Gravitational search algorithm for probabilistic design of HBPS canals

Abstract

In this study, the performance of a recent meta-heuristic technique, namely, gravitational search algorithm (GSA) is evaluated for a deterministic as well as for a probabilistic design of canals that have cross-sectional shape of horizontal bottom and parabolic sides (HBPS). The uncertainty in various input parameters is modeled using probabilistic approach and a modified chance-constrained model is presented for an optimal design of HBPS canals with the aim of minimizing the total cost, while satisfying the basic flow constraints and reliability constraints on the canal capacity and overtopping. First, the GSA method is applied to solve the HBPS canal design problem under different constraints, and its performance is evaluated by comparing with the solutions of the deterministic models by the particle swarm optimization and genetic algorithm. Then, the GSA is applied to obtain the solution of the probabilistic model and in view of multiple conflicting goals; pseudo-weight vector approach is adopted to assist in decision-making and demonstrate its applicability for arriving at a suitable decision. The results obtained suggest that the proposed GSA approach has good potential for a reliable and cost-effective design of canals.