Active adaptive control system with spectral leak

• US 5,586,189 A
• Filed: 12/14/1993
• Issued: 12/17/1996
• Est. Priority Date: 12/14/1993
• Status: Expired due to Term

• Current Assignee: Digisonix Incorporated
• Original Assignee: Digisonix Incorporated
• Inventors: Allie, Mark C., Popovich, Steven R.
• Primary Examiner(s): Coles, Sr., Edward L.
• Assistant Examiner(s): Grant, II, Jerome
• Application Number:
  US08/166,698
• Time in Patent Office: 1,099 Days
• Field of Search: 381/71, 381/94
• US Class Current: 381/71.11
• CPC Class Codes: H03H 2021/007 Computation saving measures; Accelerating measures
• CPC Class Codes: H03H 2021/0074 Reduction of the update frequency
• CPC Class Codes: H03H 21/0012 Digital adaptive filters

Abstract

An act control system introduces a control signal put transducer (14) to combine with the system input signal (6) and yield a system output signal (8). An adaptive filter model (40) has an output (46) supplying a correction signal to the output transducer to introduce the control signal. The model has a first error input (44, 232, 248) driven by an error signal from an error transducer (16) to drive the correction signal toward a value matching the system input signal to modify and/or cancel same. The model has a second error input (202, 224, 240, 268, 280) selectively driven by a spectral leak signal (212) to drive the correction signal away from such matching value by driving the correction signal towards zero, to selectively constrain performance of the model in response to a given condition of a given parameter.

Claims

1. An active adaptive control method comprising introducing a control signal from an output transducer to combine with a system input signal and yield a system output signal, sensing said system output signal with an error transducer providing an error signal, providing a spectral leak signal in response to a given condition of a given parameter, providing an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal, providing said adaptive filter model with a first error input from said error signal driving said correction signal towards a value matching said system input signal, and a second error input from said spectral leak signal constraining performance of said model by driving said correction signal towards zero.

2. The method according to claim 1 wherein said given parameter is frequency, and said given condition is a designated sub-optimum performance band, such that in said designate sub-optimum performance band, said spectral leak error signal drives said correction signal towards zero and provides sub-optimum performance of said model, and such that outside of said designated sub-optimum performance band, said error signal is maximally effective and said model optimally responds thereto and drives said correction signal toward a value matching said system input signal.

3. The method according to claim 1 comprising providing a copy of said model, and supplying said spectral leak signal through said copy to said second error input.

4. The method according to claim 1 comprising providing a filter responsive to said given parameter, and supplying said spectral leak signal through said filter to said second error input in response to said given condition of said given parameter.

5. The method according to claim 1 comprising providing a copy of said model, providing a filter responsive to said given parameter, and supplying said spectral leak signal through said filter and through said copy to said second error input in response to said given condition of said given parameter.

6. The method according to claim 5 comprising supplying the output of said filter to the input of said copy, and supplying the output of said copy to said second error input, such that said spectral leak signal is supplied first through said filter and then through said copy.

7. An active adaptive control method comprising introducing a control signal from an output transducer to combine with a system input signal and yield a system output signal, sensing said system output signal with an error transducer providing an error signal, providing a spectral leak signal from an auxiliary random signal source uncorrelated with said system input signal in response to a given condition of a given parameter, providing an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal, providing said model with a first error input from said error signal driving said correction signal toward a value matching said system input signal, providing said model with a second error input from said spectral leak signal constraining performance of said model in response to said given condition of said given parameter and driving said correction signal away from said matching value.

8. The method according to claim 7 comprising providing a copy of said model, and supplying said spectral leak signal from said auxiliary random signal source through said copy to said second error input.

9. The method according to claim 7 comprising providing a filter responsive to said given parameter, and supplying said spectral leak signal from said auxiliary random signal source through said filter to said second error input.

10. The method according to claim 7 comprising providing a filter responsive to said given parameter, providing a copy of said model, and supplying said spectral leak signal from said auxiliary random signal source through said filter and through said copy to said second error input.

11. The method according to claim 7 wherein said given parameter is frequency, and said given condition is a designated sub-optimum performance band, such that in said designated sub-optimum performance band, said spectral leak signal provides sub-optimum performance of said model, and such that outside of said designated sub-optimum performance band said error signal is maximally effective and said model optimally responds thereto and drives said correction signal toward said matching value.

