Barron Associates, Inc. Selected Current Research SAE International - - PowerPoint PPT Presentation

▶
barron associates inc selected current research
SMART_READER_LITE
LIVE PREVIEW

Barron Associates, Inc. Selected Current Research SAE International - - PowerPoint PPT Presentation

Jul 07, 2023 391 likes •696 views

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007 Barron Associates, Inc. Selected Current Research SAE International Aerospace Control & Guidance Systems Committee Boulder, Co Feb 28, 2007 David G. Ward (434) 973-1215

slide-1
SLIDE 1
  • 1-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Barron Associates, Inc. Selected Current Research

SAE International Aerospace Control & Guidance Systems Committee

Boulder, Co Feb 28, 2007 David G. Ward (434) 973-1215 ward@barron-associates.com

slide-2
SLIDE 2
  • 2-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

10 20 30 40 50 60

  • 5

5 10 15 20 25 30 35 Time (sec.) Pitch Attitude (deg.) Commanded Response Response (= 1/3 sec.) Response (= 3 sec.) Response (= 30 sec.) Response (=  sec.)

Adaptive Control

2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0

 = 0  = 0 . 1  = 0 . 1 5

*in s t a b ilit y o c c u rs fo r  > 0 . 1 7

2 4 6 8 10 12 14 16 18 20 Time (sec.) Pitch Attitude (deg.) 25 20 15 10 5

  • 5

Response (= 6.7 sec.) Response (= 10 sec.) Response (=  sec.) Shape Control Flight Control Objective

Goal: Stable flight control with limited model knowledge during wing-shape morphing

Conventional Control: Result: Inconsistent response and instability for faster morph times (>6 sec.) Result: Consistent stable response

Morph and maneuver initiated at 0 sec. Morph and maneuver initiated at 15 sec.

CFQ KFQ KI/s BPL APL KPL CPL

  • BFQ

AFQ 1/s xPL

 c

xFQ e 1/s 

xPL=[  q]T

CFQ KFQ KI/s BPL APL KPL CPL

  • BFQ

AFQ 1/s xPL

 c

xFQ e 1/s 

xPL=[  q]T

KP + KI/s P(s)

c

e  KP + KI/s P(s)

c

e 

Adaptive Control of Morphing Aircraft

AF AF SBIR SBIR Ph Phase II

  • With
  • NextGen / Northrop Grumman / VA Tech
  • Bryan Cannon, COTR

Dem Demonstr trati tion Goal

  • Real-time wind-tunnel demonstration of stable

morphing control using N-MAS wing

slide-3
SLIDE 3
  • 3-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Develop a general-purpose automated-recovery system approach that

learns appropriate recovery strategies adopts/encodes best-practices from the manned aircraft community avoids out-of-control conditions to the extent possible takes advantage of innovative actuation concepts

CUPR

NATOPS, Established Recovery Procedures, Etc. Manned Aircraft Flight Data, Piloted Simulations NATOPS, Established Recovery Procedures, Etc. Manned Aircraft Flight Data, Piloted Simulations

High-Fidelity Simulation

EAGLE EY E EAGLE EYE

UAV Upset Recovery Control Systems

COTR: Mr. Jim Busey, NAVAIR Collaboration w NASA LaRC

(Drs. Christine and Celeste Belcastro)

slide-4
SLIDE 4
  • 4-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

  • 0.2

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 5 10 15 20 25 30

Multi-input Multi-

  • utput Control Law

Control Allocation Onboard Models Cmnds

  • 0.5

0.5 1 1.5 2 2.5 3 3.5 4 4.5 2 4 6 8 10 12 14 16 18 20

Heading Overshoot Depth Overshoot

Phase I Sponsor:

  • Dr. Edward Ammeen

Head, Maneuvering and Control Division Naval Surface Warfare Center, Carderock Tel: (301) 227-5907

Innovative Methods for Optimally Mixing a Diverse Suite of Control Effectors for Marine Vehicles

  • Form

rmalize inner-loop contr trol design meth thodology to be applicable to any vehicle wi with th minimal re reconfi figurati tion and tuning - fe feedback lineari rizati tion and backste tepping

  • Dev

Develop model based tuning approaches

  • Dev

Develop meth thods to esta tablish sta tability ty, ro robustness and perf rform rmance characteri risti tics of inner-loop approaches

  • Dev

Develop state te esti timati tion appro roaches

  • Dev

Develop oute ter-loop guidance design meth thodology that can be applied to a vari riety ty of vehicle platf tforms

