<p>We address a selection problem between two microbial species by controlling the dilution rate in a continuous stirred tank reactor. Unlike earlier studies, we focus on the competition between species for two substitutable substrates instead of a single limiting resource, with each species capable of dominating the culture. The selection of the most suitable strain is formulated as a minimum-time optimal control problem. After examining the reachability of the target set, we conduct a thorough analysis of optimal solutions and singular arcs, using Pontryagin’s Maximum Principle and geometric control theory. Notably, unlike studies involving a single substrate, the target is not always reachable, and we construct a counterexample to prove it. In the case where the target can be reached, we highlight that optimal solutions exhibit turnpike-like phenomena by combining the approach with a direct optimization method, and these properties differ from those in the single-substrate case. Although there is no natural static optimization problem associated with a minimum-time control problem, our analysis enables us to characterize the turnpike point as the solution to a static optimization problem associated with an auxiliary Lagrange-type optimal control problem. These results are extensively illustrated throughout this work using direct optimization.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Optimal Separation of Two Microbial Species Competing for Two Substitutable Resources

  • Walid Djema,
  • Térence Bayen,
  • Jean-Luc Gouzé

摘要

We address a selection problem between two microbial species by controlling the dilution rate in a continuous stirred tank reactor. Unlike earlier studies, we focus on the competition between species for two substitutable substrates instead of a single limiting resource, with each species capable of dominating the culture. The selection of the most suitable strain is formulated as a minimum-time optimal control problem. After examining the reachability of the target set, we conduct a thorough analysis of optimal solutions and singular arcs, using Pontryagin’s Maximum Principle and geometric control theory. Notably, unlike studies involving a single substrate, the target is not always reachable, and we construct a counterexample to prove it. In the case where the target can be reached, we highlight that optimal solutions exhibit turnpike-like phenomena by combining the approach with a direct optimization method, and these properties differ from those in the single-substrate case. Although there is no natural static optimization problem associated with a minimum-time control problem, our analysis enables us to characterize the turnpike point as the solution to a static optimization problem associated with an auxiliary Lagrange-type optimal control problem. These results are extensively illustrated throughout this work using direct optimization.