Title: Construction safety risk assessment method of construction engineering based on improved SVM

Authors: Quan Yin; Junqing Zhou; Yi Zhou; Zhi-ping Li

Addresses: Hunan Engineering Research Center of Structural Safety and Disaster Prevention for Urban Underground Infrastructure, Hunan City University, Yiyang, 413000, China ' Hunan Chemical Geological Engineering Investigation Co., Ltd., Changsha, 410000, China ' Hunan Engineering Research Center of Structural Safety and Disaster Prevention for Urban Underground Infrastructure, Hunan City University, Yiyang, 413000, China ' Hunan Engineering Research Center of Structural Safety and Disaster Prevention for Urban Underground Infrastructure, Hunan City University, Yiyang, 413000, China

Abstract: In this study, SVM was improved to complete the design of a building safety risk assessment method. This method selects the index to design the scientific evaluation system and completes the calculation of the index weight. Then, the support vector machine is improved and optimised. The judgement matrix is established according to the nine-level scale method, and it is standardised as the input of SVM, and the building safety risk assessment grade is output through the optimal hyperplane classification. The results show that the generalisation error of this method is between 0.034 and 0.053, and the evaluation time is between 1.5 h and 2.5 h, which can better evaluate the safety risk of engineering construction.

Keywords: architectural engineering; construction safety risk; risk assessment; Cauchy function; particle swarm optimisation algorithm; SVM.

DOI: 10.1504/IJSD.2023.134411

International Journal of Sustainable Development, 2023 Vol.26 No.3/4, pp.329 - 343

Received: 13 Feb 2023
Accepted: 23 Jun 2023
Published online: 20 Oct 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article