  • Dev

Develop path th planning algori rith thm

  • Ap

Apply design meth thodology to to multiple vehicle models and evaluate te perf rform rmance

Phase II Acceptance Bound Phase II Acceptance Bound

Ph Phase II II Pl Plan

M

slide-5
SLIDE 5
  • 5-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

NON-PROPRIETARY DATA NASA SBIR/STTR Technologies Identification and Significance of Innovation Technical Objectives and Work Plan NASA and Non-NASA Applications

Damage Adaptation using Integrated Structural, Propulsion, and Aerodynamic Control

PI: D. Ward, Barron Associates, Charlottesville, VA

Topic A1.02 Integrated Resilient Aircraft Control -- Submittal No. A1.02-9516 Built on flight-test-proven reconfigurable control algorithms Compute feasible control using aerodynamic and propulsive “effectors” Compute safe operating envelope in real-time using real-time structural health monitoring

  • No excessive loads on damaged components
  • No excitation of new structural modes

Path-plan for safe landing

  • No unachievable trajectories or autopilot commands
  • No violation of structural limitations

Can be implemented using V&V’able architectures Contact: 434-973-1215

ward@barron-associates.com

Structural Health Monitoring Structural Health Monitoring Aerodynamic System ID Aerodynamic System ID Engine Diagnostics Engine Diagnostics

Integrated Damage- Adaptive Control Integrated Damage- Adaptive Control

(2) Landing Trajectory Achievable, Safe… (1) Yoke, Pedal, and Thrust Commands Achievable, Decoupled, Safe… Structural Health Monitoring Structural Health Monitoring Aerodynamic System ID Aerodynamic System ID Engine Diagnostics Engine Diagnostics Structural Health Monitoring Structural Health Monitoring Aerodynamic System ID Aerodynamic System ID Engine Diagnostics Engine Diagnostics

Integrated Damage- Adaptive Control Integrated Damage- Adaptive Control

(2) Landing Trajectory Achievable, Safe… (1) Yoke, Pedal, and Thrust Commands Achievable, Decoupled, Safe… Structural Health Monitoring Structural Health Monitoring Aerodynamic System ID Aerodynamic System ID Engine Diagnostics Engine Diagnostics

Integrated Damage- Adaptive Control Integrated Damage- Adaptive Control

(2) Landing Trajectory Achievable, Safe… (1) Yoke, Pedal, and Thrust Commands Achievable, Decoupled, Safe… Structural Health Monitoring Structural Health Monitoring Aerodynamic System ID Aerodynamic System ID Engine Diagnostics Engine Diagnostics Structural Health Monitoring Structural Health Monitoring Aerodynamic System ID Aerodynamic System ID Engine Diagnostics Engine Diagnostics

Integrated Damage- Adaptive Control Integrated Damage- Adaptive Control

(2) Landing Trajectory Achievable, Safe… (1) Yoke, Pedal, and Thrust Commands Achievable, Decoupled, Safe…

Compensate for Simultaneous Effector, Airframe, and Propulsion Damage Compensate for Simultaneous Effector, Airframe, and Propulsion Damage Simulation Demonstration of Integrated Damage Adaptive Control System Simulation Demonstration of Integrated Damage Adaptive Control System Work Tasks Define demonstration problem (GTM / AirSTAR?) Integrate representative health-monitoring system Develop integrated damage-adaptive controller Integrate autopilot and path-planning approaches Simulation demonstration Civil Aviation, Military Aviation, Space, Life-Extending Control, … Civil Aviation, Military Aviation, Space, Life-Extending Control, … Improved aircraft safety for civilian aviation Improved autonomous operations for space exploration in environments with massive uncertainties Improved autonomous operations for military vehicles (air, ground, surface, underwater, …) Life-extending control

  • Use during normal conditions to reduce wear/fatigue on

key structural components

slide-6
SLIDE 6
  • 6-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

IAG&C Through RLV Flight Envelope In Involved in all fli light phase ses

Curre rrent fo focus is on FA FAST ST

Altitude Lift-Off 1st Stage Separation 1st Stage Recovery 2nd Stage Coast 2nd Stage Burn Orbit Insertion De-Orbit Burn Entry TAEM Approach/Landing 2nd Stage Abort/Recovery Time

Mated vehicle 1st stage 2nd stage

Air Force’s Fully Reusable Access to Space Technology (FAST) Program

slide-7
SLIDE 7
  • 7-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

IAG&C for RLVs during Re-entry Go Goals/O s/Obje ject ctive ves

  • Adapt for effector failures
  • Integrate with existing RCS systems
  • Manage heating constraints, etc.

