Análise de ruídos automotivos por parâmetros psicoacústicos
Carregando...
Arquivos
Citações na Scopus
Tipo de produção
Dissertação
Data
2022
Autores
Sartori, Renato Parede
Orientador
Thomaz, C. E.
Periódico
Título da Revista
ISSN da Revista
Título de Volume
Citação
SARTORI, Renato Parede. Análise de ruídos automotivos por parâmetros psicoacústicos. 2022. 111 p. Dissertação (Mestrado em Engenharia Elétrica) - Centro Universitário FEI, São Bernardo do Campo, 2022. Disponível em: https://doi.org/10.31414/EE.2022.D.131541.
Texto completo (DOI)
Palavras-chave
Ruídos parasitas,Ruídos automotivos,Psicoacústica
Resumo
Os ruídos parasitas em veículos novos/seminovos possuem um grande impacto na satisfação dos clientes brasileiros. A indústria automobilística já detectou esta particularidade de nosso mercado em que a presença destes ruídos é diretamente relacionada a qualidade/robustez dos produtos. Devida a esta importância as empresas vêm ampliando o investimento na detecção e caracterização nessa área, embora grande parte ainda usam o método tradicional através de rodagens com especialistas tanto na fase de projetos quanto durante a produção em massa ao longo do ciclo de vida do veículo. O presente estudo consiste em desenvolver uma metodologia para investigar a relação de parâmetros da psicoacústica com a irritabilidade dos clientes. A primeira etapa deste desenvolvimento foi a aquisição dos sinais
realizada diretamente sobre veículos nas pistas de teste do grupo Peugeot Citroën na cidade de Porto Real e da Fiat na cidade de Betim. Esta atividade consistiu na gravação de dez sinais sonoros dentre as interfaces que mais provocam insatisfação/irritação dos clientes. Posteriormente esses sinais foram utilizados em duas análises exploratórias, subjetiva e objetiva. A análise subjetiva foi realizada com a participação de 52 colaboradores sem conhecimentos específicos sobre o tema. Todo o controle acústico, térmico, assim como distância entre os voluntários e o equipamento de reprodução foram baseados em estudos científicos afins. A avaliação dos áudios pelo júri foi conduzida através de uma reprodução auto-guiada permitindo que cada participante escolhesse a sequência de áudio. A avaliação seguiu a escala métrica de subjetividade da SAE(Society of Automotive Engineers). Por sua vez a análise objetiva consistiu em extrair dos sinais de áudio os parâmetros da psicoacústica como Loudness, Roughness, Sharpness e Fluctuation Strenght. Nessa etapa utilizou-se o software TestLAB para a extração dos parâmetros. A análise de dados consistiu inicialmente no cruzamento dos dados obtidos do júri através da audição com as notas fornecidas pelos
especialistas do setor para cada um dos ruídos gravados. Como desenvolvimento final, um classificador sobre o impacto de cada ruído na satisfação dos clientes, usando o algoritmo de inteligência artificial ID3(Iterative Dichotomiser 3) foi implementado com o objetivo de reduzir
a subjetividade quando comparado aos sistemas tradicionais empregados atualmente no setor automobilístico. Os experimentos exploratórios desta dissertação indicam que há correlação
entre as métricas psicoacústicas com o nível de irritabilidade provocado pelos sinais assim como uma regra geral de classificação dos tipos de ruídos automotivos considerados
Abnormal noises in new/used cars have an important impact on Brazilian customer satisfaction. The automobile industry has already detected this particularity in our market where the presence of these noises is directly related to the quality/robustness of the products. Due to this importance, industries have been increasing investment in detection and characterization in this area, although most companies still use the traditional method through runs with specialists both in the project phase and in mass production during the cars’ life cycle. This study consists of a methodology development to investigate the relation of psychoacoustics parameters with the irritability of clients. The first stage of this experiment was the acquisition of the signals carried out directly from one vehicle on the test tracks of the Peugeot Citroën group in the city of Porto Real and Fiat in the city of Betim. This activity consisted of recording ten noise sounds among the interfaces most impactful on customer dissatisfaction/irritation. Later, these signals were used in two exploratory analysis, subjective and objective. The subjective analysis was carried out with the participation of 52 employees without specific knowledge on the subject. All room parameters as acoustic and thermal control as well as the distance between the volunteers and the reproduction equipment were based on related scientific studies. The audio jury was evaluated through a self-guided playback allowing each participant to choose an audio sequence. The subjectivity metric scale used on the assessment was based on SAE(Society of Automotive Engineers) parameters. The objective analysis consisted of extracting from the signals recorded the parameters of psychoacoustics such as Loudness (parameter correlated to sound intensity), Roughness (harshness), Sharpness (spectral sensitive) and Fluctuation Strength (low frequency modulation). In this phase, the TestLAB software was used to recover the parameters from the vehicle. The analysis initially consisted of confronting the juries by listening and data provided by industry experts for each of the recorded noises. As a final development, a classifier on the impact of each noise in customer satisfaction using an ID3(Iterative Dichotomiser 3) artificial intelligent algorithm was implemented with the aim of reducing subjectivity when compared to traditional systems currently used in the automotive sector. The exploratory experiments of this dissertation indicate that there is a correlation between the psychoacoustic metrics and the irritability of the signals as well as a general rule for classifying parasitic automotive noises
Abnormal noises in new/used cars have an important impact on Brazilian customer satisfaction. The automobile industry has already detected this particularity in our market where the presence of these noises is directly related to the quality/robustness of the products. Due to this importance, industries have been increasing investment in detection and characterization in this area, although most companies still use the traditional method through runs with specialists both in the project phase and in mass production during the cars’ life cycle. This study consists of a methodology development to investigate the relation of psychoacoustics parameters with the irritability of clients. The first stage of this experiment was the acquisition of the signals carried out directly from one vehicle on the test tracks of the Peugeot Citroën group in the city of Porto Real and Fiat in the city of Betim. This activity consisted of recording ten noise sounds among the interfaces most impactful on customer dissatisfaction/irritation. Later, these signals were used in two exploratory analysis, subjective and objective. The subjective analysis was carried out with the participation of 52 employees without specific knowledge on the subject. All room parameters as acoustic and thermal control as well as the distance between the volunteers and the reproduction equipment were based on related scientific studies. The audio jury was evaluated through a self-guided playback allowing each participant to choose an audio sequence. The subjectivity metric scale used on the assessment was based on SAE(Society of Automotive Engineers) parameters. The objective analysis consisted of extracting from the signals recorded the parameters of psychoacoustics such as Loudness (parameter correlated to sound intensity), Roughness (harshness), Sharpness (spectral sensitive) and Fluctuation Strength (low frequency modulation). In this phase, the TestLAB software was used to recover the parameters from the vehicle. The analysis initially consisted of confronting the juries by listening and data provided by industry experts for each of the recorded noises. As a final development, a classifier on the impact of each noise in customer satisfaction using an ID3(Iterative Dichotomiser 3) artificial intelligent algorithm was implemented with the aim of reducing subjectivity when compared to traditional systems currently used in the automotive sector. The exploratory experiments of this dissertation indicate that there is a correlation between the psychoacoustic metrics and the irritability of the signals as well as a general rule for classifying parasitic automotive noises