Assessment of a Consolidated Algorithm for Constrained Engineering Design Optimization and Unconstrained Function Optimization

For real-life optimization problems, methods with adequate capability in exploring the search space are crucial especially when having in mind the perpetual complexity of the problems. Consequently, presenting an effective algorithm to address these problems becomes imperative. The major objective of this work is to assess the application of a consolidated algorithm in addressing constrained and unconstrained function optimization problems. Though the flower pollinated algorithm (FPA) is commonly used, it does have its limitations, including being stuck at local minima, causing premature convergence, and creating imbalances between intensification and diversification. As the FPA operates, the solution to the optimization problem relies on communication with pollen individuals. Consequently, instead of leading pollens randomly, the FPA's exploratory skills are boosted by employing the pathfinder algorithm's (PFA) components to route them to much better locations in order to avoid local optima. For that reason, the PFA has been incorporated into the FPA in order to increase its performance. The efficacy of the proposed algorithm is tested using conventional mathematical optimization functions as well as two well-known constrained engineering design optimization problems. Experimental results showed that the suggested algorithm outscored its counterparts for both constrained and unconstrained optimization problems.