基于多层隐类贝叶斯网络的舰船生存能力评估模型研究

Evaluation model of ship viability based on Hierarchical Latent Class of Bayesian networks

生存能力是舰船设计时需要考虑的一个重要性能指标,然而其涉及的因素较多,因素之间的关系错综复杂,不确定信息充斥其间.在系统分析舰船生存能力评估要素的基础上,针对评估过程中的不确定性信息难以量化处理的特点,引入基于贝叶斯网络的多层隐类模型算法对舰船生存能力进行评估,给出了模型评分原理.最后以实例说明了建模方法与评估过程,并结合专家意见分析了模型的优劣,说明该多层隐类模型的算法符合实际情况.

The ship viability is an very important capability index in the process of design. Whereas the assessment factors it involve are so many that this paper makes ship survivity and its assessment process analyzed completely and according to the characteristic of uncertainty information in the assessment process, the Bayesian Networks method based on Hierarchical Latent Class models is presented and the principle of marking for models is given on. In the last, a practical example is given to illustrate the process of estimating, and the grade of model is analysed based on experts＇ opinions to illustrate the rationality of the method.