Date of Completion

5-20-2011

Embargo Period

5-16-2011

Advisors

Hanchen Huang; Robert Gao

Field of Study

Mechanical Engineering

Degree

Master of Science

Open Access

Open Access

Abstract

L1 adaptive control for a turbofan commercial aircraft engine is developed and applied in the presences of constraint limits. Turbofan engines are a highly complex system, with a wide variety of dynamics, operating under constantly changing environmental conditions. In aircraft turbofan operation many constraint requirements must be met. When constraints, such as rotor speeds, pressures, and temperatures, exceed the prescribed limits it can decrease the life expectancy of the engine, cause component failure, or even total engine failure. Additionally stall, also known as compressor surge, which occurs when the compressor is unable to work against incoming air resulting in flow reversal, must be avoided. To decrease the chances of entering into surge or stall, control systems are developed to control stall margin, which are percentage estimates of the probability that an engine will enter surge or stall. Overall, the control of a turbofan engine is a highly nonlinear problem that must be capable of handling variable constraints so an effective controller must be capable of handling nonlinearities while operation within constraints.

Despite the nonlinearity of the engine control problem, the aero-engine industry achieves engine control using linear design logic by creating several controls at different operating points and scheduling the resulting gains based on these different operating ranges. Therefore, this thesis proposes a L1 adaptive controller to handle nonlinear uncertain systems in the presence of constraint variables. These constraint variables are maintained through a dynamic integration limiter. adaptive control theory permits transient characterization, deals with time varying uncertainties, and can create a tradeoff between tracking performance and robustness.

The theoretical foundation for the adaptive controller is developed and analyzed, genetic simulations are conducted for the theoretical controller, as well as applying the controller to a Simulink model of a twin-spool commercial aircraft engine. Additionally, special consideration is placed on a stall margin estimator and used as a limiting variable. Simulation results are found to verify the theoretical findings.

Major Advisor

Chengyu Cao

Share

COinS