Cost Optimization of Cantilever Retaining Wall Using Flower Pollination Algorithm

Abstract

The flower pollination algorithm is nature-inspired in which it mimics the pollination process of the flowers. In this study, the cost of the cantilever retaining wall is optimized, which includes the cost of concrete and the cost of reinforcement steel. The eight design variables define the geometry of the RC wall. The external stability and structural stability are considered as constraints. Overturning, sliding, eccentricity, and bearing are considered external constraints. The structural stability constraints are shear and moment failure of toe slab, heel slab, and stem. The fitness function is developed using a static exterior penalty. The study found that the FPA outperforms the non-dominated sorting genetic algorithm (NSGA-Ⅱ) in this design problem without violating any constraints. An analysis of sensitivity is carried out for the angle of internal friction of backfill soil and the angle of backfill slope. Results show that the total cost rapidly increases after 5 m height of backfill but significantly decreases with the internal friction angle of backfill increases. For backfill heights from 6 to 10 m, the change in cost is significant after the value of the inclination angle of 15°.