Time Series Analysis And Prediction Of Insects Plague

The species and amount of insect is the largest in the animal kingdom. They have a significant influence on agricultural production and human health. Pest control is the key to increase crop yields, whether pest control is successful shows the management and technical levels of agricultural economy. It provides effective information for the managers and agrotechnical personnel.In this paper, the predictions of pest plague are studied using two different methods. The first one: the stationary process of the data for the insect plague area and afflicted area is studied using time series analysis. On the basis of the above, ARMA model of insect plague area and afflicted area are proposed respectively, next the order of the models is determined and the corresponding parameter are estimated, testing the reasonableness of the model, and then the predictions of the insect plague area and afflicted area are investigated. The Second one: combined ARMA time series with Kalman filtering theory, the predictions of insect plague area and afflicted area are studied by Kalman filtering.For the same time series, comparing data analysis with the result of prediction using two different methods, the result is obtained: The latter is better than the former. The main reason is as follows: the Kalman filtering equations are composed of a lot of iterated formulas. The corresponding calculations are predicted and revised continuously. When the new observation data are got, the new filtering value is also obtained to revise the observed consequence and get better prediction results.