Algoritmi za aktivnu zaštitu od buke i vibracija

doctoral thesis

• Title (english): Algorithms for active control of noise and vibration
• Author: Martin Dadić
• Mentor: Branko Somek (mentor)
• Committee member: Zijad Haznadar (član povjerenstva)
• Committee member: Branko Somek (član povjerenstva)
• Committee member: Igor Zorić (član povjerenstva)
• Committee member: Borivoj Modlic (član povjerenstva)
• Committee member: Mladen Maletić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Electroacoustics); Zagreb
• Defense date and country: 2000-03-27, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering
• Universal decimal classification (UDC): 621.3 - Electrical engineering

Abstract

Tema ove disertacije su aktivni sustavi za zaštitu od buke i vibracija. Upotrebom pasivnih metoda zaštite od buke često nije moguće postići zadovoljavajuće rješenje, a koji put, ako je i moguće, nije financijski isplativo. Razvojem elektronike i audiotehnike, kao rješenje se sve više nameće upotreba aktivnih sustava za zaštitu od buke i vibracija, gdje se neželjeni zvučni signal buke nastoji poništiti protufaznim signalom, koji proizvodi elektronički sustav. Realizaciju učinkovitih aktivnih sustava zaštite omogućava u prvom redu upotreba suvremenih metoda adaptivne obrade signala. U ovoj radnji obrađeni su algoritmi adaptivne obrade signala koji se koriste u aktivnoj zaštiti. Tu se prije svega koristi LMS algoritam i njegove izvedenice. Kao neželjeni efekt u aktivnim sustavima zaštite koji koriste akustički senzor referentnog signala javlja se neželjena akustička povratna veza od sekundarnog izvora do senzora referentnog signala. Iako postoje akustičke mjere poput usmjerenih izvora i senzora, njihova je učinkovitost ovisna o nepromjenjivosti uvjeta prostora u kojemu se vrši zaštita. Upotrebom rekurzivnih adaptivnih filtara ovaj se problem može riješiti, i u tu svrhu se koristi u prvom redu Erikssonov FURLMS algoritam. Međutim rekurzivni adaptivni filtri nisu bezuvjetno stabilni, a i za navedeni se algoritam postavljalo pitanje konvergencije. U ovoj je radnji u slučaju akustičke povratne veze između sekundarnog izvora i senzora referentnog signala predložena upotreba filtrirano-x adaptivnog rekurzivnog filtra zasnovanog na Legendreovim funkcijama. Predložena filtarska struktura je stabilna i bez lokalnih minimuma, a usporedbom sa FURLMS algoritmom simulacijski je pokazana bolja konvergencija za realnu akustičku konfiguraciju kratke cijevi uz isti broj koeficijenata. Usporedbom dvaju algoritama uz uvjet istog broja multiplikacija po iteraciji postiže se jednaka razina preostale buke za dva algoritma.

Abstract (english)

This thesis is dealing with active noise and vibration control systems. Passive methods in noise control often cannot give satisfying result, and in many cases it is too costly even if it is actually possible. Recent development in electronics and audiotechnics allowed application of active noise and vibration control instead. In active control unwanted noise signal is acoustically summed with an antiphase signal, produced by the electronic system. Application of effective active noise and vibration control systems became possible with modern adaptive signal processing methods. In this thesis algorithms of adaptive signal processing used in active control are presented. Most used algorithms are those based on LMS algorithm. As an unwanted effect, when using active noise control with an acoustical reference sensor, there is a acoustical feedback between secondary source and reference signal sensor. There are acoustical countermeasures, such as directed sources and sensors, they all depend on invariable conditions of space where the control is performed. Application of infinite impulse response (IIR) adaptive filters can solve this problem, and there is first of all used Eriksson's FURLMS algorithm. IIR adaptive filters are not unconditionally stable, and the mentioned one was particularly discussed about its convergence. In this thesis was proposed a filtered-x adaptive IIR filter based on Legendre functions for use when such a feedback exists. The proposed filter is stable and has got unimodal performance surface. Compared with FURLMS algorithm, it is shown by simulation better convergence for the real acoustical configuration of a short duct, using the same number of coefficients. Comparing two algorithms on the basis of same number of multiplications per iteration, it is achieved the same level of residual noise.

• Keywords: Aktivna zaštita od buke; aktivna zaštita od vibracija; adaptivni filtri
• Keywords (english): Active noise control; active vibration control; adaptive filters
• Language: croatian
• URN:NBN: urn:nbn:hr:168:825409
• Study programme: Title: Doctoral study programme "Electrical Engineering and Computing"; Study programme type: university; Study level: postgraduate; Academic / professional title: doktor/doktorica znanosti, po-dručje tehničkih znanosti (doktor/doktorica znanosti, po-dručje tehničkih znanosti)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-04-02 10:33:42