Antonio J. Torija

Application of a methodology for categorizing and differentiating urban soundscapes using acoustical descriptors and semantic-differential attributes

A subjective and physical categorization of an ambient sound is the first step to evaluate the soundscape and provides a basis for designing or adapting this ambient sound to match peoples expectations. For this reason, the main goal of this work is to develop a categorization and differentiation analysis of soundscapes on the basis of acoustical and perceptual variables. A hierarchical cluster analysis, using 15 semantic-differential attributes and acoustical descriptors to include an equivalent sound-pressure level, maximum-minimum sound-pressure level, impulsiveness of the sound-pressure level, sound-pressure level time course, and spectral composition, was conducted to classify soundscapes into different typologies. This analysis identified 15 different soundscape typologies. Furthermore, based on a discriminant analysis the acoustical descriptors, the crest factor (impulsiveness of the sound-pressure level), and the sound level at 125 Hz were found to be the acoustical variables with the highest impact in the differentiation of the recognized types of soundscapes. Finally, to determine how the different soundscape typologies differed from each other, both subjectively and acoustically, a study was performed.

A neural network based model for urban noise prediction.

Noise is a global problem. In 1972 the World Health Organization (WHO) classified noise as a pollutant. Since then, most industrialized countries have enacted laws and local regulations to prevent and reduce acoustic environmental pollution. A further aim is to alert people to the dangers of this type of pollution. In this context, urban planners need to have tools that allow them to evaluate the degree of acoustic pollution. Scientists in many countries have modeled urban noise, using a wide range of approaches, but their results have not been as good as expected. This paper describes a model developed for the prediction of environmental urban noise using Soft Computing techniques, namely Artificial Neural Networks (ANN). The model is based on the analysis of variables regarded as influential by experts in the field and was applied to data collected on different types of streets. The results were compared to those obtained with other models. The study found that the ANN system was able to predict urban noise with greater accuracy, and thus, was an improvement over those models. The principal component analysis (PCA) was also used to try to simplify the model. Although there was a slight decline in the accuracy of the results, the values obtained were also quite acceptable.

Using Recorded Sound Spectra Profile as Input Data for Real-Time Short-Term Urban Road-Traffic-Flow Estimation

Road traffic has a heavy impact on the urban sound environment, constituting the main source of noise and widely dominating its spectral composition. In this context, our research investigates the use of recorded sound spectra as input data for the development of real-time short-term road traffic flow estimation models. For this, a series of models based on the use of Multilayer Perceptron Neural Networks, multiple linear regression, and the Fisher linear discriminant were implemented to estimate road traffic flow as well as to classify it according to the composition of heavy vehicles and motorcycles/mopeds. In view of the results, the use of the 50-400 Hz and 1-2.5 kHz frequency ranges as input variables in multilayer perceptron-based models successfully estimated urban road traffic flow with an average percentage of explained variance equal to 86%, while the classification of the urban road traffic flow gave an average success rate of 96.1%.

A tool for urban soundscape evaluation applying Support Vector Machines for developing a soundscape classification model

To ensure appropriate soundscape management in urban environments, the urban-planning authorities need a range of tools that enable such a task to be performed. An essential step during the management of urban areas from a sound standpoint should be the evaluation of the soundscape in such an area. In this sense, it has been widely acknowledged that a subjective and acoustical categorization of a soundscape is the first step to evaluate it, providing a basis for designing or adapting it to match peoples expectations as well. In this sense, this work proposes a model for automatic classification of urban soundscapes. This model is intended for the automatic classification of urban soundscapes based on underlying acoustical and perceptual criteria. Thus, this classification model is proposed to be used as a tool for a comprehensive urban soundscape evaluation. Because of the great complexity associated with the problem, two machine learning techniques, Support Vector Machines (SVM) and Support Vector Machines trained with Sequential Minimal Optimization (SMO), are implemented in developing model classification. The results indicate that the SMO model outperforms the SVM model in the specific task of soundscape classification. With the implementation of the SMO algorithm, the classification model achieves an outstanding performance (91.3% of instances correctly classified).

Modeling Environmental Noise Using Artificial Neural Networks

Since 1972, when the World Health Organization (WHO) classified noise as a pollutant, most industrialized countries have enacted laws or local regulations that regulate noise levels. Many scientists have tried to model urban noise, but the results have not been as good as expected because of the reduced number of variables. This paper describes artificial neural networks (ANN) to model urban noise. This model was applied to data collected at different street locations in Granada, Spain. The results were compared to those obtained with mathematical models. It was found that the ANN system was able to predict noise with greater accuracy, and therefore it was an improvement on these models. Furthermore, this paper reviews literature describing other research studies that also used soft computing techniques to model urban noise.

