Evaluation indicator and models of time-varying traffic noise based on human feeling

Effects on human vary between uniform and time-varying traffic noise with the same equivalent sound level. Firstly, the subjective evaluation experiment of time-varying traffic noise is designed and implemented in this paper to get truthful data. Secondly, taking the gender, employment and mood factors into account, some conclusions about the effects of traffic noise with different time-varying characteristics on human are obtained based on the test results. Then the paper defines the disturbance degree of traffic noise on human as the subjective indicator of noise evaluation, represented by L_m. Finally, taking multiple existing noise indicators as inputs and L_m as output, three mathematical models are established using multiple regression, stepwise regression and artificial neural network analysis respectively.