Study On Correction Of Noise Prediction Model And Optimization Of Sound Barrier For Highway Mainly To Large Vehicles

The rapid development of highway construction has greatly promoted the development of national economy. Meanwhile, it has also brought along noise pollution to different levels. Highway traffic noise has serious effect on the villages, schools and other companies along the roads. How to predict the highway traffic noise more accurately and take effective measures to reduce noise has attracted people’s great attention.Shandong province, located in Bohai Sea economic zone, has started its highway construction early and developed rapidly. Because of its location by the peninsula area, it has numerous coastal ports and large vehicles especially heavy vehicles. Compared to the mainland, Shandong province has some special characteristics of transportation. In this paper, the characteristics of highway traffic noise in Shandong province was studied on the basis of noise spot monitoring and the prediction model of traffic noise mainly to large vehicles was corrected. Furthermore, efficiency of typical highway sound barrier and optimization design of sound barrier was conducted.As shown in the research of the characteristics of traffic noise in Shandong province, traffic flow of large vehicles occupies a higher percentage, while there is no much diffenence of highway traffic noise between day and night. After further analysis and monitoring, it can be found that large vehicles traffic noise plays a dominant role in highway traffic noise of Shandong province. Because of the stability and relatively high level of large vehicles, the traffic noise of Shandong province presents a situation of little change between day and night. This phenomenon was also verified by the further study of characteristics of traffic noise on holidays. In view of this feature, typical highways in Shandong province were chosen as models to monitor and analyze their noise spectrums and equivalent frequency. The results show that the equivalent frequency is lower than 500 Hz, mainly between 200 Hz and 500 Hz. With the modification of equivalent frequency formula, the center frequency value are set as 200 Hz, 250 Hz, 315 Hz, 400 Hz, 500 Hz, 630 Hz and 800 Hz, the highway traffic noise equivalent frequency value is near 315 Hz.According to the highway traffic characteristics of Shandong province, typical highways were chosen for the spot monitoring to evaluate the accuracy of the prediction model. The results show that accuracy of prediction model in guiline is relatively high while still has certain deviation. The fitting degree of noise prediction model of single large vehicle is low while traffic noise of large vehicles plays a dominant role in Shandong province. The calculated value is lower when the car speed is low and higher when the car speed is high, which lead to the noise prediction deviation. Aiming at this defect, the correction research of sound intensity prediction model of single large vehicle was conducted. In reference to the existing formula fitting process and the characteristics of highway traffic flow in Shandong province, large vehicles were divided into four axles(including 1, 2 and 3 axles), large buses and 6 axles(including 5 axles). Calculation formulas of noise intensity of single vehicle were fitted on spot monitoring data. Through the practical verification, it shows that the prediction accuracy of highway traffic noise could be effectively improved after the detailed classification of large vehicles. Specifically, the improvement accuracy is turned to be 53.4%~66.5% in daytime and 13.7~19.0% at nighttime. In addition, correlation of traffic flow and average speed by using the mixed model of automatic encoder and BP neural network was conducted. The selected typical sample of monitoring data in Shandong province highway indicates the daytime speed error is only of 5.3%~24.0% while nighttime of 12.2%~29.7%. It can basically meet the actual requirement that can be recommended as speed prediction model of highway mainly to large vehicles.Six sound barriers-in operation phase were chosen to evaluate their noise reduction efficiency. The low reduction effect of sound barrier could be assigned to the prediction deviation of traffic flow at design stage, inaccuracy of traffic noise prediction model and the ground attenuation reduction after sound barrier construction. Traditional sound barriers mainly for traffic noise frequency of 500 Hz show lower efficiency due to the low frequency range of highway traffic noise. Combined with the ground attenuation reduction and equivalent frequency of 315 Hz, the cavity resonance theory was used in the optimization design of sound barriers in highway mainly to large vehicles. Good noise reduction efficiency was achieved when optimization design of sound barriers was used in Qingzhou to Linshu highway.