Method for Evaluating the Statistical Relationship between Sound Pressure Level and Noise Annoy- ance Based on a Nonlinear Time Series Regression Model and an Experiment

Identifying the quantitative relationship between the sound pressure level and noise annoyance for environmental noises is important from the viewpoint of noise assessment. In this study, a method for predicting the probabilistic evaluation quantities like Lx ((100−x) percentile level) and LAeq (equivalent A-weighted sound pressure level) of the noise environment is proposed by introducing a nonlinear time series regression model between the sound pressure level and noise annoyance. More specifically, the joint probability distribution is expanded in an orthonormal expansion series in which linear and nonlinear correlation information is reflected hierarchically in each expansion coefficient. Next, statistical methods for predicting the sound pressure level and the noise annoyance are proposed by introducing a nonlinear time series regression model based on the above probability distribution. The validity of the proposed method is confirmed by applying it to a set of instantaneous data on sound pressure level and noise annoyance observed in a real sound environment.