论文部分内容阅读
提出了一种适用于大规模过程系统在线优化命题的稀疏全空间 SQP算法。该算法利用解析导数构造Hessian矩阵 ,保持了系统的稀疏结构 ,并为寻优过程提供了精确的曲率信息。与传统 SQP算法相比 ,该算法的计算效率有了明显提高。对开放式方程精馏塔模型的实例计算表明 ,该算法在求解大规模过程系统优化命题时是可行和有效的。
A sparse full-space SQP algorithm is proposed for on-line optimization of large-scale process systems. The algorithm uses analytic derivatives to construct Hessian matrix, preserves the sparse structure of the system and provides accurate curvature information for the optimization process. Compared with the traditional SQP algorithm, the computational efficiency of the algorithm has been significantly improved. The calculation of the open equation distillation column model shows that this algorithm is feasible and effective in solving the optimization proposition of large-scale process system.