Bare-Bones Based Sine Cosine Algorithm for global optimization.

• The guided direction parameters of population search are modified, and the performance of standard SCA is improved by exponential decreasing strategy. • The proposed the modification of updated position equation remarkably balances exploration and exploitation significantly, so that the algorithm does not stagnate and converges prematurely to a local optimal position. • The greedy selection strategy is used in the update phase, which not only reduces the diversity spillover and keeps the stability of the solution but also provides a greedy search direction. The Meta-heuristic algorithm has become an effective solution to global optimization problems. Recently, a new meta-heuristic algorithm called sine-cosine algorithm (SCA) search algorithm is proposed, which uses the characteristics of sine-cosine trigonometric function in mathematical formulas to solve the optimal solution of the problem to be optimized. This paper presents a new variant of the SCA algorithm named Bare bones Sine Cosine Algorithm (BBSCA), which improves the exploitation ability of the solution, reduces the diversity spillover in the classical SCA search equation, and keeps the diversity of the solution very well. The proposed method uses Gaussian search equations and exponential decrement strategies to generate new candidate individuals, which use the valuable information hidden in the best individuals to guide the population to move in a better direction. At the same time, the greedy selection mechanism is adopted for the newly generated solution, which makes full use of the previously searched information to improve the individual's search ability. To evaluate the effectiveness in solving the global optimization problems, BBSCA has been tested on classic set of 23 well-known benchmark functions, standard IEEE CEC2014 and CEC2017 benchmark functions, and compared with several other state-of-the-art SCA algorithm variants. At the end of the paper, the performance of design algorithm BBSCA is also tested on classical engineering optimization problems. The numerical and simulation experimental results indicate that the proposed method can improve the performance of the algorithm and generate better statistical significance solutions in real-life global optimization problems. [ABSTRACT FROM AUTHOR]