基于遗传算法-支持向量机模型在热带气旋强度预报中的应用

Application of Genetic Algorithm-Support Vector Machine model in tropical cyclone intensity forecast

利用遗传算法对支持向量机(SVM)模型参数进行寻优,找到最优参数组合后代入SVM模型中,得到基于遗传算法的支持向量机模型(GA-SVM),利用此模型对热带气旋强度进行预报实验。该模型对热带气旋强度12 h、24 h和48 h的预报平均绝对误差分别为3.01 m/s、4.46 m/s和6.57 m/s;比最小二乘回归法的预报精度分别提高了12%、11%、14%。

The parameters for SVM model were pretreated through genetic algorithm to get the optimum parameter values,and these parameter values were used in the SVM model.Genetic algorithm-support vector machine（GA-SVM） model was obtained,which was used to make Tropical Cyclone intensity forecasting.The Mean Absolute Difference of 12 h、24 h、48 h forecasting model is 3.01 m/s、4.46 m/s、6.57 m/s.The results show the superiority of the GA-SVM compare with least square regression method（LS）.For example,its fore-casting level of 12 hour and 24 hours has improved 12 % and 11 % than LS,but the number of 48 hour has be-came 14 %.