Statistical analysis of wind load probabilistic models considering wind direction and calculation of reference wind pressure values in Liaoning Province, China
Jiaxu Li,Ming Liu,Xu Yan,Qianting Yang
DOI: https://doi.org/10.1016/j.probengmech.2023.103550
IF: 2.954
2024-01-01
Probabilistic Engineering Mechanics
Abstract:Wind pressure serves as the fundamental base for architectural design, particularly the reference wind pressure value, which affects the reliability and economy of high-rise, towering, and large-span structures. To facilitate the calculation of reference wind pressure values for design reference periods, load codes from various countries offer probabilistic distribution models for wind speed or reference wind pressure. However, the application of a uniform load probabilistic distribution model for calculating the reference wind pressure for the reference period often results in overestimation or underestimation of the reference wind pressure. To this end, this study collected annual maximum wind speed data from 34 meteorological stations in Liaoning Province, a region in Northeast China, and established extreme value type I, extreme value type III, lognormal, and generalized extreme value distribution models for the annual maximum wind speeds using the maximum likelihood method and method of moment, respectively. The goodness-of-fit test was conducted utilizing the Kolmogorov–Smirnov method. The results showed that the extreme value type III model and the method of moment achieved the best goodness-of-fit for the probabilistic distribution of annual maximum wind speeds. Based on the measured wind direction data corresponding to the annual maximum wind speed, joint probability density functions and cumulative probability density functions of wind speed and wind direction were established. Finally, the influence of different load statistical periods on the reference wind pressure was analyzed.
engineering, mechanical,mechanics,statistics & probability