基于RBF网络的操纵水动力预报  

Prediction of Ship Maneuverability Hydrodynamics Based on Radial Basis Function Neural Network

作　　者：张晓兔[1] 张乐文[1] 刘祖源[1] 

机构地区：[1]武汉交通科技大学船舶及海洋工程系,武汉430063

出　　处：《武汉交通科技大学学报》1999年第6期602-604,共3页Journal of Wuhan University of Technology(Transportation Science & Engineering)

基　　金：国家自然科学基金!19672045

摘　　要：在船舶初始设计阶段，要求设计者能够在不断调整船舶主尺度和船型系数的同时，了解所设计船舶的操纵性能指标．其中，操纵水动力导数预报的精确程度无疑对整个船舶操纵性预报工作具有至关重要的影响．文中基于径向基函数（RBF）网络提出了预报操纵水动力导数的新方法，并且与试验结果、线性回归方法预报结果进行了比较，与试验结果较为一致．At the preliminary stage of ship design,it required that ship designer should know of the maneuverability criterion of the designed ship when some main dimensions and form coefficients were modified. The precision of the maneuverability hydrodynamic derivatives was undoubtedly of great importance to predict the ship maneuverability at the preliminary stage of ship design. This paper put forward a new method of predicting maneuverability hydrodynamic derivatives based on the radial basis function neural network,and compared with the trial resu1ts and the results derived from linear regression formula,and found that the predicted results based on RBF neural network kept good agreement with the tria1*results.

关 键 词：船舶 操纵性 水动力 预报 RBF网络 

分 类 号：U661.33[交通运输工程—船舶及航道工程]