Research On Air Quality Assessment And Prediction Based On Clustering And Deep Learning

Today,with the rapid development of artificial intelligence,Deep learning is widely used in all aspects of social life.There are many factors affecting air quality,such as wind direction,wind speed,temperature,humidity and pollution source emissions.Therefore,the evaluation of air quality will have great uncertainty and the deep learning method is very suitable for research.Such objects with varying characteristics.In order to accurately assess air quality,a more scientific grading of each level of existing air quality levels,a more objective response to air quality conditions,to help us better understand the daily air conditions,this study proposes a Air quality assessment and prediction methods for class algorithms and deep learning.The data required for the study are the six main indicators of urban air pollutant AQI:sulfur dioxide(SO2),carbon monoxide(CO),nitrogen dioxide(NO2),ozone(O3),particulate matter(particle size≤10 microns),fine particulate matter(Respirable particulate matter≤2.5 microns).All data were obtained from the China Air Quality Online Monitoring Platform.These obtained air quality data are not artificially calibrated.To train various neural networks of the deep learning model,it is necessary to artificially cluster.First,cluster the unlabeled samples and calibrate the labels.Secondly,the tags are automatically matched for the uncalibrated samples;then the acquired tag samples are input into the neural network model and trained to evaluate and predict air quality.In this paper,several clustering and deep learning methods are used to evaluate and predict air quality.The first method to evaluate air quality is the improved K-means clustering method.The k-means algorithm and the layering algorithm are combined to solve the K-means aggregation.The initial clustering point selection sensitivity is also easy to fall into the local solution and other shortcomings,while having fewer iterations and visualizing the air quality results,providing more comprehensive air quality information.Air quality can then be assessed by AP clustering and VAE models,which can save a lot of computation time and more accurate understanding of real-time air quality.Finally,the LSTM model is used to predict air quality.