Electricity: motive power systems – Positional servo systems
Reexamination Certificate
2001-07-18
2004-02-03
Nappi, Robert E. (Department: 2837)
Electricity: motive power systems
Positional servo systems
C318S561000, C318S609000, C318S610000, C318S630000, C318S632000
Reexamination Certificate
active
06686716
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to motor control and more particularly, to motor control applications requiring rapid, periodic, point-to-point movement.
2. Description of the Related Art
Rapid repositioning of an object being moved is accomplished using a motorized system, such as a servo system, that typically is controlled using one of the many known closed-loop control algorithms. A motor and the object moved by the motor are commonly referred to as a “plant.” An example of such a plant is a rotating mirror that directs an image into a camera mounted in a weather satellite in which the camera photographs the earth. The motor periodically repositions the mirror to produce a series of photographs that together form an image of the earth. When variations in the plant are expected, adaptive control methods can be used to more accurately position the plant. When the movement is periodic, more specialized adaptive methods, including repetitive and learning control, are available.
Some modern periodic control systems require repositioning times that approach the minimum theoretically possible while also minimizing the peak power used. This requires a minimum-time, optimal solution that generally satisfies Pontryagin's minimum principle, described, for example, in
Optimal Control Theory
, Donald E. Kirk, Prentice-Hall Inc., pp. 227-228, 1970. This minimum-time, optimal solution requires an initial maximum acceleration followed by a maximum deceleration and is referred to in the art as bang-bang control. Bang-bang control is usually associated with an□open-loop control system. In theory bang-bang control moves a plant in a minimum amount of time. However, in practice overshoot and settling after the plant reaches its final position often occur. Variation from the desired position after the initial rapid movement from the bang-bang control usually has sinusoidal components that are referred to as residual vibration. This is a non-zero position error at the end of the bang-bang control operation, the amount of which oscillates as a damped sinusoid around the desired position as a result of the closed loop control moving the plant to eliminate the error. The residual vibration is due to disturbances, non-linlinear effects, resonance, time-related variations and other factors. Frequently, these factors cannot be accounted for, to the precision required, in a model of the control system.
Open-loop control techniques are common in the literature and have been shown to be effective in several areas of control. For example, open-loop control has been used with an input that is based on a plant model for a precision diamond turning application. In that application a hybrid (open and closed-loop) control law could be used. In another example, open-loop control has been used with a switching zone controller that causes motor torque to switch linearly from a maximum positive (acceleration) to a maximum negative (deceleration) value in a robotic application. In yet another example, a “modified bang-bang” controller uses open-loop control for rapid movement and switches to closed-loop control near the desired position to control a (military) tank gun-loader.
Conventional open-loop and open-loop/closed-loop systems, such as those discussed above, do not suffice in applications requiring rapid, periodic movement because they do not simultaneously minimize repositioning time, the peak power used and any residual vibrations. More particularly, they do not include algorithmic, repetitive tuning of open-loop input pulse amplitudes to the motor based upon position and velocity when the loop is closed to minimize residual vibrations resulting from closed-loop control.
SUMMARY OF THE INVENTION
The invention can solve the concurrent problems with conventional control systems of minimizing repositioning time, minimizing peak power used, and minimizing residual vibrations for a periodic system requiring rapid movement. Those problems can be solved by using an open-loop command such as a bang-bang control signal that is input to move the plant, then using closed-loop control to fine tune the movement to the desired position, and using a form of repetitive/learning control to adjust the bang-bang control signal parameters for subsequent periodic movements. The open-loop bang-bang control signal is determined to minimize saturation non-linearities in the electronics (e.g., due to peak voltages/currents) and in the motor (e.g., due to acceleration limits). The initial positive and negative amplitudes of the open-loop bang-bang input signal are calculated based on a model of the plant. The maximum and minimum amplitudes of the bang-bang input signal are then adjusted periodically using the repetitive/learning control.
There are advantages to using open-loop control switched to closed-loop control in applications requiring rapid, periodic movement, including the following.
