Generalized Predictive Control Parameter Adaptation Using a Fuzzy Logic Approach
Lloyd, John William
MetadataShow full item record
A method to adapt the Generalized Predictive Control parameters to improve broadband disturbance rejection was developed and tested. The effect of the parameters on disturbance rejection has previously been poorly understood and a trial and error method was used to achieve adequate results. This dissertation provides insight on the effect of the parameters, as well as an adaptive tuning method to adjust them. The study begins by showing the effect of the four GPC parameters, the control and prediction horizons, control weighting &lambda , and order, on the disturbance rejection and control effort of a vibrating plate. It is shown that the effect of increases in the control and prediction horizon becomes negligible after a certain point. This occurs at nearly the same point for a variety of &lambda 's and orders, and hence they can be eliminated from the tuning space. The control effort and closed-loop disturbance rejection are shown to be highly dependant on &lambda and order, thereby becoming the parameters that need to be tuned. The behavior is categorized into various groups and further investigated. The pole and zero locations of the closed-loop system are examined to reveal how GPC gains control and how it can fail for non-minimum phase plants. A set of fuzzy logic modules is developed to adapt &lambda with order fixed, and conversely to adapt order with &lambda fixed. The effectiveness of the method is demonstrated in both numerical simulations and laboratory experiments.
- Doctoral Dissertations 
Showing items related by title, author, creator and subject.
Integrating Collision Avoidance, Lane Keeping, and Cruise Control With an Optimal Controller and Fuzzy Controller Grefe, William Kevin (Virginia Tech, 2005-04-29)This thesis presents collision avoidance integrated with lane keeping and adaptive cruise control for a car. Collision avoidance is the ability to avoid obstacles that are in the vehicleâ s path, without causing damage ...
Kirchner, William (Virginia Tech, 2012-03-22)Human expert drivers have the unique ability to build complex perceptive models using correlated sensory inputs and outputs. In the case of longitudinal vehicle traction, this work will show a direct correlation in ...
Modeling and Simulation of a Video-on-Demand Network Implementing Adaptive Source-Level Control and Relative Rate Marking Flow Control for the Available Bit Rate Service Taylor, Elvin Lattis Jr. (Virginia Tech, 1997-12-18)The Available Bit Rate (ABR) service class for the Asynchronous Transfer Mode (ATM) protocol was originally designed to manage data traffic. ABR flow control makes no guarantees concerning cell transfer delay or cell delay ...