Research On Urban Noise Prediction And Visualization

In recent years,urban noise pollution has become a core issue in urban construction.In order to ensure that people have a good environment for work,study and rest,and to build a better green city,it is extremely important to effectively and scientifically predict urban noise.However,the traditional method of noise prediction ignores the relationship between urban areas and the nonlinear characteristics of noise data during the analysis and prediction of noise,making the accuracy of the predicted result value not very high,and there is no good Noise visual analysis method.In this paper,based on the characteristics of urban noise data,a tensor decomposition model based on kernel least squares is proposed and combined with visualization methods for analysis.The main research contents are as follows:First,this paper analyzes and calculates the similarity between noise types and the correlation between urban areas.Divide the noise data according to the proportion of different types of noise,combine the data characteristics of the noise type,calculate the similarity between the noise types based on the kernel function,and then divide the urban area based on the main street of the city,Based on the nuclear distance to calculate the correlation between urban areas.Secondly,an urban noise prediction model based on the kernel least squares tensor decomposition model is proposed.Introduce road network features,user punching and POIs related data information features,combined with the smoothness and spatiotemporal characteristics of urban noise data itself,combine the kernel least squares method with tensor decomposition method,and combine the noise type The correlation calculation algorithm and urban area correlation algorithm are introduced,and a kernel least squares tensor decomposition model is proposed to predict noise data.Thirdly,according to the characteristics of noise data,a visual analysis strategy for multi-angle analysis is proposed.Since urban noise data has the time and latitude and longitude locations of noise,visual analysis is carried out from the time dimension and the space dimension,and the time dimension and the space dimension are combined for analysis.Furthermore,there are different types of noise.The similar algorithm of the noise type is visually analyzed from the perspective of the similarity between the noise types.Finally,the prediction model of this paper is evaluated,and combined with the visual analysis strategy of this paper to visualize the experimental results.The comparative analysis of the experimental results proves the efficiency of the noise prediction model in this paper,and then combined with the actual situation,the visual analysis strategy of this paper is used to visually analyze the experimental results.