Diesel Engines Identification and Control
|Air management process in a turbocharged diesel engine is a multivariable, highly coupled nonlinear system with fast dynamics. Because of this, control algorithms with reasonably low computation times (enabling real-time application) must be used. Furthermore, testing new algorithms on a real engine is expensive. Therefore, a detailed nonlinear engine simulator based on a first-principles model must be developed. This kind of platforms provides an extensive range of possibilities in the experimentation field, because complex and innovative algorithms can be tested in a nondestructive way. Identification and control schemes based on model predictive control and local model networks are proposed in the frame of this project research.|
Fuel Cells Modelling and Control
|Fuel cell stack systems are under intensive development by several manufacturers, with the Proton Exchange Membrane (PEM, also known as Polymer Electrolyte Membrane) Fuel Cells (FC) currently considered by many to be in a relatively more developed stage for ground vehicle applications. There are three major control subsystem loops in the fuel cell system that regulate the air/fuel supply, the water management and the heat management. These tasks needs to be achieved fast and efficiently to avoid degradation of the stack voltage and sluggish net power response. Creating a control-oriented dynamic model of the overall system is an essential first step, not only for the understanding of the system behavior, but also for the development and design of model-based control methodologies.|
Embedded Real-Time Systems for Control Purposes
Embedded real-time systems are computer systems that control and react to time-critical real-world events. They are pervasive and include industrial control, telecommunications, military systems, avionics, medical equipment, and consumer electronic devices such as CD players and VCRs. Real-time systems are far more challenging to design, program, and debug than other systems because the compact systems allow little room for error.
Modern embedded systems, rely on the input from different disciplines like electronics, software engineering, IC design, sensor engineering, control theory, and radio communication during the design of these systems. These disciplines use different formalisms, such as discrete events, classes, state machines, differential equations, and electromagnetic fields. Moreover, the coupling with the electro mechanical and physical/chemical subsystems is essential for the overall system performance. Academic research is required to create models that can effectively bridge the gap between the different disciplines.
Many engineering design problems can be translated into multiobjective optimization (MO) problems. MO techniques present advantages when compared with single objective optimization techniques due to the possibility of giving a set of solutions with different trade-offs among different objectives that the problem comprises. This set of good solutions referred as nondominated solutions (none is better for all objectives) defines the Pareto set and the Pareto front (objective values for Pareto set solutions). In general, solving a MO problem is associated with the construction of the Pareto frontier. Each point of it represents one solution in the objective function space. Hence, given any pair of solutions as vectors of its objective function values, improvement in one of its components involves worsening in the others. In this way, no point is better than another in this frontier (non-dominated points), and the rest of the points are dominated by one or more points of this frontier (dominated points). Once the Pareto set is obtained, selection of a single solution is performed. This decision making process is a subjective and nontrivial procedure that depends on designer preferences.