薄膜反應器是結合反應與分離的程序強化單元,可藉由薄膜供應或分離反應物質,獲得較佳之反應器性能。混合薄膜反應器更進一步藉由間隔安排薄膜位置,提供額外的設計自由度。薄膜與混合薄膜反應器涉及非單一之性能指標,以及許多的設計與操作條件決定。 本研究針對兩種應用系統,乙烷加氧脫氫生產乙烯(Oxydehydrogenation of Ethane, ODH)與甲烷自熱蒸氣重組產氫(Autothermal Reforming of Methane, ATR),建立了薄膜反應器與混合薄膜反應器的一維擬均相數學模式,探討反應器內部的特性分佈,以及操作與設計參數影響。 本研究使用遺傳演算法,針對產量/選擇率、產量/轉化率與產量/選擇率/轉化率之雙目標函數與三目標函數,分別完成兩個系統之操作階段與設計階段多目標最佳化分析。 最佳化結果顯示,相較於薄膜反應器,使用混合薄膜反應器時,乙烷加氧脫氫與甲烷自熱蒸氣重組反應器之各項性能均可獲得提升,乙烯產量與氫氣產量分別可獲得約25%與一倍之提升。 Membrane reactor (MR) is a process intensification application, which integrates reaction and separation into one unit operation. By penetrating reaction species through a membrane, either adding into or extracting from the reactor, the performance of the reactor can be enhanced. Mixed membrane reactor (MMR) provides an extra freedom by allowing part of the reactor wall to be non-permeable wall. MR and MMR both involve non-single performance index as well as many decision-makings of design/operation conditions. In this study, two application systems of MR and MMR are investigated. They are the oxydehydrogenation of ethane (ODH) for the production of ethylene and the autothermal reforming of methane (ATR) for the production of hydrogen. One-dimensional pseudo-homogeneous models were developed for these two systems to study the internal profiles of the reactors as well as the effects of operation/design variables on the performance index. Genetic algorithm was employed to obtain the optimal solutions for binary objective functions (production rate/selectivity and production rate/conversion) and ternary objective functions (production rate/ selectivity/conversion) for the two systems using MR and MMR. Compared to MR, the optimal solutions of MMR provide better performance, the ethylene production rate and hydrogen production can be increased by 25% and 100%, respectively.