论文部分内容阅读
为了提高光学元件中频误差的识别精度,提出一种经验模态分解和支持向量机的光学元件中频误差识别模型(EMD-SVM)。首先采用经验模态分解法对光学元件表面的原始误差信呈进行分解,得到一组固有模态函数;然后采用固有模态函数和残差之和构建特征子集,并采用支持向量机建立光学元件中频误差识别模型;最后对模型性能进行了验证性测试。实验结果表明,EMD-SVM可以准确识别分布于光学元件表面的中频误差,提高了中频误差辨识的精度,可以满足光学元件质量检测的识别精度要求。
In order to improve the recognition accuracy of mid-frequency error of optical components, an empirical model decomposition and support vector machine (CMD) intermediate frequency error identification model (EMD-SVM) is proposed. Firstly, the original error signal on the surface of the optical element is decomposed by empirical mode decomposition to obtain a set of intrinsic modal functions. Then, the subset of features is constructed by using the sum of the intrinsic mode functions and the residuals, and the support vector machine is used to establish the optics Component intermediate frequency error identification model; Finally, the performance of the model was tested. The experimental results show that the EMD-SVM can accurately identify the IF errors distributed on the surface of optical elements and improve the accuracy of IF error identification, which can meet the recognition accuracy requirements of optical element quality inspection.