Improved sine cosine algorithm combined with optimal neighborhood and quadratic interpolation strategy

Abstract

The sine cosine algorithm (SCA) is a new population-based stochastic optimization algorithm, utilizing the oscillating property of the sine cosine function to balance the exploration and exploitation performance of SCA. A hybrid sine cosine algorithm based on the optimal neighborhood and quadratic interpolation strategy (QISCA) was proposed to overcome the shortcoming of updating the population guided by the global optimal individual in the sine cosine algorithm. The new algorithm uses a Stochastic Optimal Neighborhood for neighborhood updates, and it adopts a Quadratic Interpolation curve for individual updates. In addition, QISCA incorporates Quasi-Opposition Learning strategies to enhance the population’s global exploration capabilities, and improves the convergence speed and accuracy. The two simulation experiments of 23 benchmark functions and 30 latest CEC2017 test functions show that the new algorithm can better coordinate the exploration and exploitation capabilities and improve the global optimization ability, compared with the other improved sine cosine algorithm and the representative stochastic optimization algorithm. The three representative engineering problems validate the effectiveness of the new algorithm to solve practical problems.