PSO-based stability analysis of unreinforced and reinforced soil slopes using non-circular slip surface

Abstract

Slope stability analysis is one of the most intricate problems of geotechnical engineering because it is mathematically difficult to search the critical slip surface of earth slopes with complex strata owing to the involved multimodal function optimization problem. At present, a minimum factor of safety for a non-circular slip surface in a uniform and unreinforced earth slope can be calculated using several methods; however, for a reinforced soil slope, it cannot be easily calculated because of the additional effect of the reinforcement. One efficient method to search the critical slip surface is particle swarm optimization (PSO). PSO can solve complex non-differentiable problems, and its increasing ease of use has facilitated its application to multimodal function optimization problems in a variety of fields. However, the recommended PSO parameters to calculate the safety factors of unreinforced and reinforced soil slopes, namely the inertia and local and global best solution weighting coefficients, have not been sufficiently investigated. Moreover, the computational efficiency of PSO for safety factor calculation, including computational accuracy and time, has not been clarified. To calculate the unreinforced and reinforced soil slope safety factors, this study considers force and moment equilibriums, including the tensile force of the reinforcement. Firstly, the computational efficiency of the calculation process by PSO was shown to increase the maximum number of slip surface nodes in the calculation of the safety factor. Then, an analysis was carried out to investigate the safety factor sensitivity to the PSO parameters. Based on this analysis, appropriate PSO parameters for the safety factor calculation of unreinforced and reinforced soil slopes were proposed.