The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Untuk menilai turun naik tindak balas keluaran sistem akustik persekitaran sebenar yang teruja dengan input rawak sewenang-wenangnya, adalah penting untuk meramalkan keseluruhan bentuk taburan kebarangkalian berkait rapat dengan banyak indeks penilaian hingar. Lx, Leq dan sebagainya. Dalam makalah ini, kaedah penilaian jenis baharu dicadangkan dengan memperkenalkan model fungsi jenis tertib rendah dan lebih tinggi yang dipadankan dengan ramalan bentuk taburan kebarangkalian tindak balas terutamanya dari sudut pandangan berorientasikan masalah. Oleh kerana sifat bukan negatif pembolehubah keamatan bunyi, fungsi ketumpatan kebarangkalian tindak balas boleh dinyatakan secara munasabah terlebih dahulu secara teori dengan bentuk siri pengembangan Laguerre statistik. Ciri sistem antara input dan output boleh diterangkan oleh hubungan regresi antara parameter taburan (mengandungi pekali pengembangan ungkapan ini) dan input stokastik. Fungsi regresi ini boleh dinyatakan dari segi pengembangan siri ortogon. Oleh kerana, dalam persekitaran sebenar, keluaran yang diperhatikan tidak dapat tidak dicemari oleh bunyi latar belakang, fungsi regresi di atas tidak boleh digunakan secara langsung sebagai model untuk persekitaran sebenar. Mujurlah, output yang diperhatikan boleh diberikan dengan jumlah output sistem dan hingar latar belakang berdasarkan penambahan kuantiti intensiti dan momen statistik hingar latar boleh diperolehi terlebih dahulu. Jadi, model yang mengaitkan fungsi regresi dengan fungsi keluaran yang diperhatikan boleh diperolehi. Seterusnya, parameter fungsi regresi ditentukan berdasarkan kriteria ralat kuasa dua terkecil dan ukuran kebebasan statistik mengikut tahap sifat bukan Gaussian bagi fungsi keluaran yang diperhatikan. Oleh itu, dengan menggunakan fungsi regresi yang diperolehi oleh kaedah pengenalan yang dicadangkan, taburan kebarangkalian output mengurangkan hingar latar belakang boleh diramalkan. Akhir sekali, keberkesanan kaedah yang dicadangkan juga disahkan secara eksperimen dengan menerapkannya pada sistem akustik dalaman-luar yang sebenar.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Yoshifumi FUJITA, Mitsuo OHTA, "Stochastic Evaluation of Acoustic Environment with Noise Cancellation under Introduction of Hierarchically Functional Type Probability Model" in IEICE TRANSACTIONS on Fundamentals,
vol. E84-A, no. 2, pp. 467-474, February 2001, doi: .
Abstract: For evaluating the output response fluctuation of the actual environmental acoustic system excited by arbitrary random inputs, it is important to predict a whole probability distribution form closely connected with many noise evaluation indexes Lx, Leq and so on. In this paper, a new type evaluation method is proposed by introducing lower and higher order type functional models matched to the prediction of the response probability distribution form especially from a problem-oriented viewpoint. Because of the non-negative property of the sound intensity variable, the response probability density function can be reasonably expressed in advance theoretically by a statistical Laguerre expansion series form. The system characteristic between input and output can be described by the regression relationship between the distribution parameters (containing expansion coefficients of this expression) and the stochastic input. These regression functions can be expressed in terms of the orthogonal series expansion. Since, in the actual environment, the observed output is inevitably contaminated by the background noise, the above regression functions can not be directly employed as the models for the actual environment. Fortunately, the observed output can be given by the sum of the system output and the background noise on the basis of additivity of intensity quantity and the statistical moments of the background noise can be obtained in advance. So, the models relating the regression functions to the function of the observed output can be derived. Next, the parameters of the regression functions are determined based on the least-squares error criteria and the measure of statistical independency according to the level of non-Gaussian property of the function of the observed output. Thus, by using the regression functions obtained by the proposed identification method, the probability distribution of the output reducing the background noise can be predicted. Finally, the effectiveness of the proposed method is confirmed experimentally too by applying it to an actual indoor-outdoor acoustic system.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/e84-a_2_467/_p
Salinan
@ARTICLE{e84-a_2_467,
author={Yoshifumi FUJITA, Mitsuo OHTA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Stochastic Evaluation of Acoustic Environment with Noise Cancellation under Introduction of Hierarchically Functional Type Probability Model},
