Performance Evaluation of Smart Flower Optimization Algorithm Over Industrial Non-convex Constrained Optimization Problems

Abstract

The large number of dimensions, complexity and constraints of real world problems is the deriving force to assess the algorithms in diverse applications scenarios. Over a past decade, nature inspired algorithm are seen as potential solutions to solve continuous optimization problems. These are non-deterministic algorithms with proven abilities to find near optimal solutions. However, according to not a single free lunch theorem, nor an any algorithm is suitable for solving all kinds of problems. This work present performance evaluation of recent meta-heuristic algorithm that is Sun Flower Optimization Algorithm (SFOA) over Non-Convex Constrained Optimization problems industrial problems from domains like mechanical engineering, chemical processes and Process synthesis and design problems. SFOA is compared with IUDE, MAgES, iLSHADE 424 for 15 problems with varied number of equality and inequality constraints. Results reveal that SFOA performs better for two out of four industrial chemical problems and five out of six process design and synthesis problems. Problems by mechanical engineering, IUDE, MAgES, iLSHADE 424 outperform SFOA for 4 out of five problems.