1. Open-loop control can provide faster response than closed-loop control because the command signal input is coupled directly into the plant.
2. Open-loop control can be more efficient than closed-loop control because there are fewer components and response can be closer to time-optimal for point-to-point movement.
3. Linearity (no saturation) within the electronics and the plant (motor) can be maintained by limiting the highest frequency component and the maximum amplitude of the open-loop input command.
4. Closed-loop control reduces the effect of external disturbances and accurately holds the final position after the open-loop movement.
REFERENCES:
patent: 4439716 (1984-03-01), Minnich
patent: 4674029 (1987-06-01), Maudal
patent: 4742444 (1988-05-01), Gose et al.
patent: 5677609 (1997-10-01), Khan et al.
patent: 6013996 (2000-01-01), Howard et al.
patent: 6107771 (2000-08-01), Maeda
patent: 6326758 (2001-12-01), Discenzo
patent: 0461627 (1991-12-01), None
patent: 0 461 627 (1991-12-01), None
Donald E. Kirk,Optimal Control Theory An Introduction, 1970, pp. 227-228.
Hsin, et al. , “Experiments Bridging Learning and Repetitive Control”, Advances in the Astronomical Sciences vol. 95, part 2, 1997, pp. 671-690.
Katsuhiko Ogata,Discrete-Time Control Systems, Prentice-Hall, Inc., 1987, p. 626.
Dec. 1989, H. Ozisik and R.F. Keltie, “Development and implementation of a high speed open loop control technique for micropositioning of mechanical structures”, Precision Engineering, vol. 11, No. 4, Oct. 1989, pp. 225-230.
H. Ozisik and R.F. Keltie, “Implementation of an open-loop control technique for high-speed micropositioing in a single-point diamond turning process”, Precision Engineering, Vol 13, No. 2, Apr. 1991, pp. 83-94.
Zhishi, Xia and Sheldon S. L. Chang, “A Simple Robust Nonlinear Robotic Controller”, Recent Trends in Robotics, El Sevier Science Publishing Co., Inc., 1986, pp. 317-328.
Ronald L. Racicot, “Increasing Mission Reliability By Using Open-Loop Control” IEEE Transactions on Reliability, vol. 42, No. 3, Sep. 1993, pp. 384-392.
Ronald L. Racicot, “Limiting Servo Motor Torque Gradients With Near Minimum Time Repositioning”, IEEE Transactions on Control Systems Technology, vol. 1, No. 4, Dec. 1993, pp. 284-289.
Hal Broberg, Paul Lin & Joe Predina, “Periodic, Multi-Step Tracking Control for Weather Satellite Scanning Mirror”, J. of Dynamic Systems, Measurement and Control, vol. 123, Mar. 2001, pp. 85-92.
Kok Kia Chew and Masayoshi Tomizuka, “Digital Control of Repetitive Errors in Disk Drive Systems”, IEEE Control Systems Magazine, Jan. 1990, pp. 16-20.
S. Arimoto, S. Kawamura, F. Miyazaki, and S. Tamaki, “Learning Control Theory for Dynamical Systems”, IEEE Proc. Of 24thConf. On Decision and Control, Dec. 1985, pp. 1375-1380.
Predina, J., “Proposal for Infrared Temperature Sounder”, vol. 1, Technical, Nov. 29, 1991, p. 1-24 through 1-30.
Toshiharu Sugie and Toshiro Ono, “An Iterative Learning Control Law for Dynamical Systems” Automatica, vol. 27, No. 4, 1991, pp. 729-732.
Broberg, H.L., 1993, “Theory and Optimization: The Coherent Error Integrator, East-West Mirror Servomechani
Broberg Harold L.
Predina Joe Paul
Edell Shapiro & Finnan LLC
ITT Manufacturing Enterprises Inc.
Nappi Robert E.
Smith Tyrone
LandOfFree
Tuned open-loop switched to closed-loop method for rapid... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Tuned open-loop switched to closed-loop method for rapid..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Tuned open-loop switched to closed-loop method for rapid... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3339713