中文摘要：

为进一步提高支持向量机预报模型在船舶横摇运动预报中的精确度以及实时性，提出一种利用混沌理论和在线最小二乘支持向量机的实时在线预报方法。在混沌动力系统相空间重构的基础上，采用饱和关联维数法对船舶横摇运动的时间序列进行混沌特性判定，并建立混沌在线最小二乘支持向量机实时预报模型。对某船横摇运动时间序列进行预报，仿真结果表明，此模型的实时预报误差指标值RMSE在7％左右，相比于基于支持向量机和神经网络的组合预报模型，该模型能够有效提高预报精确度和收敛速度，延长预报时间。

英文摘要：

A real-time prediction method which utilized chaos theory and online least squares support vector machines was proposed to further enhance accuracy and real-time of the support vector machine prediction model in the prediction of ship rolling motion. The correlation dimension method was used to identify the chaotic characteristics of the time series of ship rolling on the basis of phase space reconstruction of the chaotic dynamical system, and then the chaotic online least squares support vector machine realtime prediction model is established. The experiments of ship roiling time series prediction were conduc- ted. The simulation results indicate that real-time prediction root-mean-square error of the proposed method is about 7% and compared with the combination prediction model based on support vector machine and neural network, this real-time prediction method can effectively improve the convergence rate and the prediction precision and extend prediction time.