One-Way Pioneer Guide Pity Beetle Algorithm: A New Evolutionary Algorithm for Solving Global Optimization Problems

To improve the convergence speed and accuracy of the Pity Beetle Algorithm(PBA), we propose an improved algorithm called one-way pioneer guide PBA (OPGPBA). First of all, given the specificity of the random sampling technology(RST), resulting in search holes, we propose a new rRST search mechanism, which integrates the random generation mode of the full-space traversal search capability. According to the algorithm's different requirements for neighborhood search, medium-scale search, large-scale search, and global search, the fixed-size mode factor is adaptively processed to make its size change adaptively with iteration. Finally, we propose a mechanism for the pioneer to guide population reproduction. The evolutionary direction of the population in each dimension is determined according to the change in the pioneer position, which effectively improves the convergence ability of the algorithm. Experiments on the CEC2013 test set show that the OPGPBA algorithm has apparent advantages in convergence speed and solution accuracy compared with the three excellent algorithms so far.