AFRL BAA w/ Boeing (Anhtuan Ngo, COTR)

slide-8
SLIDE 8
  • 8-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Architecture Archite tect cture for re-entry tr trajecto ctory command generation

ref

q

mag

 +

  • q

cmd

,  

3-DOF Plant Model 3-DOF Plant Model

Reconfigurable Control System Reconfigurable Control System Adaptive Guidance System Adaptive Guidance System Trajectory Command Generation Trajectory Command Generation

Longitudinal

  • Traj. Algorithm

Longitudinal

  • Traj. Algorithm

Lift, Drag, …

Lateral Traj.

Algorithm

Lateral Traj.

Algorithm

Cmd. Traj. States

Onboard Parameter Estimation Current estimates of lift, drag coefficients Dynamic Pressure Profile Reshaping

slide-9
SLIDE 9
  • 9-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Dynamic Pressure Profile Reshaping Commanded tr trajecto ctory dete termined by reshaping dynamic pressu ssure profile

Max glide – longest achievable trajectory - limited by heat load constraint Max dive – shortest achievable trajectory - limited by heat rate constraint

1 1.5 2 2.5 x 10

5

20 40 60 80 100 120 140 160 Altitude ~ Ft. Dynamic Press ure ~ Slug/ft s

2

Nominal Max Glide Max Dive

psf

q 

Increasing this parameter will move reference commands toward max dive trajectory with steeper than nominal glide slope Decreasing this parameter will move reference commands toward max glide trajectory with shallower than nominal glide slope

q 

Center of alt. range

x 103 8 7 6 5 4 3 2 1

Dynamic Pressure ~ N/m2

3.0 4.6 6.1 7.6

Altitude ~ m

x 104

slide-10
SLIDE 10
  • 10-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Re-entry Trajectory Retargeting In In-flight reta targeting to alte ternate te landing sites possi ssible

  • w/o

w/o exte tensive pre re-mission planning

  • 110
  • 105
  • 100
  • 95
  • 90
  • 85
  • 80
  • 75
  • 70

16 18 20 22 24 26 28 30 32 Longitude ~ Deg. Latitude ~ Deg. HAC=(21N,87W) HAC=(29N,73W) HAC=(32N,85W)

Ground Track

Rwy Rwy Rwy

X X X

Interface Point Runway Location

slide-11
SLIDE 11
  • 11-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

IAG&C Through RLV Flight Envelope

Altitude Lift-Off 1st Stage Separation 1st Stage Recovery 2nd Stage Coast 2nd Stage Burn Orbit Insertion De-Orbit Burn Entry TAEM Approach/Landing 2nd Stage Abort/Recovery Time

slide-12
SLIDE 12
  • 12-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

IAG&C for RLVs During Ascent

Dev evelop

  • p gu

guidan ance/pa path pl planni ning for

  • r:

Safe stage separation “Turn around” of vehicles Return-To-Launch-Site (RTLS) guidance law development

  • 1st stage fly back
  • 2nd stage fly back

Work rking w/ Ping Lu

1st stage fly back

AFRL BAA w/ NGC (Mike Oppenheimer, COTR)

slide-13
SLIDE 13
  • 13-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

IAG&C for RLVs During Ascent

Dev evelop

  • p gu

guidan ance/pa path pl planni ning for

  • r:

Safe stage separation “Turn around” of vehicles Return-To-Launch-Site (RTLS) guidance law development

  • 1st stage fly back
  • 2nd stage fly back

Work rking w/ Ping Lu

1st stage fly back Less risk: RTLS – should involve entry, TAEM, approach/landing designs

(potential issues: time constraints in achieving proper conditions for good TAEM soln. etc.)

