A Socio-based Cohort Intelligence Algorithm for Engineering Problems

Abstract

Mostly the design engineering problems are consisting of integer, discrete, and mixed design variables. Associated to these types of variables, the accessible search space is very limited which may increase the complexity of problems. To solve such problems, traditional optimization methods such as Newton’s method, Newton–Raphson method, Gradient methods, etc. are unable to work efficiently. In order to overcome this limitation, various bio-based (Genetic Algorithm (GA)) and social-based (Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), Firefly Algorithm (FA), etc.) optimization approaches were introduced. In the present work, a socio-based Cohort Intelligence (CI) algorithm proposed by Kulkarni et al. [1] is implemented to solve discrete variable truss structure problem, mixed variable design engineering problem, and integer variable. The CI algorithm is incorporated with a group of learning candidates which interact and compete with one another within a cohort to achieve their individual goal and further make to improve the overall cohort behavior. The variables involved with these problems are handled using simple round-off technique. Also, the well-known static penalty approach is adopted to handle the inequality constraints. Further, CI is successfully compared with several algorithms such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Mine Blast Algorithm (MBA), Harmony Search Algorithm (HS), etc. The reported solutions using CI were significantly better than that of other algorithms reviewed in literature with a lesser amount of computational cost (function evaluations and computational time).