A Parallel Slime Mould Algorithm with Boundary Rebound and Knowledge-Sharing Mechanism

The conventional slime mould method has high global search ability but has certain drawbacks such as sluggish convergence speed, limited accuracy, and the tendency to slip into a locally optimum solution. Aiming at the algorithm’s shortcomings, we propose an improved stochastic optimization algorithm—a parallel Slime Mould Algorithm with Boundary rebound and Knowledge-sharing mechanism (PKB-SMA). The main improvement measures are as follows: (1) Introduce a parallel technique to balance the algorithm’s capacity for exploration and development. (2) Propose a knowledge-sharing system to improve individual quality. (3) Using boundary rebound method to improve individual exploration ability. (4) The population was initialized by Hammersley sampling method. Several test functions are utilized to validate the performance of the PKB-SMA method, and its experimental findings are compared to those of competing algorithms.