开阔水域两船智能避碰算法评价方法研究

    Research into the evaluation method of an intelligent collision avoidance algorithm for two vessels in open water

    • 摘要: 目前国内外仍未形成针对船舶智能避碰算法的统一测评标准体系,建立主客观结合的智能避碰算法评价方法对实现船舶智能避碰具有重要意义。提出一种开阔水域下两船避碰算法能力的综合评价方法:首先,基于对《国际海上避碰规则》和专家经验的合理量化,以合规性、安全性、经济性三个核心指标构建算法评价体系;其次,根据收集的问卷调查结果,使用层次分析法及专家估测法确定指标权重,选择合适的隶属度函数建立评价模型并开发了智能避碰算法模糊综合评价系统;最后,利用构建的评价系统对典型场景进行评测。测试结果表明:评价结果公平公正,能较客观地反映算法的优劣,并能根据评分结果提供算法的改进方向。

       

      Abstract: Currently, there is still no unified evaluation standard system for intelligent collision avoidance algorithms for vessels, either domestically or internationally. The development of an intelligent collision avoidance algorithm evaluation method that combines subjective and objective criteria is therefore of great significance for the realization of intelligent collision avoidance for vessels. This paper proposes a comprehensive evaluation method for the collision avoidance capability of two vessels in open water. Firstly, an algorithm evaluation system based on three core indexes of safety, compliance and economy is constructed using the reasonable quantification of the "International Regulations for Collision Avoidance at Sea" and experts' experience. Secondly, the weights of the indexes are determined using the hierarchical analysis method and the expert estimation method based on the results of the questionnaires collected, and an appropriate subordination function is established by choosing a suitable evaluation model and developing a fuzzy evaluation model for the intelligent collision avoidance algorithm. Secondly, an evaluation model and a fuzzy comprehensive evaluation system for the intelligent collision avoidance algorithm are developed. Finally, the constructed evaluation system is used to evaluate typical scenarios. The evaluation system is then used to evaluate typical scenarios. The test results demonstrate that the evaluation is fair and impartial, objectively reflecting the algorithm's strengths and weaknesses, and providing direction for improvement based on the scoring results.

       

    /

    返回文章
    返回