The Find-Fix-Finish-Exploit-Analyze (F3EA) Meta-Heuristic Algorithm with an Extended Constraint Handling Technique for Constrained Optimization

Abstract

As a novel population-based evolutionary algorithm, the Find-Fix-Finish- Exploit-Analyze (F3EA) meta-heuristic algorithm has been introduced for numerical optimization which develops new selection operators, a new parameter-free mutation operator, and a local search operator. The algorithm takes the surface of the objective function as the battleground and mimics the F3EA targeting process of object or installation selection for destruction in the warfare. It performs the main steps of Find-Fix-Finish-Exploit-Analyze (F3EA) in an iterative manner, wherein the Find step introduces a new individual selection mechanism via imitating the military radar detection rationale; in the Fix step, it is shown that how the reality of the target monitoring process can be transformed to a single variable constrained optimization problem to obtain a local search operator; In the finish step, new solutions are generated via a new adaptive mutation operator which is developed through the simulation of projectile motion using physics equations; in the exploit step it tries to take over opportunities presented by the generated potential solution and other members of the population; Finally, in analyze step, the population is updated. In this chapter, an extended epsilon constrained handling technique is used to handle constraints within the body of F3EA meta-heuristic algorithms that is called the εF3EA. In order to evaluate the efficiency of the F3EA algorithm, nine constraint benchmark functions are used. In addition, the εF3EA algorithm is compared with eight other algorithms, which included SAFFa, COPSO, ISR, ATMES, SMES, ECHT-EP2, HCOEA, and αSimplex. Results indicate that the algorithm is the best of all on benchmark test problem instances.