moved forward from the target position
• decision block 144 whether it has slipped back from the target position.
• the unusual flexibility of the present control system comes into play because different increments of duty cycle increase are used, depending on the answer to the question whether the position has slipped back. If the driven element has slipped back, the chart at block 146 shews a 12.5% increase of the duty cycle, 'which would bring it to 52.5% if it had been 40% during the period of- the preceding on window 7 .
• the chart at block 148 shows a 6.25% increase of the duty cycle, -which would bring it to 46.25% if it had been 40% during the period of the preceding on window, to 52.5% if it had been 46.25% during the period of the preceding on window, and so on.
• the next duty cycle percentage is determined by adding the most recent duty cycle value to the increment-of-increase value. The percentages of duty cycle increase are chosen to suit the requirements of the apparatus, selecting binary values from those available in the 8-bit counter.
• the new, increased duty cycle is then returned through the loop to block 130, where the increased duty cycle is set, after which the motor is turned on again for the same on period, or "windo , " of 5 ms . 'when the total period of 50 ms has passed, a determination is made at block 142 whether the driven element has moved forward from the target position. If it still has not moved forward, another increment of duty cycle increase (either 6.25% or 12.5%) is added at block 150, and the motor triggering is repeated.
• Block 152 indicates that different strategies need to be adopted if the duty cycle reaches 100% without moving the driven element forward of the target position.
• One possibility is to reset the target at the then current position, as indicated by the solid line looping back to block 120.
• Another possibility is to get out of the pulsing loop, as indicated by the dashed line looping back to block 110.
• Figures 4A, 4B and 4C provide additional clarification by diagramming the motor driving pulses created by the pulse forming circuit. Each of the three pulse diagrams is scaled on a time, or duration, basis; but the three scales are widely different in magnitude.
• Figure A shows . the timing of signals received at AND gate 50. The clock pulses count to 256, as shown on the first line.
• the reset pulse at flip-flop 54 coincides with the 255th clock pulse, as shown on the second line.
• the pulses on line A (which create the on 'windows in the forward direction) and the pulses at Q of flip-flop 54 both appear a AND gate 60.
• the period uring which the signal on line A i positive depends on the on 'window duration, -which varies fro 100% during continuous operation of the motor to whatever value is selected as the desired on-to-off ratio during the adaptive pulsing mode used for final approach.
• the third line in Figure 4A shows an arbitrarily selected pulse length on line A; and the fourth line shows an example of the pulse width at the Q output of the flip-flop.
• the example chosen is a high duty cycle value, approximately 96%, because Q, in the illustration, remains positive from the 10th clock pulse to the 256th clock pulse.
• the duty cycle is varied by changing the number of the clock pulse cycle on which preset occurs, thereby changing the point at which Q output goes from negative to positive. If Q is caused to go from negative to positive at the 128th clock pulse, the duty cycle of the motor during each on window period (while that preset value continues) will be 50%.
• Figure 4B shows sample pulse trains during continuous operation of the motor. If maximum speed is desired, the driving pulses shown on the first line are used, providing approximately 98% of available motor energy. If a somewhat lower-speed, but continuous, motor energization is desired, a lower duty cycle may be used, such as the approximately 60% duty cycle shown on the second line.
• the millisecond scale of Figure 4B shows that the duty cycle is identical in each time segment of 2.56 ms.
• Figure 4C shows how the adaptive pulsing concept works in the final approach. Generally, it is preferred to have different duty cycle values in an outer zone and an inner zone, although this refinement is not necessary.
• the outer pulsing zone may cover the distance between .004 inch and .001 inch from destination.
• the inner, or near, pulsing zone may cover the final .001 inch.
• the adaptive pulsing entry at .004 inch has been used because it is convenient to use the faster approach mode shov.-n in Application S/N 62,416 to bring the driven element to this point, without any danger of overshoot.
• the first line of Figure 4C shows a 60% duty cycle used for adaptive pulsing in the outer zone.
• the on window period has been selected as 10% of the total on/off period, or cycle. This value can be changed, but it has been very successful in experimental apparatus. Obviously the on time percentage must be low enough not to interfere with the time-sharing requirements of other axes. As seen from the time scale in Figure 4C , the on window is 5.12 ms, and the total period, or cycle, is 51.2 ms , the motor being unable to receive electrical energy during 90% of the cycle.
• the first effort to move the driven element forward in the outer zone receives a 60% duty cycle. If forward motion does not occur, the next duty cycle, as shown, in 66.25%. I forward motion does occur as a result of this second "nudge," the duty cycle reverts back to 60%, as shown.
• a first duty cycle "nudge" of 40% is usually found to be desirable. This is preferably chosen as a value which will ordinarily not cause forward motion to occur. Until -such motion does occur, each successive on window will, as shown, receive a duty cycle -which is higher than the previous one by an increment of 5.25%. So the duty cycle will rise, in each successive on window, from 40% to 46.25% to 52.5%, and so on until forward movement occurs.

Landscapes

• Engineering & Computer Science (AREA)
• Human Computer Interaction (AREA)
• Manufacturing & Machinery (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Automation & Control Theory (AREA)
• Computer Hardware Design (AREA)
• Microelectronics & Electronic Packaging (AREA)
• Control Of Position Or Direction (AREA)

Priority Applications (1)

Applications Claiming Priority (2)

Publications (1)

Family

ID=23113752

Family Applications (1)

Country Status (5)

Citations (8)

Family Cites Families (2)

• 1982
  • 1982-07-28 EP EP19820902767 patent/EP0085098A4/en not_active Withdrawn
  • 1982-07-28 BR BR8207812A patent/BR8207812A/pt unknown
  • 1982-07-28 JP JP57502678A patent/JPS58501248A/ja active Pending
  • 1982-07-28 WO PCT/US1982/001050 patent/WO1983000571A1/en not_active Application Discontinuation
  • 1982-07-28 AU AU89007/82A patent/AU553773B2/en not_active Ceased

Patent Citations (8)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events