Vibration monitor and monitoring method

• US 5,847,658 A
• Filed: 08/13/1996
• Issued: 12/08/1998
• Est. Priority Date: 08/15/1995
• Status: Expired due to Fees

• Current Assignee: Omron Corporation
• Original Assignee: Omron Corporation
• Inventors: Ogawa, Yukio, Osada, Atsushi, Morita, Yoshiyuki, Fujimoto, Naoki, Irie, Atsushi, Furusawa, Koichi, Umeda, Hidenobu, Okura, Kiyotoshi
• Primary Examiner(s): TONG, NINA C
• Application Number:
  US08/689,787
• Time in Patent Office: 847 Days
• Field of Search: 340/683, 340/679, 340/635, 340/680, 364/550, 364/551.01, 364/551.02, 364/508, 364/474.19, 364/494, 73/570, 73/572, 73/579, 73/593, 73/602, 73/660, 73/659, 73/658
• US Class Current: 340/683
• CPC Class Codes: G01H 1/00 Measuring characteristics of vibrations in solids by using direct conduction to the detector G01H9/00, G01H11/00 take precedence

Abstract

The object of this invention is to produce a monitor which can correctly distinguish normal from abnormal vibration using automatically selected monitoring features and algorithms. The monitor can be used with a variety of devices with different vibration characteristics. The monitor according to this invention has a sensor that detects the vibration of an object. The monitor is operated in learning mode when the object is vibrating normally. The waveform of the vibration detected by this sensor is sampled over an appropriate period and analyzed. The data for each sampling feature are subjected to statistical processing, and the data which evince slight change are extracted as monitoring features. A monitoring algorithm is automatically made for these monitoring features. The algorithm selected in learning mode are used to evaluate the vibration of an object when the monitor is in operating mode. The operator does not need to set up the algorithm, and can monitor the object with a more reliable algorithm than the conventional one selected by fixed monitoring features.

Claims

1. A vibration monitor to monitor an object, comprising:

a vibration detection unit to detect vibration of said object with a vibration sensor and output a detection waveform of said vibration;

a computer unit to process said detection waveform and determine if said vibration of said object is normal or abnormal;

wherein said computer unit operates in a learning mode to automatically generate a monitoring algorithm based on said detection waveform in a normal condition, said monitoring algorithm is defined by a predetermined number of monitoring features selected from a predetermined number of sampling features in said detection waveform, and said monitoring features evince statistically slight change; and

wherein said computer unit operates in a monitoring mode to monitor said detection waveform according to said monitoring algorithm.

2. A vibration monitor according to claim 1, wherein said sampling features are maximum value, minimum value, peak-to-peak value, number of times a high-level threshold value is exceeded, number of times a low-level threshold value is exceeded, time period during which said high-level threshold value is exceeded, time period during which said low-level threshold value is exceeded, and number of times a local minimum is generated.

3. A vibration monitor according to claim 1, wherein said computer unit enters in said monitoring mode automatically after said learning mode is completed.

4. A vibration monitor according to claim 1, further comprising a manual input unit to place said computer unit in said monitoring mode after said learning mode.

5. A vibration monitor according to claim 1, further comprising an output unit to indicate that said computer unit is in said learning mode.

6. A vibration monitor according to claim 5, wherein said output unit indicates that said learning mode is completed.

7. A vibration monitor according to claim 5, wherein said computer unit enters said monitoring mode automatically after said learning mode is completed, and said output unit indicates that said computer unit is in said monitoring mode.

8. A vibration monitor according to claim 1, further comprising a manual input unit to manually select a number of said monitoring features from a plurality of said sampling features stored in a memory unit.

9. A vibration monitor according to claim 1, further comprising a manual input unit to manually define said monitoring algorithm.

10. A vibration monitor according to claim 9, wherein said monitoring algorithm is defined by a prior monitoring algorithm manually selected from a plurality of monitoring algorithms previously generated in a plurality of learning modes and stored in a memory unit.

11. A vibration monitor according to claim 10, further comprising an output unit to display output data, wherein said output data indicates a discrepancy between a current monitoring algorithm and a previous monitoring algorithm.

12. A vibration monitor according to claim 11, wherein said discrepancy is a value of monitoring features calculated by said current and previous monitoring algorithms.

13. A vibration monitor according to claim 11, wherein said output unit is a digital indicator to indicate said discrepancy.

14. A vibration monitor according to claim 11, wherein said output unit is a lamp indicator to indicate said discrepancy when said discrepancy is outside a predetermined normal range.

15. A vibration monitor according to claim 9, wherein said monitoring algorithm is defined by a last monitoring algorithm generated in a last learning mode and stored in a memory unit.

16. A vibration monitor according to claim 9, wherein said monitoring algorithm is manually defined by resetting a threshold applying to said detection waveform, and replacing said monitoring feature with a sampling feature obtained after said threshold was reset.

17. A vibration monitor according to claim 16, wherein a plurality of said monitoring features are changed by resetting at least two of said thresholds.

18. A vibration monitor according to claim 1, wherein said vibration detection unit detects a plurality of detection waveforms, and said computer unit generates said monitoring algorithm based on said plurality of detection waveforms.

19. A vibration monitor according to claim 1, further comprising an output unit to display output data, wherein said output data indicates a discrepancy of said monitoring feature values in said learning mode and said monitoring mode calculated by said monitoring algorithm.

20. A vibration monitor according to claim 19, wherein said output unit is a digital indicator to indicate said discrepancy.

21. A vibration monitor according to claim 19, wherein said output unit is a lamp indicator to indicate said discrepancy when said discrepancy is outside a predetermined normal range.

22. A vibration monitor according to claim 1, further comprising an output unit to indicate if said computer unit already contains said monitoring algorithm.

23. A vibration monitor according to claim 1, further comprising an output unit to indicate said monitoring features defining said monitoring algorithm.

24. A vibration monitor according to claim 1, further comprising a filter to separate said detection waveform into a number of frequency bands which are separately processed by said computer unit.

25. A vibration monitor according to claim 1, wherein said computer unit has a gain adjuster to adjust an output level of said detection waveform during said learning mode.

26. A vibration monitor according to claim 1, further comprising a manual input unit to input a parameter to modify said monitoring algorithm.

27. A vibration monitor according to claim 1, wherein said computer unit automatically generates said monitoring algorithm using fuzzy inference.

28. A vibration monitoring method to monitor an object, comprising the steps of:

setting a vibration monitor in a learning mode to learn a normal vibration of an object in normal condition and automatically generating a monitoring algorithm;

setting said vibration monitor in a monitoring mode to monitor said object according said monitoring algorithm;

wherein said monitoring algorithm is defined by a predetermined number of monitoring features selected from a predetermined number of sampling features in a detection waveform, and said monitoring features evince statistically slight change.