关键词: 过程优化, 过程设计, 设计裕量, 相对增益, 优先级

Abstract:

The design margin of chemical process is generally given by design experience or economic optimization. The design experience cannot guarantee the optimization of economic performance. And the economic optimization needs to solve a large-scale nonlinear optimization problem. The optimization calculation is often complicated for large-scale nonlinear optimization, and it is easy to fall into the local extreme point, and the design results sometimes is against the design experience. In this paper, the non-square relative gain array(NRGA) and the non-square relative energy gain array (NREGA) are used to describe the sensitivity relationship between independent variables and dependent variables in chemical process. The independent variables are divided into operation variables and design variables, and the dependent variables are divided into economic optimization variables and constraint variables. Because relative gain is dimensionless and normalized, operation variables and design variables can be prioritized according to the relative gains of economic optimization variables and constraint variables, and for process uncertainty, each operation variable and each design variable can be adjusted according to the priority. The margin design scheme can be found with the least impact on the process economic performance and the effectively moving of the operation point away from the constraint boundary. Finally, the series reactor is taken as an example to verify the design method. The results show that, compared with the margin design method for solving economic optimization problems, this design method obtains design results close to the economic performance, and the calculation is simple, and there is no need to solve the optimization problem.

Key words: process optimization, process design, design margin, relative gain, priority