Related Experiment Videos
A neural network based model for urban noise prediction.
N Genaro1, A Torija, A Ramos-Ridao
1Department of Computer Science and Artificial Intelligence, University of Granada, 18071 Granada, Spain. nataliag@ugr.es
The Journal of the Acoustical Society of America
|October 26, 2010
Summary
This study introduces a new model for predicting urban noise pollution using Artificial Neural Networks (ANN). The ANN model offers greater accuracy than previous methods, aiding urban planners in assessing acoustic environmental risks.
Area of Science:
- Environmental Science
- Urban Planning
- Computational Intelligence
Background:
- Noise pollution is a recognized global environmental issue, classified as a pollutant by the World Health Organization (WHO) in 1972.
- Industrialized nations have implemented regulations to mitigate acoustic pollution, yet effective tools for urban planners to assess noise levels are still needed.
- Existing urban noise modeling approaches have yielded suboptimal results, highlighting the need for improved predictive capabilities.
Purpose of the Study:
- To develop and evaluate a novel model for predicting environmental urban noise.
- To assess the efficacy of Soft Computing techniques, specifically Artificial Neural Networks (ANN), in urban noise prediction.
- To compare the performance of the ANN model against existing urban noise models.
Main Methods:
- Development of an Artificial Neural Network (ANN) model incorporating variables identified by experts as influential in urban noise.
- Application of the ANN model to real-world data collected from various street types.
- Utilized Principal Component Analysis (PCA) to explore model simplification and its impact on accuracy.
Main Results:
- The Artificial Neural Network (ANN) system demonstrated superior accuracy in predicting urban noise compared to other evaluated models.
- The ANN model provides a significant improvement for assessing and managing acoustic environmental pollution.
- Principal Component Analysis (PCA) offered a slight reduction in accuracy but yielded acceptable results for model simplification.
Conclusions:
- Artificial Neural Networks (ANN) present a highly effective tool for accurate urban noise prediction.
- The developed ANN model can assist urban planners in better evaluating and mitigating acoustic environmental pollution.
- Model simplification using PCA is a viable option when a slight trade-off in accuracy is acceptable.