Methods and apparatus for data smoothing

• US 6,782,297 B2
• Filed: 05/31/2001
• Issued: 08/24/2004
• Est. Priority Date: 05/24/2001
• Status: Expired due to Fees

• Current Assignee: Eric Paul Tabor
• Original Assignee: Eric Paul Tabor
• Inventors: Tabor, Eric Paul
• Primary Examiner(s): PATEL, RAMESH B
• Application Number:
  US09/872,195
• Publication Number:
  US 20020183868A1
• Time in Patent Office: 1,181 Days
• Field of Search: 70/50, 70/52, 70/55, 708/300
• US Class Current: 700/55
• CPC Class Codes: G05B 23/0221 Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods

Abstract

A method and apparatus for smoothing data comprises adjusting data according to a first smoothing technique and selectively adjusting the data according to a second smoothing technique. The method and apparatus applies a first smoothing technique to a selected datum to adjust a value. The method and apparatus also compares preceding adjusted data to preceding raw data to generate a comparison result and may apply a second smoothing technique to the selected datum according to whether the comparison result meets a threshold. In addition, the method and apparatus of various embodiments may calculate a predicted value of the selected datum and apply another smoothing technique to the selected datum according to whether the predicted value meets a second threshold.

Claims

1. A method of smoothing data, comprising:

applying a first smoothing technique to a selected datum to adjust a value of the selected datum;

comparing a plurality of preceding adjusted data to a plurality of preceding raw data to generate a comparison result;

applying a second smoothing technique to the selected datum to adjust the value of the selected datum according to whether the comparison result meets a first threshold;

calculating a predicted value of the selected datum; and

applying a third smoothing technique to the selected datum to adjust the value of the selected datum according to whether the predicted value meets a second threshold.

2. A method according to claim 1, further comprising initially determining whether smoothing is appropriate for the selected datum.

3. A method according to claim 2, wherein determining whether smoothing is appropriate comprises at least one of:

comparing the value of the selected datum to a threshold;

determining a number of data received; and

determining a deviation of a plurality of data from a value.

4. A method according to claim 2, wherein determining whether smoothing is appropriate comprises:

determining a difference between the value of the selected datum and an adjusted value of a preceding datum; and

comparing the difference to a threshold.

5. A method according to claim 1, wherein at least one of the first smoothing technique and the second smoothing technique comprises at least one of:

random smoothing;

random walk smoothing;

moving average smoothing;

simple exponential smoothing;

linear exponential smoothing;

seasonal exponential smoothing; and

exponential weighted moving average smoothing.

6. A method according to claim 1, wherein at least one of the first smoothing technique and the second smoothing technique comprises adjusting the value of the selected datum substantially in accordance with the equation:

7. A method according to claim 1, wherein:

the first smoothing technique comprises adjusting the value of the selected datum substantially in accordance with the equation;

8. A method according to claim 1, wherein comparing the plurality of preceding adjusted data to the plurality of preceding raw data comprises:

comparing an adjusted value of each of the plurality of preceding adjusted data to a corresponding value of each of the plurality of preceding raw data; and

generating the comparison result according to whether substantially all of the adjusted values of the plurality of preceding adjusted data share a common relationship with respect to the corresponding values of the plurality of preceding raw data.

9. A method according to claim 1, wherein calculating the predicted value of the selected datum comprises calculating a slope according to a plurality of preceding data.

10. A method according to claim 9, wherein the slope is calculated using at least one of:

a least squares fit through technique;

a line regression technique;

or

an N-points centered technique.

11. A method according to claim 1, wherein the second threshold corresponds to a difference between the adjusted value of the selected datum and the predicted value of the selected datum to a range.

12. A method according to claim 1, wherein the third smoothing technique comprises adjusting the value of the selected datum substantially in accordance with the equation:

13. A method according to claim 1, wherein at least one of the second smoothing technique and the third smoothing technique is configured to improve the tracking of the adjusted value of the selected datum to a selected raw data.

14. A method according to claim 1, wherein at least one of the second smoothing technique and the third smoothing technique is configured to enhance the smoothing of the data.

15. A method according to claim 1, wherein the data comprises at least one of automatic test equipment test data and financial data.

16. A computer readable program for smoothing data, the computer readable program configured to cause a computer to:

apply a first smoothing technique to a selected datum to adjust a value of the selected datum;

compare a plurality of preceding adjusted data to a plurality of preceding raw data to generate a comparison result;

apply a second smoothing technique to the selected datum to adjust the value of the selected datum according to whether the comparison result meets a first threshold;

calculate a predicted value of the selected datum; and

apply a third smoothing technique to the selected datum to adjust the value of the selected datum according to whether the predicted value meets a second threshold.

