无约束优化演示文件修改版.pptVIP

Figure 3.13. Powell’s conjugate directions method, Example 3.6. * 接近最优点时用牛顿法，远离时用柯西法。 * Ref: Xin-SheYang. Engineering Optimization: An Introduction with Metaheuristic Applications. John Wiley Sons, Inc, 2010 * 7.9.1Cauchy算法 单变量优化 7.9.1Cauchy算法 7.9.2Newton方法 = 1 7.9.3修正Newton方法 单变量优化 7.9.4Marquardt方法 7.9.4Marquardt方法 7.9.5共轭梯度法 单变量优化 7.9.6准Newton方法 单变量优化 矩阵A对称正定，保证二阶收敛 A是对Hessian矩阵的近似 7.9.6准Newton方法 BFGS方法(Broyden–Fletcher–Goldfarb–Shanno) 7.9.7信赖域方法 信赖域内对近似函数优化 信赖域缩小 信赖域迁移 7.9.8基于梯度的通用算法 7.9.8基于梯度的通用算法 7.9.10各种方法比较 Rosenbrock函数 Rosenbrock函数 7.9.10各种方法比较 齿轮惯性问题 齿轮惯性问题 7.9.10各种方法比较 Wood 函数 Wood函数 Clearly, depending upon whether the function to be optimized is continuous or discontinuous or depending upon the nature of its domain, different methods will have to be employed to conduct the search for the optimum. A method that is effective for continuous functions will probably not be effective for discontinuous functions, although the converse may be possible. a function does not have to be continuous to be unimodal. Unimodality is an extremely important functional property used in optimization. * Figure 2.7. Local and global optima * Figure 2.10. Case (i) and case (ii) of Theorem 2.3. * * Figure 3.1. Contours of a multimodal function. * Figure 3.3. Two-variable nonlinear function of Example 3.1 * FIGURE 9.10 Various steps in the univariate search method. * Figure 3.5. Construction of new simplex. * Figure 3.6. Expansion and contraction of a simplex * Figure 3.7. Pattern search iterations for Example 3.3 * Figure 3.8. Quadratic with cross terms. Figure 3.9. A quadratic without cross terms. * Figure 3.10. Conjugacy in two dimensions. * Figure 3.11. Conjugacy from a single point * Figure 3.12. Construction of conjugate directions in three dimensions. * 参考书 最优化理论与算法（第2版）——清华大学研究生公共课教材·数学系列 陈宝林 编著 清华大学出版社，2005 7 无约束优化 单变量无约束优化 多变量无约束优化 7.1单变量函数f(x)性质 连续-离散 单峰 图1 单峰函数 7.2 最优化准则 目标函数对设计变量的一阶导数为0： 第一个不为0的高阶导数为 2.1 n为奇数，则x*为鞍点 2.2 n为偶数，则x*是一个极值点 如果 0, x*是一个极小值点 如果 0，