unified communication & desktop video collaboration

National Aeronautics and Space Administration

www.nasa.gov

Model-Based Engine Control

Joe Connolly Amy Chicatelli & Sanjay Garg

[email protected]

3rd GRC PCD Workshop Feb. 28 – Mar. 1, 2012, Cleveland, OH

1


www.nasa.gov

Outline

• Overview

• Motivation • Approach

• Results

• Conclusions and Future Work

2


www.nasa.gov

Model-Based Control and Diagnostics

Ground Level

Engine Instrumentation • Pressures • Fuel flow • Temperatures • Rotor Speeds

Actuator Commands • Fuel Flow • Variable Geometry • Bleeds

Ground-Based Diagnostics • Fault Codes • Maintenance/Inspection Advisories

Validated Sensors

On-Board

Fault Detection and Isolation

Logic

Performance Estimates

Sensor Estimates

Sensor Measurements

Actuator Positions

“Personalized” Engine Control

On-Board Model & Tracking Filter

• Efficiencies • Flow capacities • Stability margin • Thrust

• Real-time performance estimation • Enhanced gas path diagnostics • Performance enhancing engine control


www.nasa.gov

MBEC Overview • Subsonic Fixed Wing (SFW) / Efficient Propulsion and

Power (EPP) Sub-project Objective: – Create, advance, and validate technologies and engine component

performance prediction tools that will reduce thrust specific energy consumption and contribute to the performance metric's of N+3 vehicle concepts

• Model-Based Engine Control (MBEC) helps achieve this objective by developing controls logic technology to allow more efficient operation of the engine with a tighter control of stability margins – MBEC Objective: Develop an on-board model based engine

control architecture and demonstration simulation for a commercial aircraft engine that provides reduced specific fuel consumption while maintaining safe operation throughout the engine life

4


www.nasa.gov

• Thrust cannot be measured and hence is indirectly controlled through regulating a measured variable which correlates with thrust – Fan Speed, Engine Pressure Ratio (EPR)

• Throttle to Thrust relationship changes over engine life • Surge Margins cannot be measured. Safe margins are indirectly maintained by the accel and decel limits

• Conservatism provided in selecting a “nominal” operating line to maintain safe margins throughout engine life

Typical Current Engine Control

5


www.nasa.gov

MBEC Motivation • The goal is to use an on-board “adaptive” model of the

engine to provide accurate estimates of unmeasured parameters that are used in the control system design – MBEC will enable operation of an engine with less conservative

safety margins, since all safety margins currently are designed to an end of life engine, whereas an on-board model can provide a more accurate margin estimate for the actual condition of the flight engine

• The longer-term pay-off is to have a “personalized” control for each engine, which adapts to the condition of the engine to not only maintain the most efficient operation throughout its lifetime, but also increase the useful operating life. – MBEC will enable consistent PLA to thrust performance for the

entire engine life cycle and possibly improve engine life expectancy by using an accurate estimate of thrust as a direct feedback variable for control through an on-board model

6


www.nasa.gov

MBEC What’s New? • MBEC Concept has been around since early 1990s

– The challenge has been to “adapt” an on-board model to the engine condition so as to get “accurate” enough estimates of “unmeasured” quantities of interest

• The “Optimal Tuner” based performance and condition estimation technique provides an innovative solution to this challenge – Simulation studies have shown that Thrust, Hot Section

Temperatures, and LPC/HPC Surge Margins can be “accurately” estimated even with engine degradation with usage

7


www.nasa.gov

MBEC Approach • Using C-MAPSS40k as the baseline, integrate an optimal-tuner

based on-board model and investigate new control architecture

Throttle (degrees)

Setpoint (EPR, Nf)

Error (EPR, Nf)

Wf (lb/s)

Outputs (EPR, Nf) (Nc, T, P)

Power Management

Setpoint Controller

Wf Cmd (lb/s)

Flight Condition (alt, Mn, Temp)

Actuators Engine Protection Logic

Sensors

Setpoint

Nf Max

Nc Max

Ps3 Max

Accel

Ps3 Min

RU Min

Min

Max

Protection Logic

• Provide direct Thrust control using estimated Thrust instead of regulating EPR/Nf

• Replace the Accel schedule with a HPC Stall Margin regulator using estimated SM

8


www.nasa.gov

SFW – EEP – MBEC Phased Approach

• Phase I: Linear Operating Point Control Design • Implement Optimal-tuner based model and evaluate estimation

accuracy for parameters of interest • Implement direct thrust control and investigate benefits • Implement HPC SM regulator and evaluate accuracy • In a pseudo-linear environment, investigate the capability to maintain

desired min SM with the accel schedule replaced by HPC SM regulator

• Phase II: Nonlinear Control Design • Develop the model-based control architecture over the whole operating

envelope • Investigate performance benefits while maintaining safety throughout

the engine operating life

• Phase III: Investigate potential benefits of using MIMO Model-Based Control

9


www.nasa.gov

Verification of the Optimal Tuner Estimation

• The C-MAPSS40k linear engine model is used as the truth model for the MBEC simulation and will subsequently be referred to as the engine

– The goal here is to illustrate that the estimation scheme will accurately model the engine thrust as shown in the plot.

Command Controller Engine Estimation Error

Thrust

Disturbance/ Health

External Inputs

Safety Margins

Fan Speed/Engine Pressure Ratio

Limiting Parameters

10


www.nasa.gov

MBEC Direct Thrust Control • A MBEC simulation is developed by integrating the continuous

time optimal tuner, thrust controller, and a linear operating point C-MAPSS40k model

Thrust Command Controller Engine

Optimal Tuner Kalman Estimation

Error

Est. Thrust

Disturbance/ Health

External Inputs

Surge Margin

Estimated Thrust

Limiting Parameters

True Thrust

Thrust variation with EPR Control Tight Thrust response with MBEC

11


www.nasa.gov

Surge Margin Estimation Analysis • Ability of the Optimal-tuner based model to accurately

estimate the HPC SM was analyzed. • The transient accuracy of the estimation is illustrated for a

2o PLA change with a 50% deteriorated engine

• The SM estimation accuracy is not sufficient for direct SM regulation – The optimal-tuners were

selected to minimize Thrust errors

• Need to redefine the cost function for optimal-tuner selection to include minimization of SM estimation error

12


www.nasa.gov

Analysis with Modified Optimal Tuners • With Optimal-tuner selection modified to include SM

estimation error in the cost function, a more accurate transient estimate of SM is obtained.

13


www.nasa.gov

MBEC Next Step • Implement a pseudo-linear implementation of accel

schedule and compare performance with MBEC implementation of min HPC SM regulator.

14


www.nasa.gov

Conclusions and Future Work • The demonstrated MBEC architecture using a linear C-MAPSS40k

engine model and a continuous time optimal filter has been shown to provide a tight thrust control given various engine deterioration levels

• The ability to provide direct control of thrust should provide improved on-wing engine life

• Investigation of the HPC SM estimation demonstrated that the Optimal-tuner selection has to be done with the cost function to be minimized reflecting the desired estimation objectives

• Current approach is to complete the linear model investigations by mid June 2012, and move on to nonlinear MBEC implementation on C-MAPSS40k

15

unified communication & desktop video collaboration

Documents