Multivariable Control of Systems with Multiple Hard Nonlinearities (Constraint Enforcement)
This area of research focuses on the development of systematic design methods for implementing real-time constraint enforcement within multivariable control systems.
Relevant application areas include:
Relevant Control Challenges
- Aerospace systems; e.g. hypersonic vehicles (waveriders/gliders); angle-of-attack (AOA), side-slip-angle (SSA), Mach number, aero-control surface position and rate deflection, and propulsion limiting
- Robotic systems; e.g. spacing and speed between cooperating robots
- Thermal processes; e.g. temperature limiting
- Low power DC-DC converters; e.g. duty cycle limiting
Relevant control challenges include:
Objectives and Goals
- model-based look ahead for variable saturation prediction
- maintaining multivariable bandwidth, directionality, command following, disturbance attenuation, noise attenuation properties of original control system (to the extant possible)
- control, state, and output limiting or constraint enforcement
- sensitivity to model uncertainty
- sensitivity to measurement uncertainty (i.e. sensor noise)
- complexity and real-time computation issues
The main objective of this research is to develop systematic methods for robust real-time constraint enforcement.
Model-based look ahead methods, observer-based methods,
generalized predictive control (GPC), model predictive control (MPC).
Collaborators and Sponsors
This work has been supported by the National Science Foundation (NSF), AFOSR, Eglin AFB, DARPA, Boeing, Honeywell, NASA.