A dimensional learning squirrel search algorithm based on roulette strategy
Daiquan Wen, Lin Huo · 2022 Asia Conference on Algorithms, Computing and Machine Learning (CACML) · 2022
Squirrel search algorithm (SSA) is an effective intelligent optimization algorithm, but it also has the problems of slow convergence speed and easy to fall into local optima. To solve these problems, a dimensional learning squirrel search algorithm (DLSSA) based on roulette strategy is proposed. Roulette strategy was introduced to update the position of squirrels on normal trees to increase the diversity of the algorithm and avoid falling into local optimal solution. Using the method of dimensional learning to guide the location update of squirrels on oak trees, the excellent information of hickory tree can be accurately obtained to increase the efficiency of information flow transmission. Cauchy mutation was performed on squirrel on hickory tree to accelerate the convergence of the algorithm. Optimization experiments are carried out on 10 benchmark functions. Compared with other 7 algorithms, DLSSA has obvious advantages in the success rate of optimization and convergence speed.