Research On Noise Prediction And Visualization Of Urban Road Traffic Based On Parameter Learning

With the advance of urbanization and the expansion of road transportation network, noise problem of road traffic caused by motor vehicle in city has become increasingly prominent, it not only impacts on the life and work of nearby residents, but also obstructs the booming development of road transportation. Therefore, it's very urgent to scientifically and effectively predict the traffic noise and its impact on the surrounding area of urban road.Firstly, it analyzes the evaluation indexes and their scope of the traffic noise about urban road. Besides, it describes the characteristics, hazards and factors of road traffic noise,presents the advantages, disadvantages and scope of the usual evaluation indexes, and ultimately selects the LAeq(equivalent continuous A-weighted sound level) as the evaluation index of this article. Also, it illustrates the national criteria for evaluating the road traffic noise.Secondly, it improves FHWA(Federal Highway Administrator) model. It describes the predicting models widely used in the world, and analyzes the influence factors and applicable conditions of the model. By combing the characteristics of urban road traffic in our country, it selects FHWA model to predict noise, compares and analyses the results so as to correct the model based on the actual monitoring.Thirdly, it proposes predicting model of traffic noise based on parameter learning. Based on the improved model, it introduces the concept of parameter learning, uses machine learning methods to train the model, which can correct the model parameters by using its abilities of self-learning and self-organization. After analyzing the experimental results, it shows that the model has high accuracy and practicality.Finally, it uses GIS(geographic information system) software to draw traffic noise maps of road around Olympic Sports Center and Wutai Mountain Stadium located in Nanjing city.It researches GIS and its application in the field of noise predicting about road traffic, and illustrates the process of drawing noise map. Using the data obtained by above model,combining geographic information basic data of regional traffic, it draws traffic noise map of the target area, which can visually show the distribution of traffic noise surrounding urban road.