National Technical University of Athens (NTUA)
Local and global active noise control systems in vehicle cabins
Abstract
dc:descriptionExcessive noise in vehicle cabins, particularly from the propulsion systems and their interaction with the environment (e.g., wind, road, waves),significantly diminishes the acoustic comfort of passengers. At low frequencies, where acoustic energy is typically concentrated in these disturbances, the feasibility of using conventional passive sound insulation techniques is limited due to the increased weight and thickness of these structures as the wavelength of the acoustic disturbances increases. The primary objective of this doctoral thesis is to devise and evaluate innovative systems for global and local active noise control in the cabins of small aircraft and yachts. The proposed methods focus on enhancing the performance of existing techniques, while also exploring the potential for integration into real-world applications. Initially, the focus is on developing a local active noise control system, known as the active headrest, which is installed at the front headrest of the passenger seat. The purpose of this system is to create a quiet zone around the ears, where a substantial reduction in sound pressure level is achieved, and the zone must be large enough to allow for gentle head movements while maintaining the reduced noise level. The development of this system involved an initial evaluation of a low-complexity linear adaptive control algorithm, with the aim of facilitating its implementation in digital signal processing systems and enabling the operation of multiple such systems, without requiring powerful processors. The mixed error FxLMS algorithm, an alternative to the multichannel FxLMS algorithm, was found to be an effective solution for local noise control systems, as it reduced computational complexity without significantly impacting performance. In the second phase of the system’s development, a method for estimating acoustic pressure in the passenger’s ears was incorporated, using microphones positioned a few centimeters away. This method, based on linear extrapolation, allowed for the relocation of the quiet zone to the desired location without the need for physical microphones, which are often constrained by spatial limitations. Finally, to achieve greater sound level reduction, a nonlinear algorithm was employed, based on a simple single-level functional link neural network. This algorithm combined mixed error and linear extrapolation techniques, offering the benefits previously mentioned. The system demonstrated improved performance, particularly for real-world acoustic disturbances recorded in vehicle cabins. In the last part of this thesis, an investigation was conducted into the sound field of the entire cabin when multiple local control systems were operational simultaneously. The findings revealed that although quiet zones persisted in front of the headrest, the sound levels in the sub-regions of the cabin remained high. To address this issue, a global active noise control system was devised, which utilized active sound absorbers. The examination of this system demonstrated that by strategically positioning the absorbers, a reduction of over 10 dB in sound pressure levels could be attained in all areas of a small cabin.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Mylonas, Dimitrios
- Μυλωνάς, Δημήτριος
Subjects
dc:subject × 19- Τοπικός ενεργητικός έλεγχος θορύβου
- Ολικός ενεργητικός έλεγχος θορύβου
- Ακύρωση θορύβου
- Θόρυβος καμπίνας αεροσκάφους
- Local active noise control
- Global active noise control
- Noise cancellation
- Active sound absorber
- Aircraft cabin
- Active headrest
- Filtered-x least mean squares
- Functional link neural network
- Quiet zones
- Επιστήμη Μηχανολόγου Μηχανικού
- Επιστήμες Μηχανικού και Τεχνολογία
- Μηχανική ήχου
- Mechanical Engineering
- Engineering and Technology
- Audio engineering
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/57305