Cooperative co-evolutionary comprehensive learning particle swarm optimizer for formulation design of explosive simulant

Generally, the actual explosive is not suitable for the training of security personnel due to its danger. Hence, it is significant to create the simulant as similar as possible to the real explosive, where the difficulties are derived from finding safe compounds from the compound database and their related proportion. In this paper, a cooperative co-evolutionary comprehensive learning particle swarm optimizer is proposed to obtain the formulation design of explosive simulant. To be specific, the proposed algorithm employs particle swarm optimization as the optimizer and creates two cooperative populations focusing on finding compounds and their proportions, respectively. Moreover, a comprehensive cooperative strategy is designed to improve the solution diversity and thus enhance the search performance. To the best of our knowledge, this is the first attempt to employ evolutionary algorithm to design explosive simulant formulation. Comprehensive experiments are conducted on several typical explosives and results demonstrate the superiority of the proposed algorithm in comparison to other algorithms.