Optimization of an Air Pressure System: A Multi-Objective Control and Modeling Approach

This paper presents the application of a methodological approach based on multi-objective optimization for identifying a nonlinear model and designing controllers for an air pressure system. First, the parameters of the system's nonlinear model are determined based on experimental data obtained from the process. Subsequently, the methodology is extended to tuning cascade PI controllers, where experimental tests are performed in the laboratory. This methodology allows a designer to have a detailed view of the proposed multi-objective problem (both in the modeling and controlling the air pressure system). Depending on the level of detail required, the design objectives of the multi-objective problem are proposed. In this case, two objectives are proposed to evaluate the model's performance and three objectives for evaluating the control system's quality. Among the advantages offered by this methodology is the easy understanding of the conflicts that arise between the design objectives, which allows for selecting an optimal solution according to the preferences established by a designer and based on detailed information on the system's performance. In this case, a conflict between the dynamics of the pressures of the air pressure system could be evidenced, both in the modeling and in the control. The proposed methodology also allows the analysis of the behavior of the air pressure system in a global framework without loss of information, which would not be feasible in a traditional single-objective analysis that agglutinates all the information in a single design objective. Although the decision-making stage of the proposed methodology is more complex than a single-objective analysis, it is justified as it provides detailed and complete information to a designer to make accurate and effective decisions on how to model and control the compressed air pressure system effectively.