Application of an Improved SVM Algorithm for Wireless Sensor Networks in the Prediction of Air Pollution

As is known to all, human's daily life can not be separated from the atmospheric environment, and the quality of human life is directly affected by the quality of the atmospheric environment. In recent decades, the rapid development of China's economy, industry, transportation and other industries cause that concentration of the hazardous material in the air is much higher than the identification standard. It not only affects everyone's life and production, but also hinders the development of the whole society. Based on this, with the help of the wireless sensor technology, this paper builds a wireless sensor networks for the acquisition and transmission of various air pollutants data. This network can effectively monitor the current existence of most of the pollutant gas. Secondly, according to the received data, this paper introduces the support vector regression model based on ant colony optimization to forecast the concentration of pollutants in the air. Because the prediction accuracy of support vector machine is largely determined by the selection of parameters, the selection of training parameters is optimized by using ant colony algorithm in order to get the optimized support vector machine prediction model. At last, we use the modified model to predict the concentration of PM2.5 with nonlinear data. Experimental results show that the proposed improved support vector prediction algorithm is effective, and is significantly better than the other two prediction algorithm.