AFRL BAA w/ NGC (Mike Oppenheimer, COTR)

Most technical risk: Guidance/control design for stage separation and proper maneuvering to begin RTLS

slide-14
SLIDE 14
  • 14-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Progress to Date

  • P. Lu’s Approach at TRL 1

Progress to Date (TRL approaching 2)

  • Code completely reconstructed to allow

integration into Northrop’s simulation environment

  • Convergence issues addressed
  • Numerous Monte Carlo experiments

successfully completed

Engine performance variations/wind profile dispersions

Near Term Plans

  • Working on autonomous decision making

Coasting - vs - no-coasting option (key attribute under abort scenarios – especially for engine-out cases)

  • Developing adaptive elements

28.5o inclination

Ground Track View

51.6o inclination

500 1000 1500 2000 2500 3000 3500 200 400 600

Altitude vs downrange

dow nrange(km) alt itude (km) 200 400 600 800 2000 4000 6000 8000

Inertial velocity

V (m/s) time (s) 185.2 km orbit 500 km orbit

1st stage burnout 2nd stage ignition 2nd stage ignition

coast coast 1st stage burnout 2nd stage ignition 2nd stage ignition Altitude vs Downrange Velocity History Launch Profiles for Various Coast Times

Sample 3-DOF Results:

slide-15
SLIDE 15
  • 15-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

IAG&C Through RLV Flight Envelope

Altitude Lift-Off 1st Stage Separation 1st Stage Recovery 2nd Stage Coast 2nd Stage Burn Orbit Insertion De-Orbit Burn Entry TAEM Approach/Landing 2nd Stage Abort/Recovery Time

slide-16
SLIDE 16
  • 16-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Rapid Mission Planning and Optimization Rapid, relia iable resp sponse

  • Mission ready within 2 hours, 24 hours a day, 7 days a week

Tool to to enable rapid planning …

RMP& O “Toolbox” RMP& O “Toolbox”

Vehicle specific properties Mission specific parameters Current atmospheric conditions Other launch conditions/constraints Map of feasible trajectories &

  • ther required

information Map of feasible trajectories &

  • ther required

information As automated as possible IVHM information ?? Choice of launch vehicles

  • r 2nd stage

weapons Choice of launch vehicles

  • r 2nd stage

weapons

AFRL SBIR Phase II (Michael Bolander, COTR)

slide-17
SLIDE 17
  • 17-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Phase II Start

Main Ef Effo forts rts Pro Proposed:

  • Fu

Fully develop ra rapid mission planning software ware to toolbox: “Th The Mission Planner” shall be complete ted to to a high TRL TRL to to be used by AF & industry try

  • “Real-world” demonstration of the Mission Planner: a complete mission plan shall

be completed

  • “FAST launch vehicle” simulation model developed for demonstrations
  • Pr

Primary ry Fe Feature res of th the Mission Planner: r:

  • Employ trajectory generation algorithms developed in past AF IAG&C programs:

very fast optimization routines designed for in-flight use – for full mission:

  • Ascent / On-orbit / Entry / TAEM / A&L
  • Built-in analysis capabilities:
  • choice of model fidelity & vehicle characteristics
  • Monte Carlo/dispersion analyses & trade studies
  • Fault/failure modes analysis
  • Hybrid / mission optimization
slide-18
SLIDE 18
  • 18-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Current Progress

Simulation Platform Generate Trajectory Libraries for Each Flight Phase

Launch & Ascent On-orbit / De-orbit Burn Re-entry Post-Entry (TAEM, A/L, Terminal)

User Defined Studies Perform Specific Trade Studies Monte Carlo Experiments Abort/Failure Scenarios Worst-on-Worst Analysis Specific Dispersions Nominal Case Studies

2nd Stage Vehicle Model (6-DOF) 2ndStage Vehicle Model (6-DOF) 2nd Stage Vehicle Model (6-DOF) 2nd Stage Vehicle Model (3-DOF) 1st Stage Vehicle Model (6-DOF) 1st Stage Vehicle Model (6 -DOF) 1st Stage Vehicle Model (6-DOF) 1st Stage Vehicle Model (3-DOF) 2nd Stage Vehicle Model (6-DOF) 2nd Stage Vehicle Model (6-DOF) 2nd Stage Vehicle Model (6-DOF)

Choose Vehicle Model(s) Initialization Define Problem: Requirements, Constraints Initial/End Conditions … Global/Hybrid Mission Optimization Present Results Plots, Tabulated Data, … Footprints, Trajectories, … Present Results Plots, Tabulated Data, … Footprints, Trajectories, …

1st Stage Vehicle Model (6-DOF) 1st Stage Vehicle Model (6-DOF)

Vehicle Model Library Build Model Variations Vehicle Characteristics

Mass/Geometric Properties Engine Performance Aerodynamic Data 2nd Stage Vehicle Model (6-DOF) 1st Stage Vehicle Model (6-DOF) 1st Stage Vehicle Model (6-DOF)

