基于多变量加权一阶局域法多步预报模型的舰船摇荡预报

摘要

针对舰船各个自由度摇荡之间存在耦合的实际，提出了一种多变量加权一阶局域法多步预报模型（MAO-LMM）。对Chen’s吸引子仿真数据及某型船真实海浪下航行的摇荡数据进行了预报分析，结果显示，多变量加权一阶局域法多步预报模型的预报精度高于单变量加权一阶局域法多步预报模型，提高了舰船摇荡预报的有效性。%A model of Multivariate Add-weighted One-rank Local-region Multi-steps Method (MAOLMM) is proposed aiming at various degrees of freedoms for ships’ actual coupling effect. A forecast test is oper-ated with Chen’s simulation data. Simulation results show that the precision of Multivariate Add-weighted One-rank Local-region Multi-steps Method is better than Single Add-weighted One-rank Local-region Multi-steps Method (SAOLMM). And it raised the effectiveness of predicting ship motions.