摘要
应用神经网络技术建立船舶稳性知识库，该方法建模简单，船型参数表达比较全面，预报精度高．根据船舶初步设计阶段特点及影响船舶稳性的相关因素，提出了目标函数形式及建立知识库的方法．用一艘实船数据对知识库进行检验，结果表明：基于神经网络建立的船舶稳性知识库，能利用过去的数据资料对其训练，获取知识并有效地预报设计船舶的稳性性能．
A method to establish the knowledge base of ship stability based on neural-network is presented. According to the feature of the preliminary design stage and the elements of affecting the stability, the form of the stability function and the method of setting up the base is raised. The base is examined by the data of a real ship. The results show that the method is cleay and simple, the base based on the BP neural-network can be trained by the data of old ship stability, the prediction of new ship stability is more accurate by the base.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
1996年第12期75-79,共5页
Journal of Shanghai Jiaotong University
关键词
神经网络
船舶稳性
知识库
neural-network
ship stability
knowledge base
