海浪水压场的混沌特性分析及预测模型

The Analysis of Wave Hydrodynamic Pressure Signals With Chaos Method and the Prediction Model

海浪水压场作为水中兵器水压引信最严重的干扰源,对其混沌特性进行研究具有极为重要的意义.依据判断系统是否具有混沌特性的2个重要指标:关联维数和最大Lyapunov指数,分别通过G-P算法和小数据量的Wolf算法进行了计算.对一定数量的海浪水压场数据所进行的计算表明,海浪水压场具有分数的关联维数和正的Lyapunov指数,表现出明显的混沌特性.在此基础上依据混沌信号的相空间重构理论使用支持向量机对海浪水压场进行了预测建模,结果表明非线性混沌模型较常规的线性AR模型具有更高的预测精度.

As a major interference source to the pressure fuze of the underwater ordnance, the research of chaos character to the wave hydrodynamic pressure is important. For the correlation dimension and the largest Lyapunov exponent is the two key indexes to judge the chaos character of the system, they are calculated respectively by the G-P algorithm and the Wolf algorithm. The calculated results of many samples display the wave hydrodynamic pressure have fractional correlation dimension and positive largest Lyapunov exponent. Based on the results, the wave hydrodynamic pressure is modeled by support vectors machine with the phase-space reconstructing theory, and the results show the nonlinear model is more precise than the linear AR model.