An Improved Squirrel Search Algorithm for Optimization

Squirrel search algorithm (SSA) is a novel meta-heuristic optimization algorithm which is more appropriate for solving unimodal, multimodal, and compound problems. However, SSA has shortcomings such as limited accuracy and slow convergence. In order to improve its global search ability and convergence speed, this paper proposes an improved squirrel search algorithm (ISSA). First, the social part of the particle swarm algorithm is introduced into the position update formula on the ordinary tree to improve the search range and convergence speed of SSA. Second, a new location update condition is added, which can balance between exploration and exploitation of ISSA. Third, in order to prevent falling into a local optimal solution, an adaptive strategy of the minimum seasonal constant is proposed. The proposed ISSA has been tested by 12 benchmark functions and compared with the results of six other intelligent optimization algorithms. The effectiveness and superiority of ISSA are shown in average convergence value and convergence speed.