Supporting Models Winds, Atmosphere, Subsystem Models (Actuators, Sensors, …) Mated Vehicle 1st Stage 2nd Stage Decisions Choice of 1st, 2nd Stages Mission Requirements Fuel Required, etc… Decisions Choice of 1 st, 2nd Stages Mission Requirements Fuel Required, etc…

Software toolbox architecture development Graphical User Interfaces Constructed

slide-19
SLIDE 19
  • 19-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Preliminary Experiments Conducted

  • Nominal asc

scent & desce scent stu studies

Mission Trade Studies Engine performance tradeoffs for ascent Mass tradeoffs for reentry & terminal guidance

  • Anomaly Stu

tudies

Engine failure during 1st stage burn Effects of mass-to-orbit margin studied Drag anomalies during reentry & terminal guidance

slide-20
SLIDE 20
  • 20-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Backup Module System Module Safety Wrapper Backup Module System Module Safety Wrapper

Safety Wrapper Backup Module

Generic Run-Time V&V Safety Wrappers

Backup Module System Module Safety Wrapper

Verification Data Validation Data

Inputs Outputs

Order of operation checks, etc check input/output bounds, system behavior, etc.

Generic individual safety wrapper for one system module Safety wrapper for overall system comprised of a multitude of subcomponent safety wrappers Incremental degradation: shut down only those sub- components not working, allowing other advanced components to continue

  • peration

System (Aircraft) SW Implementation SW Design Failure Classes that can be Accommodated

AF SB SBIR Pha Phase II

  • With Lockheed Martin
  • Wendy Chou, COTR

Demonstra tration Goal

  • Real-time HIL demonstration of V&V of

intelligent UAV mission planning and control software

slide-21
SLIDE 21
  • 21-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Problem Statement: Develop a general-purpose MATLAB™-based modeling and simulation environment to support Air Force test and evaluation efforts …

AUSPEX Modeling Environment

Sponsor: Air Force Flight Test Center TPOC: David Kidman, 773 TS / ENFP Scope / Enabling Technologies: Concept:

Model Development: Automated local modeling / data partitioning Structure learning polynomials and hybrid (empirical-CFD-first principles) models Model Calibration: Parameter estimation to update model parameters to match test data Model Updating: Hypersurface blending to merge local model results with existing databases

1 2 3 4 Model Comparison Model Development Model Calibration Model Use

Data Quality Assurance Data Quality Assurance Data Quality Assurance Data Quality Assurance Data Quality Assurance

WT-Based Model FT-Calibrated Model (incl. Addt’l Dyn.) WT-Based Model WT-Based Model FT-Calibrated Model (incl. Addt’l Dyn.) Wind Tunnel (WT) Data Addt’l WT Data Flight Test (FT) Data FT Data Addt’l FT Data

Algorithms & flexible high-level framework to enable development and/or updates of simulation databases with test data…

slide-22
SLIDE 22
  • 22-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

AUSPEX Modeling Environment

Sponsor: Air Force Flight Test Center TPOC: David Kidman, 773 TS / ENFP Current Application:

Engine Inlet Modeling …

F/A-22: F119 Engine Inlet (Barron) F-35: F135 Engine Inlet (AEDC)

Engine Model Calibration…

Ongoing Application : Test Data Contractor High-Fidelity Transient Thrust Deck

Interactive analysis / data comparison Output-error tuning of thrust deck free parameters to match test data

slide-23
SLIDE 23
  • 23-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

CAESAR CAESAR

Control-law A Automated E Evaluation through Simulation-based and A Analytic Routines

Identification & Significance of Innovation

V&V Major Barrier to Advanced Control Laws

  • many advanced control approaches work well in tests
  • certifying for fleet vehicles very difficult

Automated Test is Key Technology

  • V&V is expensive and time consuming
  • Automated test key to making V&V manageable

Approach

Combine Analysis and Simulation-based Test

  • use worst-case analysis tools to gain insight
  • select simulation-based test conditions based on results

Provide Uncertainty Tools for Simulink

  • uncertainty blocks for non-linear Simulink models
  • linearization capability for uncertain Simulink models

Design general SW framework

  • facilitate design, analysis, V&V of control systems
  • permit performance validation for arbitrary systems

Open-Architecture MATLAB Implementation

  • plug in user analysis and app-specific performance SW
  • body of advanced algorithms and toolboxes in MATLAB

NASA Applications

Design, Analyze, & Validate Technologies for…

  • AvSP/SAAP
  • VSP/AuRA
  • SATS
  • ARES (Mars Airplane)

Non-NASA Applications

Design, Analyze, & Validate Technologies for…

  • commercial, & GA aircraft safety
  • military aircraft survivability & performance
  • UAV autonomy
  • adaptable robots, power systems, etc.

