舰船最小阻力估算的智能学习算法

Intelligent learning algorithm for estimating the minimum resistance of ships

考虑舰船阻力性能受船体参数及航行环境条件的综合影响,数据类型繁多且复杂,最小阻力估算难度较大,研究基于智能学习的舰船最小阻力估算算法。此算法使用RANS控制方程和湍流模型分析舰船运动状态,由重叠网格技术分析舰船运动时粘性流场信息后,将舰船船型参数与运动状态数据、流场数据,作为基于GRNN智能优化学习的舰船最小阻力估算模型输入向量,在改进果蝇优化算法寻优设置GRNN的平滑因子后,利用此网络构建最小阻力估算模型,学习拟合舰船阻力与输入向量之间关系,输出最小阻力估算结果。实验结果显示:在舰船纵倾与升沉运动中,所提算法估算的最小阻力较准确。

Considering the comprehensive influence of ship parameters and navigation environment conditions on the resistance performance of ships,there are numerous and complex data types,and the difficulty of estimating the minimum resistance is high.Therefore,research is needed on an intelligent learning based algorithm for estimating the minimum resist-ance of ships.This algorithm uses RANS control equations and turbulence models to analyze the motion state of ships.After analyzing the viscous flow field information during ship motion using overlapping grid technology,the ship type parameters,motion state data,and flow field data are used as input vectors for the ship minimum resistance estimation model based on GRNN intelligent optimization learning.After improving the fruit fly optimization algorithm to optimize and set the smooth-ing factor of GRNN,this network is used to construct the minimum resistance estimation model,learn and fit the relation-ship between ship resistance and input vector,and output the minimum resistance estimation result.The experimental results show that the proposed algorithm estimates the minimum resistance accurately in the longitudinal tilt and heave motion of the ship.