12. An active adaptive control method comprising introducing a control signal from an output transducer to combine with a system input signal and yield a system output signal, sensing said system output signal with an error transducer providing an error signal, providing an adaptive filter model having an error input from said error signal and having an output supplying a correction signal to said output transducer to introduce said control signal providing a stopband filter supplying a filtered signal to said model, said filter blocking frequencies in a given stopband at which control of said system input signal by said model is desired, said filter passing frequencies in a given passband at which control of said system input signal by said model is undesired.

13. The method according to claim 12 comprising driving said control signal toward a value matching said system input signal only for frequencies in said stopband.

14. The method according to claim 12 comprising driving said control signal toward a value matching said system input signal for frequencies in said stopband, and driving said control signal away from a value matching said system input signal for frequencies in said passband.

15. The method according to claim 12 comprising providing an auxiliary random signal source supplying a random input signal to said stopband filter uncorrelated with said system input signal.

16. The method according to claim 12 comprising providing a copy of said model having an output supplied to said model.

17. The method according to claim 12 comprising providing said adaptive filter model with a second error input, and supplying said filtered signal to said second error input.

18. The method according to claim 17 comprising providing an auxiliary random signal source supplying a random input signal to said stopband filter uncorrelated with said system input signal, providing a copy of said model having an input from said stopband filter and an output supplied to said second error input.

19. An active adaptive control method comprising introducing a control signal from an output transducer to combine with a system input signal and yield a system output signal, providing an adaptive filter model having an output supplying a correction signal to said output transducer to introduce said control signal to combine with said system input signal and yield said system output signal, providing said adaptive filter model with a pair of error inputs, driving the first of said error inputs to drive said correction signal toward a value matching said system input signal, selectively driving the second of said second error inputs to drive said correctional away from said matching value by driving said correction signal towards zero, and supplying an error signal to said first error input from an error transducer sensing said system output signal, and selectively supplying a spectral leak signal to said second error input in response to a given condition of a given parameter such that

in the presence of said given condition, said spectral leak signal drives said correction signal towards zero, and

in the absence of said given condition, said error signal drives said correction signal towards said matching value.

20. An active adaptive control method comprising introducing a control signal from an output transducer to combine with a system input signal and yield a system output signal, sensing said system output signal with an error transducer providing an error signal, providing an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal, providing said adaptive filter model with a first error input from said error signal driving said correction signal towards a value matching said system input signal, providing said adaptive filter model with a second error input, providing a copy of said model, providing a spectral leak signal in response to a given condition of a given parameter and supplying said spectral leak signal through said copy to said second error input to constrain performance of said model by driving said correction signal away from said matching value.

21. An active adaptive control system comprising an output transducer introducing a control signal to combine with a system input signal and yield a system output signal, an error transducer sensing said system output signal and providing an error signal, an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal and having a first error input from said error signal driving said correction signal towards a value matching said system input signal, and having a second error input, a spectral leak signal source responsive to a given condition of a given parameter and supplying a spectral leak signal to said second error input to constrain performance of said model by driving said correction signal away from said matching value, wherein said spectral leak signal source comprises a copy of said model having an output supplied to said second error input.

22. The invention according to claim 21 wherein said spectral leak signal source further comprises a filter responsive to a given condition of a given parameter and having an output supplied to the input of said copy.

23. The invention according to claim 22 wherein said spectral leak signal source further comprises an auxiliary random signal source uncorrelated with said system input signal and having an output supplied to the input of said filter, such that said spectral leak signal is supplied from said auxiliary random signal source through said filter and through said copy to said second error input.

24. An active adaptive control system comprising an output transducer introducing a control signal to combine with a system input signal and yield a system output signal, an error transducer sensing said system output signal and providing an error signal, an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal and having a first error input from said error signal driving said correction signal towards a value matching said system input signal, and having a second error input, a spectral leak signal source responsive to a given condition of a given parameter and supplying a spectral leak signal to said second error input to constrain performance of said model by driving said correction signal away from said matching value, and a second adaptive filter model adaptively modeling the transfer function from said output transducer to said error transducer, a first auxiliary random signal source supplying a random signal to said second adaptive filter model uncorrelated with said system input signal, and wherein said spectral leak signal source comprises a second auxiliary random signal source supplying a random signal to said second error input uncorrelated with said system input signal.