17. A computer readable program according to claim 16, further comprising initially determining whether smoothing is appropriate for the selected datum.

18. A computer readable program according to claim 17, wherein determining whether smoothing is appropriate comprises at least one of:

comparing the value of the selected datum to a threshold;

determining a number of data received; and

determining a deviation of a plurality of data from a value.

19. A computer readable program according to claim 17, wherein determining whether smoothing is appropriate comprises:

determining a difference between the value of the selected datum and an adjusted value of a preceding datum; and

comparing the difference to a threshold.

20. A computer readable program according to claim 16, wherein at least one of the first smoothing technique and the second smoothing technique comprises at least one of:

random smoothing;

random walk smoothing;

moving average smoothing;

simple exponential smoothing;

linear exponential smoothing;

seasonal exponential smoothing; and

exponential weighted moving average smoothing.

21. A computer readable program according to claim 16, wherein at least one of the first smoothing technique and the second smoothing technique comprises adjusting the value of the selected datum substantially in accordance with the equation:

22. A computer readable program according to claim 16, wherein:

the first smoothing technique comprises adjusting the value of the selected datum substantially in accordance with the equation;

23. A computer readable program according to claim 16, wherein comparing the plurality of preceding adjusted data to the plurality of preceding raw data comprises:

comparing an adjusted value of each of the plurality of preceding adjusted data to a corresponding value of each of the plurality of preceding raw data; and

generating the comparison result according to whether substantially all of the adjusted values of the plurality of preceding adjusted data share a common relationship with respect to the corresponding values of the plurality of preceding raw data.

24. A computer readable program according to claim 16, wherein calculating the predicted value of the selected datum comprises calculating a slope according to a plurality of preceding data.

25. A computer readable program according to claim 24, wherein the slope is calculated using at least one of:

a least squares fit through technique;

a line regression technique;

or

an N-points centered technique.

26. A computer readable program according to claim 16, wherein the second threshold corresponds to a difference between the adjusted value of the selected datum and the predicted value of the selected datum to a range.

27. A computer readable program according to claim 16, wherein the third smoothing technique comprises adjusting the value of the selected datum substantially in accordance with the equation:

28. A computer readable program according to claim 16, wherein at least one of the second smoothing technique and the third smoothing technique is configured to improve the tracking of the adjusted value of the selected datum to a selected raw data.

29. A computer readable program according to claim 16, wherein at least one of the second smoothing technique and the third smoothing technique is configured to enhance the smoothing of the data.

30. A computer readable program according to claim 16, wherein the data comprises at least one of automatic test equipment test data and financial data.

31. A method of smoothing data, comprising:

initially adjusting an initial value of a selected datum according to a first smoothing technique; and

supplementarily adjusting the value according to a second smoothing technique if at least one of the initial value and the initially adjusted value meets a threshold.

32. A method according to claim 31, further comprising adjusting at least one of the initial value, the initially adjusted value, and the supplementarily adjusted value of the selected datum according to a third smoothing technique if at least one of the initial value, the initially adjusted value, and the supplementarily adjusted value meets a second threshold.

33. A method according to claim 32, wherein the second threshold corresponds to a difference between the at least one of the initial value, the initially adjusted value, and the supplementarily adjusted value, and a predicted value of the at least one of the initial value, the initially adjusted value, and the supplementarily adjusted value.

34. A method according to claim 31, further comprising initially determining whether smoothing is appropriate for the selected datum.

35. A method according to claim 34, wherein determining whether smoothing is appropriate comprises at least one of:

comparing the initial value to a threshold;

determining a number of data received; and

determining a deviation of a plurality of data from a value.

36. A method according to claim 34, wherein determining whether smoothing is appropriate comprises:

determining a difference between the initial value of the selected datum and an adjusted value of a preceding selected datum; and

comparing the difference to a threshold.

37. A method according to claim 31, wherein at least one of the first smoothing technique and the second smoothing technique comprises at least one of:

random smoothing;

random walk smoothing;

moving average smoothing;

simple exponential smoothing;

linear exponential smoothing;

seasonal exponential smoothing; and

exponential weighted moving average smoothing.

38. A method according to claim 31, wherein the first smoothing technique comprises initially adjusting the initial value of the selected datum substantially in accordance with the equation:

39. A method according to claim 38, wherein the second smoothing technique comprises supplementarily adjusting the initially adjusted value of the selected datum substantially in accordance with the equation:

40. A method according to claim 31, wherein the data comprises at least one of automatic test equipment test data and financial data.