基于改进型遗传算法的海上无人集群协同作业需求−功能动态匹配研究

Research on dynamic matching between requirements and functional models for unmanned maritime swarm collaborative operation based on improved genetic algorithm

  • 摘要:
    目的 针对海上无人集群协同作业需求动态性强导致的协同作业需求与功能模型之间匹配难的问题,提出一种融合基于精英保留分层选择策略和基于自适应概率的交叉与变异策略的改进遗传算法(IGA),以实现协同作业需求与功能模型的动态匹配。
    方法 分析获取的动态作业需求数据以及多类无人设备逻辑模型作为海上无人集群协同作业需求−功能动态匹配方法的输入,并根据海上无人集群协同作业的特点完成基因编码,再通过基于精英保留分层选择策略和基于自适应概率的交叉与变异策略,有效平衡算法的全局和局部搜索性能,进而动态生成协同作业需求−功能的最优匹配方案。
    结果 结果表明,在相同的算法参数下,所提方法所生成的最优匹配方案的适应度值,相较于蜜蜂算法(BA)和人工蜂群算法(ABC),分别平均提高了7.8%和7%,运行时间分别平均加快了5倍与21倍。
    结论 所提理论与方法可应用于海洋装备设计与运行优化研究领域,有助于提升海洋装备的智能化程度。

     

    Abstract:
    Objective Aiming at the challenges in matching the requirements and functional models caused by the dynamic nature of the requirement model for unmanned maritime swarm collaborative operations, an improved genetic algorithm (IGA) is proposed. This algorithm combines a hierarchical selection strategy based on elite retention with a crossover and mutation strategy based on adaptive probabilities to realize dynamic matching between the collaborative requirements and functional models.
    Methods The method uses dynamic requirement data and multiple unmanned equipment logic models as inputs for dynamic matching between collaborative operation requirements and functional models of unmanned maritime swarms. Gene coding is performed based on the characteristics of unmanned maritime swarm collaborative operations. The hierarchical selection strategy based on elite retention and the crossover and mutation strategies based on adaptive probabilities are used to balance the global and local search performance. Afterwards, the optimal matching solution between the requirement and functional model could be dynamically generated.
    Results The results show that under the same algorithm parameters, the fitness of the proposed method is averagely 7.8% and 7% higher, and the running time is averagely 5 and 21 times faster, than those of the bee algorithm and the artificial bee colony algorithm respectively.
    Conclusion The proposed theory and method could be applied to optimizing maritime equipment design and operations, which could improve the intelligence of maritime equipment.

     

/

返回文章
返回