论文部分内容阅读
高斯过程是一种具有严格的统计学习理论基础,在处理高维数、非线性、小样本的复杂回归问题中具有较高精度的机器学习方法。针对可靠度领域中采用传统响应面法求解隐式功能函数结构可靠度精度不足的问题,采用高斯过程回归模型重构隐式功能函数,并与改进传统响应面法相结合,提出了一种基于高斯过程响应面方法的结构可靠度分析。研究分析表明,该方法在处理隐式功能函数的可靠度问题方面具有结果可靠且计算效率高的优势。
Gaussian process is a kind of machine learning method with rigorous statistical learning theory and high accuracy in dealing with complex regression problems of high-dimensional, nonlinear and small samples. Aiming at the problem of using the traditional response surface method to solve the structural reliability of the implicit functional function in the field of reliability, the Gaussian process regression model is used to reconstruct the implicit functional function. Combined with the improved traditional response surface method, a Gaussian Structural Reliability Analysis of Process Response Surface Methodology. Research and analysis show that this method has the advantages of reliable results and high computational efficiency in dealing with the reliability of implicit function.