Active acoustic and structural vibration control without online controller adjustment and path modeling

• US 6,601,054 B1
• Filed: 08/11/2000
• Issued: 07/29/2003
• Est. Priority Date: 08/16/1999
• Status: Expired due to Fees

• Current Assignee: Maryland Technology Corp.
• Original Assignee: Maryland Technology Corp.
• Inventors: Yu, Lei, Lo, James T.
• Primary Examiner(s): Davis, George B.
• Application Number:
  US09/637,406
• Time in Patent Office: 1,082 Days
• Field of Search: 706/30, 706/14, 706/23
• US Class Current: 706/30
• CPC Class Codes: G05B 13/027 using neural networks only
• CPC Class Codes: G05B 17/02 electric
• CPC Class Codes: G05B 5/01 electric
• CPC Class Codes: G06N 3/044 Recurrent networks, e.g. Hopfield networks

Abstract

Active vibration control (AVC) systems without online path modeling and controller adjustment are provided that are able to adapt to an uncertain operating environment. The controller (250, 280, 315, 252, 282, 317, 254, 319) of such an AVC system is an adaptive recursive neural network whose weights are determined in an offline training and are held fixed online during the operation of the system. AVC feedforward, feedback, and feedforward-feedback systems in accordance with the present invention are described. An AVC feedforward system has no error sensor and an AVC feedback system has no reference sensor. All sensor outputs of an AVC system are processed by the controller for generating control signals to drive at least one secondary source (240). While an error sensor (480, 481) must be a vibrational sensor, a reference sensor (230, 270, 295, 305, 330) may be either a vibrational or nonvibrational sensor. The provided AVC systems reduce or eliminate most of such shortcomings of the prior-art AVC systems as use of an error sensor, relatively slow convergence of a weight/waveform adjustment algorithm, frequent adjustment of a path model, use of a high-order adaptive linear transversal filter, instability of an adaptive linear recursive filter, failure to use a useful nonvibrational reference sensor, failure to deal with the nonlinear behavior of a primary or secondary path, weight adjustment using control predicted values, use of an identification neural network, and online adjustment of the weights of a neural network.

Claims

1. An active vibration control system for attenuating at least one primary vibration by injecting at least one cancelling vibration in an uncertain operating environment, comprising

(a). reference sensor means, comprising at least one reference sensor, for deriving at least one reference signal containing information about said at least one primary vibration;

(b). a recursive neural network, comprising a plurality of weights, which processes said at least one reference signal to generate at least one control signal; and

(c). secondary source means, comprising at least one secondary source, for converting said at least one control signal into said at least one cancelling vibration,

wherein said weights are held fixed online during the operation of said active vibration control system.

2. The active vibration control system of claim 1, wherein said at least one primary vibration is at least one acoustic vibration.

3. The active vibration control system of claim 1, wherein said at least one primary vibration is at least one structural vibration.

4. The active vibration control system of claim 1, wherein said at least one primary vibration is an engine exhaust noise and said at least one secondary source is a speaker.

5. The active vibration control system of claim 1, wherein said reference sensor means comprises a plurality of reference sensors, and said secondary source means comprises a plurality of secondary sources.

6. The active vibration control system of claim 1, wherein said reference sensor means comprises at least one nonvibrational sensor.

7. The active vibration control system of claim 1, wherein said reference sensor means comprises at least one vibrational sensor.

8. The active vibration control system of claim 1, wherein said recursive neural network is a multilayer perceptron with interconnected neurons and tapped delay lines.

9. The active vibration control system of claim 1, wherein said recursive neural network is a multilayer perceptron with interconnected neurons, output feedbacks and tapped delay lines.

10. An active vibration control system for attenuating at least one primary vibration by injecting at least one cancelling vibration in an uncertain operating environment, comprising

(a). reference sensor means, comprising at least one reference sensor, for deriving at least one reference signal containing information about said at least one primary vibration;

(b). error sensor means, comprising at least one error sensor, for sensing the combination of said at least one primary vibration and said at least one cancelling vibration to provide at least one error signal;

(c). a recursive neural network, comprising a plurality of weights, for processing said at least one reference signal and said at least one error signal to generate at least one control signal; and

(d). secondary source means, comprising at least one secondary source, for converting said at least one control signal into said at least one cancelling vibration,

wherein said weights are held fixed online during the operation of said active vibration control system.

11. The active vibration control system of claim 10, wherein said at least one primary vibration is at least one acoustic vibration.

12. The active vibration control system of claim 10, wherein said at least one primary vibration is at least one structural vibration.

13. The active vibration control system of claim 10, wherein no error sensor is placed at at least one objective point.

14. The active vibration control system of claim 10, wherein said reference sensor means comprises at least one nonvibrational sensor.

15. The active vibration control system of claim 10, wherein said at least one primary vibration is at least one acoustic vibration and at least one structural vibration.

16. The active vibration control system of claim 10, wherein said reference sensor means comprises at least one vibrational sensor.

17. The active vibration control system of claim 10, wherein said recursive neural network is a multilayer perceptron with interconnected neurons and tapped delay lines.

18. The active vibration control system of claim 10, wherein said recursive neural network is a multilayer perceptron with interconnected neurons, output feedbacks and tapped delay lines.

19. The active vibration control system of claim 10, wherein said reference sensor means comprises a plurality of reference sensors, said error sensor means comprises a plurality of error sensors, and said secondary source means comprises a plurality of secondary sources.

20. An active vibration control system for attenuating at least one primary vibration by injecting at least one cancelling vibration in an uncertain operating environment, comprising

(a). error sensor means, comprising at least one error sensor, for sensing the combination of said at least one primary vibration and said at least one cancelling vibration to provide at least one error signal;

(b). a recursive neural network, comprising a plurality of weights, for processing said at least one error signal to generate at least one control signal; and

(c). secondary source means, comprising at least one secondary source, for converting said at least one control signal into said at least one cancelling vibration,

wherein said plurality of weights are held fixed online during the operation of said active vibration control system.

21. The active vibration control system of claim 20, wherein said at least one primary vibration is at least one acoustic vibration.

22. The active vibration control system of claim 20, wherein said at least one primary vibration is at least one structural vibration.

23. The active vibration control system of claim 20, wherein said at least one primary vibration is at least one acoustic vibration and at least one structural vibration.

24. The active vibration control system of claim 20, wherein no error sensor is placed at at least one objective point.

25. The active vibration control system of claim 20, wherein said error sensor means comprises a plurality of error sensors, and said secondary source means comprises a plurality of secondary sources.

26. The active vibration control system of claim 20, wherein said recursive neural network is a multilayer perceptron with interconnected neurons and tapped delay lines.

27. The active vibration control system of claim 20, wherein said recursive neural network is a multilayer perceptron with interconnected neurons, output feedbacks and tapped delay lines.