year={2001},
volume={E84-A},
number={2},
pages={467-474},
abstract={For evaluating the output response fluctuation of the actual environmental acoustic system excited by arbitrary random inputs, it is important to predict a whole probability distribution form closely connected with many noise evaluation indexes Lx, Leq and so on. In this paper, a new type evaluation method is proposed by introducing lower and higher order type functional models matched to the prediction of the response probability distribution form especially from a problem-oriented viewpoint. Because of the non-negative property of the sound intensity variable, the response probability density function can be reasonably expressed in advance theoretically by a statistical Laguerre expansion series form. The system characteristic between input and output can be described by the regression relationship between the distribution parameters (containing expansion coefficients of this expression) and the stochastic input. These regression functions can be expressed in terms of the orthogonal series expansion. Since, in the actual environment, the observed output is inevitably contaminated by the background noise, the above regression functions can not be directly employed as the models for the actual environment. Fortunately, the observed output can be given by the sum of the system output and the background noise on the basis of additivity of intensity quantity and the statistical moments of the background noise can be obtained in advance. So, the models relating the regression functions to the function of the observed output can be derived. Next, the parameters of the regression functions are determined based on the least-squares error criteria and the measure of statistical independency according to the level of non-Gaussian property of the function of the observed output. Thus, by using the regression functions obtained by the proposed identification method, the probability distribution of the output reducing the background noise can be predicted. Finally, the effectiveness of the proposed method is confirmed experimentally too by applying it to an actual indoor-outdoor acoustic system.},
keywords={},
doi={},
ISSN={},
month={February},}
Salinan
TY - JOUR
TI - Stochastic Evaluation of Acoustic Environment with Noise Cancellation under Introduction of Hierarchically Functional Type Probability Model
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 467
EP - 474
AU - Yoshifumi FUJITA
AU - Mitsuo OHTA
PY - 2001
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E84-A
IS - 2
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - February 2001
AB - For evaluating the output response fluctuation of the actual environmental acoustic system excited by arbitrary random inputs, it is important to predict a whole probability distribution form closely connected with many noise evaluation indexes Lx, Leq and so on. In this paper, a new type evaluation method is proposed by introducing lower and higher order type functional models matched to the prediction of the response probability distribution form especially from a problem-oriented viewpoint. Because of the non-negative property of the sound intensity variable, the response probability density function can be reasonably expressed in advance theoretically by a statistical Laguerre expansion series form. The system characteristic between input and output can be described by the regression relationship between the distribution parameters (containing expansion coefficients of this expression) and the stochastic input. These regression functions can be expressed in terms of the orthogonal series expansion. Since, in the actual environment, the observed output is inevitably contaminated by the background noise, the above regression functions can not be directly employed as the models for the actual environment. Fortunately, the observed output can be given by the sum of the system output and the background noise on the basis of additivity of intensity quantity and the statistical moments of the background noise can be obtained in advance. So, the models relating the regression functions to the function of the observed output can be derived. Next, the parameters of the regression functions are determined based on the least-squares error criteria and the measure of statistical independency according to the level of non-Gaussian property of the function of the observed output. Thus, by using the regression functions obtained by the proposed identification method, the probability distribution of the output reducing the background noise can be predicted. Finally, the effectiveness of the proposed method is confirmed experimentally too by applying it to an actual indoor-outdoor acoustic system.
ER -