Abstract
When an optimal control problem requires an important computational effort, a metaheuristic algorithm (MA) can be a useful approach. An MA is conceived to solve a specific optimal control problem having a characteristic objective function. This algorithm solely yields only an optimal offline solution. The desideratum to have a closed-loop implementation can be fulfilled through a supplementary “tool”, the Receding Horizon Control (RHC) structure. This paper addresses a particular case and integrates Evolutionary Algorithms into the RHC structure. The main objective is to propose a general harmonization between the Controller of the closed loop and the Evolutionary Algorithm. Some details concerning the implementation of the closed loop and Controller are described. The impact of the RHC’s prediction technique upon the control sequences’ encoding is also analyzed. Two general structure Controllers are proposed, one of them conceived to cope with restrictive time constraints. Practical ideas have been illustrated through a case study: the well-known optimal control of a fed-batch reactor for ethanol production. This time, our implementation achieves a closed-loop solution. The results from the programs and simulation series validate the Controllers, EAs, and the closed-loop structure. Generally speaking, the association between RHC and EA can be a realistic solution to optimal process control.
Highlights
Optimal control of a dynamic system is a usual task in process engineering
A metaheuristic algorithm (MA), like simulated annealing, genetic algorithm, evolutionary algorithm (AE), ant colony systems, particle swarm optimization, etc., is not by itself prepared to achieve a closed-loop control structure. That is why it needs to be integrated into a structure like Receding Horizon Control (RHC) or Model Predictive Control ([11,12])
This paper presented some implementation aspects regarding the optimal control of a process using a Controller based on an evolutionary algorithm
Summary
Optimal control of a dynamic system is a usual task in process engineering. If the process has mathematical properties sufficient to apply theoretical control laws, this task can sometimes be achieved without significant computational complexity. Other times the computational effort is important, or we do not dispose of known control techniques In this case, metaheuristic algorithms integrated within appropriate control structures constitute a realistic solution. A metaheuristic algorithm (MA), like simulated annealing, genetic algorithm, evolutionary algorithm (AE), ant colony systems, particle swarm optimization, etc., is not by itself prepared to achieve a closed-loop control structure. That is why it needs to be integrated into a structure like Receding Horizon Control (RHC) (see [8,9,10]) or Model Predictive Control ([11,12]). Both structures use prediction techniques that could involve MAs because prediction is usually optimization
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
