| 1. |
- Håkansson, Lars, et al.
(författare)
-
NOISE CANCELLING HEADSETS FOR SPEECH COMMUNICATION
- 2002
-
Ingår i: Noise Reduction in Speech Applications. - Boca Raton, Florida : CRC Press. - 0849309492 ; , s. 305-327
-
Bokkapitel (övrigt vetenskapligt/konstnärligt)abstract
- Headsets for speech communication are used in a wide range of applications. The basic idea is to allow hands-free speech communication, leaving both hands available for other tasks. One typical headset application is aircraft pilot communication. The pilot must be able to communicate with personnel on the ground and at the same time use both hands to control the aircraft. Communication headsets usually consist of a pair of headphones and a microphone attached with an adjustable boom. Headphone design varies widely between different manufacturers and models. In its simplest form, the headphone has an open construction providing little or no attenuation of the environmental noise. In headsets designed for noisy environments, the headphones are mounted in ear cups with cushions that provide some attenuation. The microphone is primarily designed to pick up the speech signal, but if the headset is used in a noisy environment, the background noise will also be picked up and transmitted together with the speech. As a consequence, speech intelligibility at the receive end will be reduced, possibly to zero. To increase the speech-to-noise ratio, it is common to use a directional microphone that has a lower sensitivity to sound incident from other directions than the frontal direction. In addition to this, the microphone electronics are usually equipped with a gate function that completely shuts off the microphone signal if its level drops below a threshold value. The purpose of the gate is to open the channel for transmission only when a speech signal is present. Headsets are frequently used in noisy environments where they suffer from problems of speech intelligibility. Even if an ear-cup type headset is used, the attenuation is relatively poor for low frequencies. Low frequency noise has a masking effect on speech, which significantly reduces the speech intelligibility. Several cases have been reported in which the sound level of the communication signal was increased to hazardous levels by the user to overcome this low frequency masking effect [1,2]. Ear exposure to the communication system resulted in hearing damage, such as hearing loss, tinnitus and hyperacusis.
|
|
| 2. |
- Johansson, Sven, et al.
(författare)
-
A Novel Multiple-Reference Algorithm for Active Control of Propeller-Induced Noise in Aircraft Cabins
- 2000
-
Rapport (övrigt vetenskapligt/konstnärligt)abstract
- The cabin noise inside propeller aircraft is essentially dominated by strong tonal components at harmonics of the blade passage frequency of the propellers. In order to achieve an efficient reduction of such a periodic low frequency noise, it is advisable to use an active noise control system based on adaptive narrowband feedforward techniques. The feedforward controller presented in this paper exploits narrowband assumptions by using complex-valued filtering and complex modeling of control paths. This paper introduces a multiple reference controller based on the novel actuator-individual normalized Filtered-X Least-Mean-Squares (FX LMS) algorithm. This algorithm combines low computational complexity with high performance. The algorithm is of the LMS-type. However, owing to the novel normalization of the algorithm it can also be regarded as a Newton-type algorithm. A comparison between the actuator-individual normalized FX LMS algorithm and the ordinary normalized FX LMS algorithm is presented. The results demonstrate better performance in terms of convergence rate and tracking properties when the Newton-like actuator-individual normalized FX LMS algorithm is used as compared with the conventional normalized LMS algorithm. The evaluation was performed using noise signals recorded inside the cabin of a twin engine propeller aircraft during flight. The paper also discusses variants of the actuator-individual normalized FX LMS algorithm.
|
|
| 3. |
- Johansson, Sven, et al.
(författare)
-
Active Control of Sound using the Actuator-Individual Normalized Filtered-X LMS Algorithm
- 2001
-
Konferensbidrag (refereegranskat)abstract
- The control algorithm is a central part of an active noise control system. The computational power of the DSP hardware is limited, and therefore algorithms with high computational complexity allow fewer control sources and sensors to be used, often with reduced noise attenuation as a result. This paper presents results from practical experiments using an ANC system based on the actuator-individual normalized Filtered-X LMS algorithm. The algorithm combines low computational complexity with high performance. For that reason the algorithm is suitable for use in large multi-channel systems in order to reduce the computional power required by the control system.
|
|
| 4. |
- Johansson, Sven, et al.
(författare)
-
ACTIVE HEADSET FOR ENHANCED SPEECH INTELLIGIBILITY IN INTERCOM SYSTEMS
- 2001
-
Konferensbidrag (refereegranskat)abstract
- Low frequency noise has a masking effect on speech, which significantly reduces the speech intelligibility. Usually, in order to overcome the masking effect the sound level of the communication signal is increased. Exposure to the communication system then causes hearing damage. A feasible approach is to reduce the low frequency noise by using active techniques combined with a method for reducing the noise in the intercom microphone signal. In order to achieve an efficient low frequency noise attenuation inside the headset an active hybrid headset is proposed, and in order to suppress the background noise in speech signals spectral subtraction is used. Improved low frequency noise attenuation enables lower communication levels and reduces the risk for hearing damage.
|
|
| 5. |
- Johansson, Sven, et al.