25. The invention according to claim 24 wherein said random signals from said first and second auxiliary random signal sources are uncorrelated with each other.

26. An active adaptive control system comprising an output transducer introducing a control signal to combine with a system input signal and yield a system output signal, an error transducer sensing said system output signal and providing an error signal, an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal and having a first error input from said error signal driving said correction signal towards a value matching said system input signal, and having a second error input, a spectral leak signal source responsive to a given condition of a given parameter and supplying a spectral leak signal to said second error input to constrain performance of said model by driving said correction signal away from said matching value, wherein said adaptive filter model comprises an IIR filter having an FIR filter A and another FIR filter B each having a first error input from said error signal and each having a second error input, and wherein said spectral leak signal source comprises a copy of filter A having an output supplied to said second error input of filter A, and a copy of filter B having an output supplied to said second error input of filter B.

27. The invention according to claim 26 comprising an auxiliary random signal source uncorrelated with said system input signal, a stopband filter having an input from said auxiliary random signal source, a third summer summing the input filter A and the output of said stopband filter and supplying the output resultant sum to update filter A, a fourth summer summing the input filter B and the output of said stopband filter and supplying the output resultant sum to update filter B.

28. The invention according to claim 26 comprising an auxiliary random signal source uncorrelated with said system input signal, a stopband filter having an input from said auxiliary random signal source, wherein the output of said stopband filter provides an input to said copy of filter A and provides an input to said copy of filter B.

29. The invention according to claim 26 comprising a first summer summing said first and second error inputs to filter A and supplying the output resultant sum to update filter A, and a second summer summing said first and second error inputs to filter B, and supplying the output resultant sum to update filter B.

30. The invention according to claim 29 comprising a first multiplier multiplying the input to filter A and the error signal from said error transducer, and a second multiplier multiplying the input to filter B and the error signal from said error transducer, and wherein said first summer sums the output of said first multiplier and the output of said copy of filter A and supplies the output resultant sum as a weight update signal to filter A, and said second summer sums the output of said second multiplier and the output of said copy of filter B and supplies the output resultant sum as a weight update signal to filter B.

31. The invention according to claim 30 comprising a third multiplier multiplying the output of said copy of filter A and the input to said copy of filter A, and supplying the output resultant product as said second error input to filter A at said first summer, and a fourth multiplier multiplying the output of said copy of filter B and the input to said copy of filter B, and supplying the output resultant product as said second error input to filter B at said second summer.

32. The invention according to claim 31 comprising a second adaptive filter model adaptively modeling the transfer function from said output transducer to said error transducer, an auxiliary random signal source supplying a random signal to said second adaptive filter model uncorrelated with said system input signal, a first copy of said second model having an input receiving the input to filter A, wherein said first multiplier multiplies the output of said first copy of said second model and the error signal from said error transducer and supplies the output resultant product as said first error input to said first summer, a second copy of said second model having an input receiving the input to filter B, wherein said second multiplier multiplies the output of said second copy of said second model and the error signal from said error transducer and supplies the output resultant product as said first error input to said second summer.

33. The invention according to claim 32 comprising a second auxiliary random signal source uncorrelated with said system input signal and uncorrelated with said first mentioned auxiliary random signal source, a stopband filter having an input from said second auxiliary random signal source, wherein the output of said stopband filter provides the input to said copy of filter A and provides the input to said copy of filter B.

34. The invention according to claim 29 comprising a first multiplier multiplying the input to filter A and the output of said first summer and supplying the output resultant product as a weight update signal to filter A, and a second multiplier multiplying the input to filter B and the output of said second summer and supplying the output resultant product as a weight update signal to filter B.

35. The invention according to claim 34 comprising a third summer summing the input to filter A and the input to said copy of filter A and supplying the output resultant sum to said first multiplier for multiplication with the output of said first summer to provide the weight update signal to filter A, a fourth summer summing the input to filter B and the input to said copy of filter B and supplying the output resultant sum to said second multiplier for multiplication with the output of said second summer to provide the weight update signal to filter B.

