The maximum wind velocity of automatic weather station (AWS) in Caofeidian Industrial Zone from 2007 to 2008 and the maximum wind velocity of meteorological stations at Tanghai, Laoting and Luannan from 1972 to 2008 were resolved into u and v components, and the regression models of u and v components were established respectively. Using the maximum wind speed data in Tanghai, Laoting and Luannan from 1972 to 2006, the data series of maximum wind speed in Caofeidian Industrial Zone during 1972-2006 were rebuilt. The observing and rebuilding data in Caofeidian were compared with its observing data in Tanghai, and the results show that the correlation coefficient of the maximum wind speed between Tanghai and Caofeidian is quite steady and there are some differences in wind direction between the two stations. When the wind speed is slower, the difference of wind direction between the two stations is larger. When the wind speed is larger, the difference of wind direction is smaller. The observed data of Caofeidian and Tanghai show that the times of same bearing are 28.3 percent in spring, 31.5 percent in summer, 24.7 percent in autumn, and 33.1 percent in winter. The rebuilding data of Caofeidian and the observed data in Tanghai show that the times of same bearing are 46 percent in spring, 29.4 percent in summer, 42.7 percent in autumn, and 54 percent in winter. In order to investigate the reality of the rebuilding data， we take 120 cases of high wind in different months as examples and analyze the relationship between rebuilding data and weather system. In the 120 cases, 114 cases meet obvious weather systems and 5 cases do not. The probability distribution of rebuilding maximum wind speed is obvious. Through hypothesis test, the daily maximum wind speed in Caofeidian follows a Gamma distribution. The series of maximum wind speed in Caofeidian can be used for the analysis and evaluation of wind disasters.