【摘要】 应用可视化方法进行化工过程优化设计:应用平行坐标系进行多维矢量可视化,并在此基础上提出了一种在n 维空间用于多维矢量线性回归的“扫描线”算法。在优化问题中该算法可以将问题简化到一或二维,从而以二到三维笛卡尔坐标为基础的传统可视化方法可用于优化设计。将这种方法与遗传算法相结合,不仅可以得到较好的优化结果,还可得到一些“额外信息”以使用户理解基于可视化的优化设计过程并接受该结果。更多还原

【Abstract】 The visualization method is used for chemical process optimal design.Parallel Coordinate System is used here for the multi-dimensional vectors’visualization and a scan line algorithm is developed for the identification of lines to which clusters of vectors areclose in n-dimensional space.The lines can be used for the dimension reduction of the optimal problem to just one or two dimensions,so the traditional visualization methods based on two or three dimensional Cartesian Coordinate System can be used for the optimal de-sign.Combined the proposed approach with Genetic Algorithm,not only the optimum results but also some addition information is ob-tained for the user to understand the optimization process and accept the result.更多还原