Title:
Implementation and tuning of an extended expert control system for helicopter autorotation and development of a nonlinear model of electric drives to be used in the optimization of torque performance
Implementation and tuning of an extended expert control system for helicopter autorotation and development of a nonlinear model of electric drives to be used in the optimization of torque performance
Author(s)
Repola, Caroline R.
Advisor(s)
Rogers, Jonathan
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
This thesis covers two separate investigations under the topic of control. The first is the design and tuning of a fuzzy logic controller for Human-in-the-Loop (HITL) helicopter autorotation. The second is the exploration of an optimized pulse pattern for the control of an electric drive with focus on the development of the mathematical model of the drive.
Part One of this thesis discusses the autorotation controller. Helicopter autorotation is the operation a pilot performs when power is no longer supplied to the main rotor and an emergency landing is required. A controller was developed that allowed an autonomously controlled helicopter to perform an autorotation, an ‘expert skill’ more easily learned by human pilots. This controller is used in this thesis to create a tool that brings the computer and human together. The tuning process of the autorotation controller is described in detail. The controller used has five stages of operation; the transitions between these stages occur through a fuzzy logic determination. The results of the tuning bring about a successful autorotation in a simulated environment. The specific model of the controller developed in this thesis can be used in a different system to supply commands to a human pilot, aiding in the decisions during an autorotation.
Part Two of this thesis covers the development of the mathematical model of an electric drive and an optimization scheme to find a ‘better’ switching sequence for control. The goal of the model is to use it to find a better switching sequence, where better means fewer switching events as well as hitting targets of other key performance indicators (KPIs). The idea explored in this thesis is controlling the drive based on direct manipulation of the switches instead of indirectly through voltage or current. The mathematical model focusing on the switches is important to develop to facilitate the exploration of this control. Two different methods for developing this model are described. The first is a manually switched model based on examining every possible state of the drive. The second method is a non-smooth differential algebraic equation (DAE) approach, a more sophisticated mathematical approach that describes every state of the drive in one set of equations. An optimization scheme using model predictive control (MPC) is described. The focus of the optimization is the torque output of the motor and the number of switching events. The optimization would use the model developed in the thesis.
Sponsor
Date Issued
2017-12-11
Extent
Resource Type
Text
Resource Subtype
Thesis