基于遗传算法的海气耦合模型参数反演

为了提高模型参数反演效率和精度,基于遗传算法和系统优化思想,从一个简单的海气耦合模式出发,反演模式参数和强迫项.试验结果表明:遗传算法和系统优化结合的方法,可以有效优化模式参数和强迫项,进而修正模式的预报误差.同时,该方法具有参数反演过程不与模型结构发生耦合,计算量小和操作性强、反演效率高等优点,为在实际应用中改善模式参数和初值、提高预报准确率提供了一个可借鉴的方法和途径.

Parameters invers ion of coup led a ir-sea mode l based on genet ic a lgor ithm

In o rde r to im prove the e ff iciency and accuracy o f them ode l param e te rs, the non linea r dynam icEN SO m odel pa ram e ters and fo rcing te rm s w e re inve rsed from a sim ple coup led air-seam ode l based on theinve rsion gene t ic a lgo r ithm and the idea o f sy stem op t im iza t io n. T he resu lts ind ica te th at the m e thod canop t im ize them ode l param ete rs and the fo rcing item s to am end the m ode l p redic tion e rro rs, th at the inv ersion p ro cess is no t coup led w ith the mo de l stru ctu re in the m ethod, and th at w ith them e rits o f sm a ll computa t ion am ount, st rong operab ility and h ig h inv ersion e f ficiency, them e thod g iv es a reference fo r im p rovingthe m ode l pa ram e ters, the in itia l va lues and the p red ict ion accuracy ra te