论文部分内容阅读
当布拉格光栅轴向存在大的应变梯度时,其反射光谱的形状会被扭曲,甚至出现多峰,发展非均匀应变分布重构方法对于结构健康监测技术具有重要意义。但采集反射光谱时的测试噪声会显著影响应变分布重构的精度。为此,提出了采用支持向量机对含噪的反射光谱进行回归预处理,并运用适应度排序改进的遗传优化算法结合传输矩阵反射光谱构建方法识别布拉格光栅轴向非均匀应变分布的方法。该方法将反射光谱视为时间序列,利用支持向量回归的全局优化和泛化能力进行噪声抑制,从而回归出有效的反射光谱;通过传输矩阵方法将光栅轴向应变分段均匀化,利用改进的遗传算法进行并行重构。对多种应变分布形式下的应变重构进行了仿真研究,结果表明,支持向量机方法可以有效地进行反射光谱回归,提高非均匀应变分布重构的精度。
When the Bragg grating has a large strain gradient in the axial direction, the shape of the reflection spectrum will be distorted and even appear as a multi-peak. Therefore, it is of great significance to develop a non-uniform strain distribution reconstruction method for structural health monitoring technology. However, the acquisition of reflection spectrum of the test noise will significantly affect the accuracy of strain distribution reconstruction. In order to solve the above problem, a new method based on support vector machine (SVM) is proposed to predict the nonuniform strain distribution of the Bragg grating by using regression genetic algorithm with modified genetic algorithm and transmission matrix reflection spectrum. In this method, the reflection spectrum is regarded as time series, the global optimization and generalization ability of support vector regression are used to suppress the noise, and the effective reflection spectrum is regressed. The axial strain distribution of the grating is homogenized by the transfer matrix method, Genetic algorithm for parallel reconstruction. The strain reconstruction under various forms of strain distribution is simulated. The results show that SVM can effectively reflect the spectrum and improve the accuracy of the non-uniform strain distribution reconstruction.