Optimum Design of Lined Channels using Advanced Jaya Algorithm and Feed-Forward Artificial Neural Networks

The design of lined channels is expressed as a constrained optimization problem, where the primary objective is to minimize construction costs. To achieve universal applicability of the results, a dimensionless form of the cost function is employed for the lined channel optimization problem. This ensures that the optimized design can be adapted to any specific scenario. To tackle the optimization problem, an advanced version of the Jaya (AJaya) algorithm is introduced. The performance of the AJaya algorithm is assessed using two benchmark structural optimization design examples. Subsequently, AJaya is applied to the optimum design of the lined channel, using dimensionless design variables and the objective function. The optimization process for lined channels must account for various cross-sectional shapes and an extensive range of design variable combinations. As a solution, artificial neural networks are employed to predict the optimal dimensionless design parameters. The results demonstrate that this methodology is an effective tool for the optimal design of lined channels.