New Simple Trigonometric Algorithms for Solving Optimization Problems

Abstract

Heuristics play a key role in solving optimization problems with complex functions. They are popular as efficient heuristics are capable of providing results quickly with acceptable solution quality. Population-based heuristics are stochastic and hence, several iterations and trials are needed to achieve the expected accuracy and convergence to the global optimum. This article proposes two new, simple; population-based trigonometric algorithms, Sine (AB) and Cosine (AB). The algorithms are validated using forty well-known benchmark test functions available in the literature. The results are compared with a similar popular Sine Cosine Algorithm and the computational results show that the performance of Sine (AB) and Cosine (AB) are better than Sine Cosine Algorithm. Wilcoxon Signed-Rank and Friedman tests are carried out for statistical analyses. In addition to unconstrained functions, three real-world, constrained problems are solved to have a more intensive analysis of the proposed algorithms.

Keywords

• optimization
• population-based heuristic
• un-constrained optimization
• sine cosine algorithm
• trigonometric algorithm