Teager-based method and system for predicting limit cycle oscillations and control method and system utilizing same

• US 6,004,017 A
• Filed: 04/30/1997
• Issued: 12/21/1999
• Est. Priority Date: 11/07/1996
• Status: Expired due to Fees

• Current Assignee: Madhavan, Poovanpilli G.
• Original Assignee: Madhavan, Poovanpilli G.
• Inventors: Madhavan, Poovanpilli G.
• Primary Examiner(s): Grant, William
• Assistant Examiner(s): Calcano, Ivan
• Application Number:
  US08/848,316
• Time in Patent Office: 965 Days
• Field of Search: 364/270.3, 364/157, 364/162, 364/271, 364/724.1, 364/176, 364/152, 364/474.17, 364/148.09, 702/56
• US Class Current: 700/71
• CPC Class Codes: G05B 5/01 electric

Abstract

A data-driven method and system are provided for predicting limit cycle oscillations such as chatter to obtain a classifier signal which, in turn, may be utilized by a control method and system. The method and system utilize newly-developed algorithms based on Teager algorithm which are applied to "vibration" data to predict when chatter phenomenon may emerge during an ongoing machining process. Using real data from a turning operation on a lathe, the onset of chatter can be predicted so that sufficient time is available to deploy an automatic control strategy such as reduction or random variation of spindle speed to quench the "chatter". The method and system preferably use the generalized Teager measure.

Claims

1. A method for predicting limit cycle oscillations, the method comprising the steps of:

generating a signal based on oscillatory behavior;

statistically processing the signal to obtain a generalized Teager measure;

monitoring instantaneous changes in the generalized Teager measure; and

generating a classifier signal when the instantaneous changes are more than a predetermined threshold value for a predetermined time period whereby the classifier signal is a prediction of the limit cycle oscillations.

2. The method as claimed in claim 1 wherein the limit cycle oscillations are undesirable vibrations.

3. The method of claim 2 wherein the undesirable vibrations are machine tool chatter.

4. The method of claim 1 wherein the step of monitoring includes the step of estimating the generalized Teager measure of the signal and wherein the instantaneous changes are based on the time-frequency distribution.

5. The method of claim 4 wherein the time-frequency distribution is a spectrogram.

6. The method of claim 1 wherein the electrical signal is a vibration signal.

7. The method of claim 1 wherein the electrical signal is a sampled vibration signal.

8. A method for controlling limit cycle oscillations of a machine tool having a spindle, the method comprising the steps of:

generating an electrical signal based on oscillatory behavior of the object;

statistically processing the electrical signal to obtain a function;

monitoring instantaneous changes in the function;

generating a classifier signal when the instantaneous changes are more than a predetermined threshold value for a predetermined time period whereby the classifier signal is a prediction of the limit cycle oscillations; and

generating a spindle speed control signal adapted to control the speed of the spindle of the machine tool based on the classifier signal.

9. A system for predicting limit cycle oscillations, the system comprising:

a detector for generating a signal based on oscillatory behavior;

a signal processor for statistically processing the signal to obtain a generalized Teager measure;

a monitor for monitoring instantaneous changes in the generalized Teager measure; and

a signal generator for generating a classifier signal when the instantaneous changes are more than a predetermined threshold value for a predetermined time period whereby the classifier signal is a prediction of the limit cycle oscillations.

10. The system of claim 9 wherein the limit cycle oscillations are undesirable vibrations.

11. The system of claim 10 wherein the undesirable vibrations are machine tool chatter.

12. The system of claim 9 wherein the monitor includes an estimator for estimating the generalized Teager measure of the signal and wherein the instantaneous changes are based on the time-frequency distribution.

13. The system of claim 12 wherein the time-frequency distribution is a spectrogram.

14. The system of claim 9 wherein the electrical signal is a vibration signal.

15. The system of claim 9 wherein the electrical signal is a sampled vibration signal.

16. A system for controlling limit cycle oscillations of a machine tool having a spindle, the system comprising:

a detector for generating an electrical signal based on oscillatory behavior of the object;

a signal processor for statistically processing the electrical signal to obtain a function;

a monitor for monitoring instantaneous changes in the function;

a first signal generator means for generating a classifier signal when the instantaneous changes are more than a predetermined threshold value for a predetermined time period whereby the classifier signal is a prediction of the limit cycle oscillations; and

a second signal generator for generating a spindle speed control signal adapted to control the spindle of the machine tool based on the classifier signal.