中文摘要：

主要研究了基于捕食搜索策略遗传算法的支持向量机参数优化方法在风切变识别方面的应用。首先，利用已有的仿真雷达数据生成仿真雷达扫描图，通过Ostu自动阈值分割提取风切变区域获取样本图像。然后，对样本图像进行二层小波分解，求取各子带小波系数的均值和标准差作为特征向量。最后，利用捕食搜索策略的遗传算法优化支持向量机的核函数参数，对特征向量进行识别分类。实验结果表明，该算法识别率可达97．3％，在低空风切变的识别中具有良好的可行性。

英文摘要：

Application in low - level wind shear was studied of the approach that parameter optimization method of support vector machine （SVM） based on genetic algorithm of predatory search strategy. Firstly, the existed simula- tion lidar data was emploied to generate the lidar scanning images, and the sample images of wind shear regions was extracted by threshold segmentation. Then, two levels wavelet decomposition on sample images was comple- ted, the feature vectors was obtained by strike mean and standard deviation of each sub - band of wavelet coeffi- cients. Finally, genetic algorithm based on predatory search strategy was used to optimize the parameter of kernel function included in SVM, so the optimized SVM could classify the eigenvectors of different types in order to get the better recognition accuracy rate. The simulation demonstrates that the recognition rates can be up to 97.3 % , it is well used in the recognition of low - level wind shear.