The subjective effect of low frequency content in road traffic noise

Based on subjective listening trials, Torija and Flindell [J. Acoust. Soc. Am. 135, 1-4 (2014)] observed that low frequency content in typical urban main road traffic noise appeared to make a smaller contribution to reported annoyance than might be inferred from its objective or physical dominance. This paper reports a more detailed study which was aimed at (i) identifying the difference in sound levels at which low frequency content becomes subjectively dominant over mid and high frequency content and (ii) investigating the relationship between loudness and annoyance under conditions where low frequency content is relatively more dominant, such as indoors where mid and high frequency content is reduced. The results suggested that differences of at least +30 dB between the low frequency and the mid/high frequency content are needed for changes in low frequency content to have as much subjective effect as equivalent changes in mid and high frequency content. This suggests that common criticisms of the A-frequency weighting based on a hypothesized excessive downweighting of the low frequency content may be relatively unfounded in this application area.

A general procedure to generate models for urban environmental-noise pollution using feature selection and machine learning methods.

The prediction of environmental noise in urban environments requires the solution of a complex and non-linear problem, since there are complex relationships among the multitude of variables involved in the characterization and modelling of environmental noise and environmental-noise magnitudes. Moreover, the inclusion of the great spatial heterogeneity characteristic of urban environments seems to be essential in order to achieve an accurate environmental-noise prediction in cities. This problem is addressed in this paper, where a procedure based on feature-selection techniques and machine-learning regression methods is proposed and applied to this environmental problem. Three machine-learning regression methods, which are considered very robust in solving non-linear problems, are used to estimate the energy-equivalent sound-pressure level descriptor (LAeq). These three methods are: (i) multilayer perceptron (MLP), (ii) sequential minimal optimisation (SMO), and (iii) Gaussian processes for regression (GPR). In addition, because of the high number of input variables involved in environmental-noise modelling and estimation in urban environments, which make LAeq prediction models quite complex and costly in terms of time and resources for application to real situations, three different techniques are used to approach feature selection or data reduction. The feature-selection techniques used are: (i) correlation-based feature-subset selection (CFS), (ii) wrapper for feature-subset selection (WFS), and the data reduction technique is principal-component analysis (PCA). The subsequent analysis leads to a proposal of different schemes, depending on the needs regarding data collection and accuracy. The use of WFS as the feature-selection technique with the implementation of SMO or GPR as regression algorithm provides the best LAeq estimation (R(2)=0.94 and mean absolute error (MAE)=1.14-1.16 dB(A)).

A model for the rapid assessment of the impact of aviation noise near airports

This paper introduces a simplified model [Rapid Aviation Noise Evaluator (RANE)] for the calculation of aviation noise within the context of multi-disciplinary strategic environmental assessment where input data are both limited and constrained by compatibility requirements against other disciplines. RANE relies upon the concept of noise cylinders around defined flight-tracks with the Noise Radius determined from publicly available Noise-Power-Distance curves rather than the computationally intensive multiple point-to-point grid calculation with subsequent ISO-contour interpolation methods adopted in the FAAs Integrated Noise Model (INM) and similar models. Preliminary results indicate that for simple single runway scenarios, changes in airport noise contour areas can be estimated with minimal uncertainty compared against grid-point calculation methods such as INM. In situations where such outputs are all that is required for preliminary strategic environmental assessment, there are considerable benefits in reduced input data and computation requirements. Further development of the noise-cylinder-based model (such as the incorporation of lateral attenuation, engine-installation-effects or horizontal track dispersion via the assumption of more complex noise surfaces formed around the flight-track) will allow for more complex assessment to be carried out. RANE is intended to be incorporated into technology evaluators for the noise impact assessment of novel aircraft concepts.

Differences in subjective loudness and annoyance depending on the road traffic noise spectrum.

There is at present no consensus about the relative importance of low frequency content in urban road traffic noise. The hypothesis underlying this research is that changes to different parts of the spectrum will have different effects depending on which part of the spectrum is subjectively dominant in any particular situation. This letter reports a simple listening experiment which demonstrates this effect using typical urban main road traffic noise in which the low frequency content is physically dominant without necessarily being subjectively dominant.

Framework for Predicting Noise–Power–Distance Curves for Novel Aircraft Designs

Along with flight profiles, noise–power–distance curves are the key input variable for computing noise exposure contour maps around airports. With the development of novel aircraft designs (incorporating noise-reduction technologies) and new noise-abatement procedures, noise–power–distance datasets will be required for assessing their potential benefit in terms of noise reduction around airports. Noise–power–distance curves are derived from aircraft flyover noise measurements taken for a range of aircraft configurations and engine power settings. Clearly then, empirical noise–power–distance curves will be unavailable for novel aircraft designs and novel operations. This paper presents a generic framework for computationally generating noise–power–distance curves for novel aircraft and situations. The new framework derives computationally the noise–power–distance noise levels that are normally derived experimentally, by estimating noise level variations arising from technological and operational changes with respect to a baseline scenario, where the noise levels are known or otherwise estimated. The framework is independent of specific prediction methods and can use any potential new model for existing or new noise sources. The paper demonstrates the methodology of the framework, discusses its benefits, and illustrates its applicability by deriving noise–power–distance curves for an unconventional approach operation and for a future concept blended wing–body aircraft.