| With the rapid development of cities,road traffic and rail transit construction is accelerated,but traffic noise also follows.Long-term exposure to traffic noise can lead to annoyance,sleep disturbance,hypertension,cardiovascular disease,and cognitive impairment.To improve residents’life quality and happiness,it is urgent to control noise pollution caused by urban traffic as soon as possible.The evaluation and prediction of urban traffic noise can realize the regulation and prevention of existing noise problems and further predict the possible events of planned lines to reserve the implementation space for treatment in advance.Traffic noise is closely related to urban road conditions and the traffic environment.Most studies on traffic noise are focused on plain cities,while few factors such as roadway gradient and railway bridge are considered in mountainous cities.As a mountainous city,Chongqing is characterized by various road alignments and complex traffic systems.The types and characteristics of traffic noise sources are different from those of plain cities.This paper considers two kinds of noise(road traffic noise and rail transit noise)that seriously affect peoples’daily lives.Noise prediction models are used to analyze the characteristics of traffic noise in mountainous cities in series.The road traffic noise prediction model is developed by introducing roadway gradient,while the rail transit noise prediction model is focused on the railway bridge.The main research work and results are as follows:(1)The sound source characteristics of urban road traffic noise are analyzed by revealing the mechanism of a microscopic sound source.A noise monitoring test of a full-scale test track(RIOHTRACK)is carried out by the Research Institute of Highway Ministry,which adopts the close-proximity(CPX)trailer method.A multilevel Bayesian regression model is established based on the field data to reveal the relationship between key variables and tire/pavement noise.The influence of different factors on tire/pavement noise is obtained.There is a significant positive correlation between vehicle speed and tire/pavement noise.Tire type and pavement type have a grading effect for the tire/pavement noise.At higher speed,the sound absorption and noise reduction characteristics of porous asphalt pavement are reflected.(2)According to the characteristics of roadway gradient in mountainous cities,traffic noise monitoring tests in different gradient intervals are carried out.Based on an artificial neural network,the prediction model of per-vehicle radiated noise and traffic flow noise in wide gradient intervals are constructed,which is helpful to realize the prediction,control,and management of road traffic noise in mountainous cities.For the test set,R~2 and MAE of the well-fitted per-vehicle radiation model are 0.8688 and 1.73d B,respectively;R~2 and MAE of the well-fitted traffic flow noise radiation model are0.8218 and 0.73 d B,respectively.(3)Aiming at the mountainous city rail transit system,taking the railway bridge(Dongshuimen Yangtze River Bridge)as the research object,based on the bridge parameters and modal measurement results,ANSYS software is used to establish the dynamic model of the railway bridge,wheel,and rail under moving load.The response time history of vibration displacement,velocity,and acceleration in bridge side span is calculated by the transient dynamic analysis method.The response spectrum of vibration displacement,velocity,and acceleration of wheel and rail is obtained by the harmonic response analysis method.(4)According to the causality between vibration and noise,the calculated structural vibration response results are input as acoustic boundary conditions into LMS Virtual.Lab Acoustics simulation software.The acoustic finite element method is used to realize the numerical simulation of acoustic radiation in the frequency domain.The acoustic boundary element method is used to realize the numerical simulation of acoustic radiation in the time domain.Based on the simulation results of the established vehicle-rail-bridge system model,the acoustic characteristics,spatial and temporal distribution of acoustic radiation of bridge structural noise,and wheel-rail rolling noise of railway bridge are analyzed.The acoustic energy of bridge-borne is mainly concentrated at 31.5 Hz,80 Hz,and 125 Hz.Wheel-rail noise acoustic energy is mainly concentrated at 250 Hz,400 Hz,1250 Hz,and 2000 Hz.Vibration and noise reduction measures can be mainly taken for these peak frequencies.(5)Field data of bridge structure noise and wheel/rail noise are obtained by setting measuring points near the railway bridge wheel/rail system and sensitive areas around the bridge.The difference correction formulation is constructed by using a cubic spline curve to correct the simulation results of the established vehicle-railway-bridge system model and the actual measured values.Based on the modified simulation results of the vehicle-railway-bridge system,a railway bridge noise spatial prediction model is established,and the applicability of the model is evaluated by another validation test.The field points near the wheel-rail contact surface can better predict the wheel-rail noise,and the field points underneath the bridge can better predict the bridge-borne noise.High-frequency noise decays faster than low-frequency noise.The difference between the simulation results and the measured results of 50%data is less than 2 d B.The percentage error range is from 0.17%to 6.50%.The model can evaluate and predict the noise level of typical rail transit systems in mountainous cities.This paper established the tire/pavement noise prediction model,per-vehicle emission noise prediction model,traffic noise prediction model,and railway bridge noise spatial prediction model.All of these models will be beneficial to realize the noise assessment to the complex traffic environment in mountainous cities and the noise events prediction of planning lines,which can provide references for relevant departments to control traffic noise pollution. |
PDF Full Text Request