Based On Rough Set And Neural Network Inthe Subway Platform Air Quality Comprehensive Evaluation Research

With the improvement of comprehensive national power and economic and social rapid development,the pace of urban life is also accelerating.With its fast and convenient characteristics,MTR has developed rapidly in China and has become the preferred mode of transportation for people to work.More and more people take the subway,people in the subway station stagnation time increases,but also to the subway station air quality problems are increasingly exposed: air pollution,lack of ventilation,particulate pollution,noise pollution and so on.In this context,people gradually began to pay attention to the subway station air quality,but also greatly aroused the domestic and foreign experts and scholars on the subway station air quality comprehensive evaluation research.In this paper,the air quality problem in the platform area of the subway system is studied,and the four stations of Beijing: Xizhimen Station,Fuxingmen Station,Lingjing Hutong Station and Andri Beijie Station are selected.The temperature,relative humidity,wind speed,noise CO₂,Formaldehyde,TVOC and PM₁₀.The test period was from 7:00 to 9:00,from 13:00 to 15:00 and 17:00 to 19:00,The statistic and analysis of the test data are carried out,and the main influencing factors of the station air quality index are explored.Based on the statistical analysis of the objective measured data,this paper,based on the combination of rough set theory and BP neural network theory,builds a comprehensive evaluation model of the air quality of the subway platform with MATLAB platform.It was applied to the comprehensive evaluation of the air quality of the station,and the classification of the station air quality was finally obtained.Combining with the attribute value and the evaluation result after discretization,the influence of each index on air quality grade evaluation of subway station is as follows: temperature,relative humidity,noise,PM₁₀,wind speed and CO₂.And the evaluation results are analyzed,proving the rationality of the model.This is of great theoretical and guiding significance for establishing a practical air quality evaluation system and improving the air quality of metro.The following conclusions were drawn during the testing and research process:?1?The average relative humidity,wind speed,CO₂ concentration,formaldehyde and TVOC of the four sites are below the recommended standard range,the average noise level exceeds the recommended standard,and the temperature,PM₁₀ concentration is only met at the non-transfer station Recommended standards.?2?Passenger volume and PSD are the most important factors influencing the air quality of subway platform.The air temperature,noise,CO₂ concentration and PM₁₀ concentration were significantly higher in the subway stations where the traffic volume was relatively large.than those in non-transfer stations with small passenger flow rate.On the subway platforms with PSDs,wind speed,noise,PM₁₀ concentration Lower than the subway platform without PSD.?3?The air temperature,relative humidity and wind speed of the subway platform are relatively stable in the day.The heat dissipation of the personnel and equipment is the main influencing fa ctor of the air temperature of the subway platform,and the humidity dissipation of the personnel,temperature and outdoor climate of the subway station can not be ignored."Piston wind" and whether the PSD have a greater impact on the site wind speed.?4?The noise,CO₂ concentration,formaldehyde,TVOC and PM₁₀ concentration showed a tendency to change with time in the daytime.The noise,CO₂ concentration and PM₁₀ concentration were higher in the early and late peak periods than those in the off-peak periods,while the concentration of formaldehyde and TVOC decreased with time.