Release Time:2019-03-12  Hits:

Indexed by: Journal Article

Date of Publication: 2016-10-12

Journal: NEUROCOMPUTING

Included Journals: Scopus、EI、SCIE

Volume: 209

Page Number: 46-56

ISSN: 0925-2312

Key Words: Slope stability; System probabilistic analysis; LSSVM; Particle swarm optimization; Response surface

Abstract: This paper presents an intelligent response surface method for evaluating system failure probability of soil slopes based on least squares support vector machines (LSSVM) and particle swarm optimization. A novel machine learning technique LSSVM is adopted to establish the response surface to approximate the limit state function based on the samples generated by computer experiments. Subsequently, the proposed response surface is utilized in conjunction with Monte Carlo simulation to obtain the desired reliability estimation. The hyper-parameters which are crucial to the performance of LSSVM are selected by a swarm intelligence algorithm called particle swarm optimization. Experimental results on three examples show that the proposed system reliability analysis method is promising for soil slopes with obvious system effects. (C) 2016 Elsevier B.V. All rights reserved.