Main research fields
Control theory
We develop new algorithms for the robust control of complex dynamical systems, particularly those exhibiting discontinuities, nonlinearities, constraints, scarce measurements, cross-coupling, etc.
-
In every real control loop there exist physical limits, structural constraints or dynamic restrictions that affect the achievable closed-loop performance. In multivariable systems, the effects and demands of these “wide sense” system constraints are worsened because of the directionality and crossed interactions present in this kind of plants. We seek for solutions to this problem mostly based on Sliding Mode Reference Conditioning (SMRC) approaches, both by delimiting a given signal in the loop or by imposing virtual system constraints. -
High order sliding mode (HOSM) algorithms reduce the chattering problems inherent to conventional sliding mode control while preserving most of its robustness features. Our research is oriented to develop modified versions of the classical HOSM algorithms as well as new design tools by using time scaling and coordinate transformations.
Control of renewable energy systems
We develop new algorithms for the control of renewable energy systems. We are particularly interested in the alleviation of mechanical loads in wind turbines and in the integration of renewable energies to the grid.
-
Commercial wind turbines are usually controlled to capture as much as energy as possible from the wind, limiting the power efficiency at high wind speeds to avoid overloading. This is typically achieved by collectively pitching the blades. Gain scheduled PI controllers are conventionally used to command the pitch actuators. Because of their inherent limitations, they can hardly be tuned and optimized to satisfactorily accomplish more demanding new objectives like active power control and mitigation of the mechanical loading. We seek for higher performance control systems based on advanced control techniques such as Linear Parameter Varying control, Sliding Mode control and passivity based control. New topologies, machines and control structures are also investigated in order to improve operation under grid fault conditions or and to provide ancillary services like grid stability support. -
Renewable energy systems connected to isolated grids and microgrids usually comprise several energy resources (wind, photovoltaics, etc.) and storage devices (electrolyzers, supercapacitors). These system components are coupled by means of power electronic devices, which play a fundamental role in the integration of the system to the grid. Sliding mode control offers a very attractive, simple and high-performance option to control the electronic converters. We investigate novel system topologies and control strategies to maximize renewable production of electricity and hydrogen while minimizing the impact on the grid power quality and fulfilling the inherent restrictions of devices operation.
Control of biological processes and systems
Some of the main difficulties in the control of biological systems are the nonlinearities and uncertainties in their dynamic behavior, as well as the lack of on-line and accurate measurements. We develop robust observers and controllers in an effort to overcome these problems.
-
Automatic and semi-automatic regulation of glycemia in type 1 diabetes mellitus patients is a long standing control challenge. The main drawback of both open- and closed-loop therapies is the excess of insulin delivered after the meals, which in many cases produces severe hypoglycemia events. Intra-patient variability, delays and lack of accurate on-line blood glucose level measurement impose severe limitations to the achievable performance. We develop control algorithms to reduce the events of hypoglycemia without increasing the risk of hyperglycemia. -
Biotechnology uses enhanced and engineered microorganisms as cell factories to produce high-added values specialty metabolites (e.g. amino acids, vitamins, food additives, biofuels, biofilms, tissues). Biotechnological engineering is of paramount importance for the future of health, chemical, food and other process industries, as well as for the environment. Yet, the current state-of-the-art, characterized by uncertainty and lack of in-depth real-time knowledge about the process state, forces industry to operate their bioprocesses at too conservative, suboptimal and not intensified regimes, so as to avoid undesirable microorganism physiological states. Such practices cause poor efficiency, lack of process stability and increased waste of product and energy. In order to surmount these difficulties, there is a need of identifying desirable metabolic states and of developing bioreactor optimization, monitoring and control methods so as to lead the system to the desired state in the course of a process. The members of our group have been participating in different projects within this line, ranging from biomass growth in fed-batch reactors to lab-scale production of bioplastics, microalgae and tissue.
Control of autonomous systems
In the last years, we have opened a new research line on control and guidance of autonomous systems, either marine, terrestrial or aerial. One-side or high-order sliding mode algorithms have been mainly explored.
-
Problems of both single and multiple robotic systems have been tackled from the perspective of constrained control, among them:
-
Path following speed auto-regulation in terms of joint constraints.
-
Robust and dynamic robot workspace delimitation.
-
Redundancy resolution and trap avoidance schemes.
-
Reactive collision avoidance.
-
Multiple robots task priority and coordination policies.
-
Coordination of multi-agent systems (AUV’s).
-
