Psychoacoustic quantities and their relevance for sound-quality optimization of switched reluctance machines

Abstract

This paper gives a brief introduction to noise development in switched reluctance machines (SRMs) and presents psychoacoiustic quantities that allow an objective eva.luation of their sound qua.lity by taking into accouLnt the charcteristics of the human hearing system. The relevance of these quantities is deteined by listening tests and statistical analyses. The goal of this work is to demonstrate that a scceessfLl acoustic optimization of a SRM should be petformed based on both technical and psychoa.coustic know-how. 12K 2005 -I2're.xkn ISBN PH FINCIRUTHl Sbas&t DEl 2005 -L)i--;c11i ISBN :90-75815-08-5 P.7X HIE 2005 -. PS.

Effectiveness of Noise Reducing Measures in Switched Reluctance Drives

“…It has been shown in [1] that the determination of the sole sound pressure level is not sufficient for the representation of annoyance reception and that additional quantities such as the loudness, sharpness and tonality have to be taken into account. Therefore, the influence of the dc-link voltage level is experimentally analyzed with respect to Zwicker-loudness and sharpness for the two-phase 8/4 SRM and it can be shown again that a lower dc-link voltage benefits the pleasantness of the sound since it reduces both, loudness and sharpness (Fig.22).…”

“…Finally, noise perception marks the end of the signal path, since the non-linearities and psycho-acoustic effects in the human ear and brain (A-weighting, etc.) play an important role for the annoyance of sound [1].…”

“…However, it has been shown in [1] that additional quantities such as Zwicker-loudness [17], sharpness and tonality [18] have to be taken into account to represent the annoyance of sound in SRM well. Zwicker loudness is basically related to the sound pressure level additionally taking masking effects into account.…”

Aproximación metodológica a la obtención de modelos de percepción de molestias a partir de parámetros psicoacústicos en vehículos ferroviarios. Aplicación a trenes de alta velocidad.

“…Las escalas discretas, tanto la verbal como la numérica, como se ha comentado anteriormente, limitan en mayor o menor medida la respuesta del usuario ya que el rango de respuestas está acotado. Para evitar este efecto, Kasper et al (2005) propone la combinación de ambos métodos mediante escala de categorías. Este método fue utilizado para estudiar la calidad del ruido creado por maquinas eléctricas.…”

“…En general, el número de participantes según el género está equilibrado (Kuwano et al;2004;Ma y Yano;2005). En algunos de los estudios examinados hay una tendencia a utilizar más hombres que mujeres (Gonzalez et al, 2003) y otros ni siquiera especifican la distinción de géneros (Boullet et al, 2007;(Patsouras et al, 2002); (Kasper et al, 2005 Por último, hay que tener en cuenta que no todos los sujetos que participan en la evaluación tienen la misma capacidad para identificar los atributos de los sonidos con la misma precisión. Otto et al (2001) afirma que alrededor del 10 % de los evaluadores previsiblemente tendrán baja fiabilidad.…”