(författare)
-
Comparison of Multiple- and Single-Reference MIMO Active Noise Control Approaches Using Data Measured in a Dornier 328 Aircraft
- 2000
-
Ingår i: International Journal of Acoustics and Vibration. - : Int. Inst. of Acoustics and Vibration (IIAV). - 1027-5851 .- 2415-1408. ; 5:4, s. 77-88
-
Tidskriftsartikel (refereegranskat)abstract
- In many applications of noise control, the greatest annoyance is caused by periodic low frequency noise. Successful reduction of such noise can often be achieved by using an active noise control system with narrowband feedforward control. If several noise sources contribute to the sound field, a multiple-reference control system is usually required. This type of system allows the reference signals from each noise source to be processed individually within the controller, thereby enabling individual control of the sound field from each noise source. The present paper addresses the problem of controlling noise from two sources that are more or less synchronized. A typical application is the control of propeller-generated noise within a twin propeller aircraft. To find out whether a multiple-reference controller is necessary, or if a single-reference controller is sufficient, the performance of a single- versus twin-reference control algorithm is evaluated in a comparative study. The study is performed as a computer simulation (off-line evaluation) using real-life data recorded in a Dornier 328 under different flight conditions. The results demonstrate that the twin-reference controller performs better than the single-reference whenever there is a slight deviation in the rotational speed of the two propellers. The paper also treats the generation of reference signals. The approach presented is based on a fixed sampling rate and uses a sliding FFT filtering technique.
|
|
| 6. |
- Johansson, Sven, et al.
(författare)
-
Experimental Performance Evaluation of a Multi-Reference Algorithm for Active Control of Propeller-Induced Cabin Noise
- 2000
-
Rapport (övrigt vetenskapligt/konstnärligt)abstract
- A noisy environment dominated by low frequency noise can often be improved through the use of active noise control. This situation arises naturally in propeller aircraft where the propellers induce periodic low frequency noise inside the cabin. The cabin noise is typically rather high, and the passenger flight comfort could be improved considerably if this level were significantly reduced. This paper discusses the operation and robustness of a narrowband feedforward active noise control system in a practical installation. The ANC system used 8 control sources and 11 control microphones, and the control algorithm was the multi-reference actuator-individual normalized filtered-x least mean squares algorithm. The experiment was performed in a full-scale fuselage section of a SAAB 340 aircraft. To produce the propeller noise, loudspeakers mounted in a ring around the fuselage were used. Results are presented from a series of experiments on the active control of propeller-induced cabin noise. Among the ``flight" conditions evaluated were: conditions where the ``propellers" were completely synchronized, and conditions with constant as well as time-varying frequency beat between ``left and right propellers."
|
|
| 7. |
- Sjösten, Per, et al.
(författare)
-
A Practical, Fast and Cost-efficient Algorithm Multiple Input, Multiple Output Active Noise Control Applications
- 2001
-
Konferensbidrag (refereegranskat)abstract
- Active noise control has proven to be an efficient solution to low frequency noise problems in many different applications. A large part of these applications are concerned with harmonic noise control and typically require control systems with several inputs and outputs. As the systems grow, the demand for processor capacity increases rapidly, resulting in large and expensive hardware platforms. As multiple-input, multiple-output noise control is restricted to the control of periodic noise, a controller structure that is adapted to the signal type may significantly reduce the requirements on the hardware capacity. This paper discusses a complex, time-domain controller that is designed for the control of harmonic components. The structure of the controller is simple, easily implemented and can easily be extended to handle any number of noise references and any number of harmonics. The convergence properties of a multiple-input, multiple-output control system depends largely on the acoustic coupling between the active sources and the control sensors. To get accurate and stable control it is necessary to use normalization, i.e. a weighting function that optimizes the controller for each control source. With the presented controller structure, the use of normalization is straightforward and a number of different approaches for normalization is discussed as well as examples from practical inplementations.
|
|
| 8. |
- Sjösten, Per, et al.
(författare)
-
Considerations on Large Applications of Active Noise Control, Part I : Theory
- 2003
-
Ingår i: Acta Acoustica united with Acustica. - : S. Hirzel Verlag. - 1610-1928 .- 1861-9959. ; 89:5, s. 822-833
-
Tidskriftsartikel (refereegranskat)abstract
- The application of active noise or vibration control to a real situation is usually a great challenge. The step from computer simulations, or even lab studies, to the actual implementation is often much larger than first expected. This is especially true for multiple-error, multiple-output control, in which careful planning and a very great deal of preparatory work is crucial to the success of the project. The present paper is the first in a series of two, that treat the process of preparing and evaluating an active noise control system. Part I gives a general discussion on proposed methods from a theoretical standpoint. Part II describes the application of these methods in practical measurements in an aircraft application. The emphasis is on multiple-error, multiple-output control but the general guidelines haves proven to be useful even in small applications. The discussion covers some acoustical aspects as well as the signal processing view. A method, based on simulated annealing using the metropolis sampler is suggested for searching the suboptimal positions for the control sources and sensors, including a thorough discussion on the preparation of input data for the optimisation process. Finally, a version of the complex lms algorithm, normalised individually for each actuator, is discussed.
|
|
| 9. |
- Sjösten, Per, et al.
(författare)
-
Considerations on Large Applications of Active Noise Control, Part II : Experimental Results
- 2003
-
Ingår i: Acta Acoustica united with Acustica. - : S. Hirzel, Germany. - 1610-1928 .- 1861-9959. ; 89:5, s. 834-843
-
Tidskriftsartikel (refereegranskat)abstract
- This paper treats the practical aspects of the prediction, implementation and verification of an active noise control system. The system was implemented in a mock-up of a Saab 340 propeller aircraft. A search algorithm, based on simulated annealing using the metropolis sampler was used to determine the positions for the control sources and sensors. The sound field was generated by a specially designed noise simulation system, using 12 individually controlled loudspeakers and a DSP--based signal generator. A version of the individually normalized complex lms algorithm was implemented for the control of the sound field. The prediction and verification of the active control system, was made in 20 evaluation positions, distributed at ear level above the seats. This is the second paper in a series of two, concerning the preparation, selection and verification of an active noise control system. Part I contains a theoretical treatment with some general guidelines for the implementation. Paper II is focused on the practical aspects of the implementation and verification.
|
|
