دانلود رایگان مقاله استفاده از یک الگوریتم ژنتیک برای طراحی و بهینه سازی فرآیند تقطیر فشار نوسان

Abstract

The design and optimization of pressure-swing distillation (PSD) have a critical impact on its economics. An optimization method based on simulated annealing algorithm (SAA) was proposed. The move generator and cooling schedule of the SAA were discussed, and suitable parameter settings were investigated. Two cases of PSD with and without heat integration were optimized by the SAA-based optimization method using procedures of pressure specified and pressure optimized. The results of the process without heat integration were compared with conventional optimization methods. For the acetone-methanol system, the total annual cost (TAC) shows a 5.69% decrease with the pressure specified and a 17.32% decrease with the pressure optimized. For the methanol-chloroform system, the TAC shows a 1.79% decrease with the pressure specified and a 9.04% decrease with the pressure optimized. The SAA-based optimization method has the advantages of a high probability to obtain the global optimum, automatic calculation, and less computing time.

4. Conclusions

A method of design and optimization of the PSD process based on the SAA was proposed in this paper. To handle the problem of optimizing the PSD process with continuous variables and discrete variables,the move generator and cooling schedule ofthe SAA were discussed, and suitable parameter settings for the PSD were investigated. The cases of methanol-chloroform and acetone-methanol systems were studied. The results show that the present method can optimize the PSD process with no, partial, and full heat integrations. A SAA-based optimization process starting with different initial values of PSD design variables results in the same optimum design, which demonstrated that the global optimal design is obtained. The optimization results of the PSD process without heat integration show that an improved optimum design with a minor TAC is obtained by the SAA-based optimization method compared with those of the conventional optimization methods. The optimization results with the pressure optimized are substantially improved compared with those ofthe pressure specified. Therefore, this step is important for the optimization of the pressures in the two columns of PSD. The SAA-based optimization method has the advantages of automatic optimization, less computing time, and a greater chance to obtainthe global optimal design. This work will be helpful to optimize the PSD processes for separating other systems.