In this paper,the statistical model research of the main air pollution prediction in Shijiazhuang was introduced,including the theory of the regression analysis,prediction selection factors,the regression equation of the main air pollution and its correlative coefficient,the model popularization and applications.

An attempt is made to establish three statistical models for vertical distribution of atmospheric refractive index based on the theory and approach with the aid of conventional radiosonde record from the selected 20 stations in China.

The general law of layer’s predtictivity is investigated in an agrometeorological con text,whereupon is constrticted a statistical model for the variation in the physiological tendericyversus time and meteorological production as a function of temperature-humidity index(THI).

A new propagation prediction method for urban microcellular systems is presented based on measurement data and statistic model.

Structural Abnormality Detection Using Statistical Pattern Recognition Technique;

The basic theory of statistical system identification approach (Bayesian model updating and stochastic finite element model updating), statistical pattern recognition approach and probabilistic neural network approach is introduced, and the statistical identification for structural health diagnosis in civil engineering is discussed comprehensively.

A supervision and sorting technique on the basis of statistical pattern recognition by using color as characteristic space was proposed according to the properties of plant disease image.

Two-dimensional Euler model was used to simulate long-range transport and se.