摘要
以有限元法求解模拟移动床的稳态TMB模型和动态SMB模型,提出基于Pareto非劣解集的多目标双种群遗传粒子群算法;利用动态SMB模型仿真模拟移动床色谱吸附分离过程,以分离纯度和性能指标分别作为约束条件和目标函数进行多目标操作优化设计。仿真结果表明,SMB模型较之TMB模型更真实可靠,双种群遗传粒子群算法也较单一种群的遗传算法或粒子群算法具有更好的收敛性和鲁棒性,能有效地对模拟移动床操作条件进行优化,提高其分离性能和经济效益。
Both dynamic SMB model and steady TMB model were solved with finite elements method.Multi-objective genetic algorithm（GA） and particle swarm optimization（PSO） based on Pareto non-inferior solutions set were combined to optimize the operating condition.Double populations were chosen when one using GA and another using PSO and evolving independently but accompanied by individual migration.The results show that the SMB approach outperforms the TMB,and the GA and PSO algorithms have better convergence and robustness.The optimized operating conditions are effective to improve the separation performance.
出处
《化工自动化及仪表》
CAS
北大核心
2011年第4期427-431,共5页
Control and Instruments in Chemical Industry
基金
国家高技术研究发展计划863项目(2009AA04Z161)
