Application of metaheuristic optimization algorithms to evaluate the geometric and kinematic parameters of downbursts. (November 2022)

• Record Type: Journal Article
• Authors: Xhelaj, Andi; Burlando, Massimiliano
• Is Part Of: Advances in engineering software. Volume 173(2022)
• Publication Date: 2022-11
• Subjects: Downburst analytical model -- Metaheuristic algorithm -- Differential evolution -- Teaching-learning-based optimization -- Single-objective optimization -- Downburst kinematic parameters; Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2022.103203
• Languages: English
• ISSNs: 0965-9978

Abstract

Highlights: Development of a global optimization procedure to estimate the main downbursts parameters. The performance efficiency of two global metaheuristic optimization algorithm is compared. Ten different real scale thunderstorm downburst outflows are simulated. Abstract: Strong localized downbursts generated in thunderstorms can produce surface winds very dangerous for civil structures and infrastructures. Modelling and simulating such severe wind systems is therefore extremely important for structural safety and design wind speed evaluation. This paper deals with the downburst wind field simulation by means of an optimization algorithm that uses a downburst analytical model, previously developed by the authors, and two metaheuristic algorithms, namely the Differential Evolution (DE) and the Teaching-Learning-Based Optimization (TLBO), for the downburst kinematic and geometrical parameters evaluation. The optimization problem minimizes the relative error between recorded and simulated wind speed and direction time histories. A comparison is made between the performance of two algorithms for ten thunderstorm events measured in north-western Italy between October 2011 and October 2015. Both algorithms provide solutions which are coherent with the downburst parameters values present in literature. TLBO outperforms DE since it has a faster convergence rate to the optimal solution.