为了对营运船舶的操纵性进行预报,基于SPSS软件的相关分析和统计功能,采用多项回归法对59艘船舶的K、T指数统计资料进行分析,得出具有一定精度的K′、T′指数估算公式。考虑船型数据之间非线性因素的影响,构造了四元二阶多项式回归模型。增加了渔船等中、小型船舶,使回归方程更具有普遍性。所构建的船舶操纵性指数回归模型较原有的回归模型将K′、T′的复相关系数精度分别提高了8.5%和7.2%,在保证原回归方程显著性的基础上兼顾了数据拟合的精度和回归多项式的简洁性。因此,该模型对操船者在航海实践中评估船舶的操纵性有一定的实用价值。
In order to forecast maneuverability of ships,based on the correlation analyses and statistics functions of the SPSS software,two formulae of estimating K and T are obtained through multinomial regression analysis of 59 ships' K and T statistical data.Considering the nonlinear effects of ship forms,two four-element,second-order polynomial regression equations are set up.Also,this regression equation can be used more extensively by including some smaller ships,such as fishing vessels.Compared with existing regression model,the model improves the precisions of the multiple correlation coefficients of K and T by 8.5% and 7.2% respectively.Besides keeping the explicitness of existing regression equation,the polynomial regression equation improves precision of data fitting and simplicity of regression polynomial.It is valuable in navigation practice for sailors to evaluate ship maneuverability.