36. The invention according to claim 35 comprising a second adaptive filter model adaptively modeling the transfer function from said output transducer to said error transducer, an auxiliary random signal source supplying a random signal to said second adaptive filter model uncorrelated with said system input signal, a first copy of said second model having an input receiving the input to filter A and an output supplied to said third summer, wherein said third summer sums the output of said first copy of said second model and the input to said copy of filter A and supplies the output resultant sum to said first multiplier for multiplication with the output of said first summer to provide the weight update signal to filter A, a second copy of said second model having an input receiving the input to filter B and an output supplied to said fourth summer, wherein said fourth summer sums the output of said second copy of said second model and the input to said copy of filter B and supplies the output resultant sum to said second multiplier for nultiplication with the output of said second summer to provide the weight update signal to filter B.

37. The invention according to claim 36 comprising a second auxiliary random signal source uncorrelated with said system input signal and uncorrelated with said first mentioned auxiliary random signal source, a stopband filter having an input from said second auxiliary random signal source, wherein the output of said stopband filter provides an input to said copy of filter A and provides an input to said third summer and provides an input to said copy of filter B and provides an input to said fourth summer.

38. An active adaptive control system comprising an output transducer introducing a control signal to combine with a system input signal and yield a system output signal, an error transducer sensing said system output signal and providing an error signal, an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal and having a first error input from said error signal driving said correction signal towards a value matching said system input signal, and having a second error input, a spectral leak signal source responsive to a given condition of a given parameter and supplying a spectral leak signal to said second error input to constrain performance of said model by driving said correction signal away from said matching value, wherein said adaptive filter model comprises an IIR filter having an FIR filter A and an FIR filter B, a second adaptive filter model adaptively modeling the transfer function from said output transducer to said error transducer, a first auxiliary random signal source supplying a random signal to said second model uncorrelated with said system input signal, a first copy of said second model having an input from the input to filter A, a second copy of said second model having an input from the input to filter B, a second auxiliary random signal source supplying a random signal uncorrelated with said system input signal, a stopband filter having an input from said second auxiliary random signal source, a copy of filter A having an input from said stopband filter, a copy of filter B having in input from said stopband filter, a first multiplier multiplying the output of said first copy of said second model and the error signal from said error transducer, a second multiplier multiplying the output of said second copy of said second model and the error signal from said error transducer, a third multiplier multiplying the output of said copy of filter A and the input to said copy of filter A provided from the output of said stopband filter, a fourth multiplier multiplying the output of said copy of filter B and the input to said copy of filter B provided from the output of said stopband filter, a first summer summing the outputs of said first and third multipliers and supplying the output resultant sum as a weight update signal to filter A, a second summer summing the outputs of said second and fourth multipliers and supplying the output resultant sum as a weight update signal to filter B.

39. An active adaptive control comprising an output transducer introducing a control signal to combine with a system input signal and yield a system output signal, an error transducer sensing said system output signal and providing an error an adaptive filter model having an output outputting a correction signal to said output transducer to introduce said control signal and having a first error input from said error signal driving said correction signal towards a value matching said system input signal, and having a second error input, a spectral leak signal source responsive to a given condition of a given parameter and supplying a spectral leak signal to said second error input to constrain performance of said model by driving said correction signal away from said matching value, wherein said adaptive filter model comprises an IIR filter having an FIR filter A and an FIR filter B, a second adaptive filter model adaptively modeling the transfer function from said output transducer to said error transducer, a first auxiliary random signal source supplying a random signal to said second model uncorrelated with said system input signal, a first copy of said second model having an input from the input to filter A, a second copy of said second model having an input from the input to filter B, a second auxiliary random-signal source supplying a random signal uncorrelated with said system input signal, a stopband filter having an input from said second auxiliary random signal source, a copy of filter A having an input from said stopband filter, a copy of filter B having an input from said stopband filter, a first summer summing the output of said first copy of said second model and the output of said stopband filter providing the input to said copy of filter A, a second summer summing the output of said second copy of said second model and the output of said stopband filter providing the input to said copy of filter B, a third summer summing the output of said copy of filter A and the error signal from said error transducer, a fourth summer summing the output of said copy of filter B and the error signal from said error transducer, a first multiplier multiplying the outputs of said first and third summers and supplying the output resultant product as a weight update signal to filter A, a second multiplier multiplying the outputs of said second and fourth summers and supplying the output resultant product as a weight update signal to filter B.