Slope Stability Evaluation Based on Projection Pursuit and Attribute Mathematics

Abstract

Evaluation of slope stability is a complex system problem of uncertainty. A new model for slope stability evaluation was established based on the Attribute Recognition Model (ARM) and the Projection Pursuit (PP) optimized by the Particle Swarm Optimization (PSO), which was named as Attribute Recognition Model based on Projection Pursuit Weight (ARM-PP). The attribute measurement of single index was computed through constructing the attribute measurement functions, and the synthetic attribute measurement was calculated by the weight of PP optimized by PSO. Confidence criterion was used to recognize the slope stability. The attribute mathematical theory could successfully resolve the comprehensive evaluation problem with a number of fuzzy attribute. Furthermore, ARM-PP adopted the PP to determine weight, avoiding the subjectivity and randomicity, and ensuring objectivity and accuracy of the evaluation. Case study showed that ARM-PP was feasible and precise in slope stability evaluation.