Source: NASA LaRC

slide-24
SLIDE 24
  • 24-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

ROME Real-time Observation-based Margin Estimation

Observed Behavior Plant Model Uncertainties Instability margin model mismatch Plant Model “used-up” margin “remaining” margin Mapping Function

Estimate transfer functions during flight test

Compute confidence intervals for estimates Incorporate additional data as test progresses Compare to off-line worst-case margin estimates

slide-25
SLIDE 25
  • 25-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Phase II Completion

  • Demonstr

stration of several run-ti time V&V monito tors

  • Shut down effector blender – revert to classical control assignment
  • Shut down dynamic inversion – revert to classical P-I-D controller
  • Command hedging – reduce desired dynamics to FQ Level II or III
  • Shut down complex logic in outer-loop guidance – command return to

base waypoint following

waypoint n ‘infinite’ turns

A

waypoint n ‘a few’ turns

B

waypoint n ‘infinite’ turns

A

waypoint n ‘infinite’ turns

A

waypoint n ‘a few’ turns

B

waypoint n ‘a few’ turns

B

A B A B

Check for breakdown in waypoint following Command trajectory back to base

slide-26
SLIDE 26
  • 26-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Phase II Enhancements Automatic generation of Simulink (autocodable) monitors Develop sets of monitors for a variety of critical parameters Automatically generate parameters/bounds required by monitors Monitoring for triplex architectures Guaranteed bounds for on-board neural networks

slide-27
SLIDE 27
  • 27-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

Flight Testing Production Vehicles - DACS

RASCLE Simulation-based Analysis RASCLE Simulation-based Analysis Analysis Method Analysis Method

Analysis Methods, Software Tools, and Novel System Designs

V&V Through the Control Law Life Cycle

Production Control Law System

Measured Responses Effector Cmds. Pilot Cmds.

In-Line Retrofit Control Module

Measured Responses Actuators

r r u  x ^ Production Control Law System

Measured Responses Effector Cmds. Pilot Cmds.

In-Line Retrofit Control Module

Measured Responses Actuators

r r u  x ^

Automated Off-Line Test Of Stability, Robustness, and Performance (with MuSyn) Run-Time V&V (with

Lockheed)

Retrofit Flight Controls Real-Time Monitoring of Safety Margins

(with MuSyn)

COTR: Celeste Belcastro, NASA Langley COTR: Christine Belcastro, NASA Langley

slide-28
SLIDE 28
  • 28-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

FUNDING ($K)—Show all funding contributing to this project

FY05 FY06 FY07 FY08 FY09 AFOSR Funds 100 375 375

TRANSITIONS

Deb,Tao,Burkholder,Smith, “An adaptive inverse control scheme for a synthetic jet actuator,” Proc. 2005 ACC Deb,Tao,Burkholder,Smith, “Adaptive compensation of nonlinearities in syn. jet actuators,” AIAA GNC 2006 Deb,Tao,Burkholder,Smith, “Adaptive compensation control

  • f synthetic jet actuator arrays for airfoil virtual shaping,” to

appear AIAA Journal of Aircraft

STUDENTS, POST-DOCS

Dipankar Deb (U. of Virginia); Patrick Shea (U. of Wyoming)

LABORATORY POINT OF CONTACT

  • Dr. James Myatt, AFRL/VACA, WPAFB, OH

APPROACH/TECHNICAL CHALLENGES The control inputs for synthetic jet actuators to achieve a desired aerodynamic effect have been shown to be complex The control algorithms are based on the adaptive inverse technique developed specifically to compensate for sensor and actuator nonlinearities that are imperfectly known ACCOMPLISHMENTS/RESULTS  Designed, fabricated, and tested synthetic jet actuators and wind tunnel model  Designed and implemented a real-time control and data acquisition system  Wind tunnel tests over a wide range of conditions Long-Term PAYOFF: Closed-loop flight control without mechanical control surfaces, plus expanded flight envelopes OBJECTIVES Design, implement, and test adaptive control algorithms to achieve closed-loop flight control

  • bjectives using synthetic jet actuators for virtual

surface shaping Demonstrate closed-loop control performance using an innovative, tailless wind tunnel model with integrated synthetic jet actuators

With U. of Virginia (Gang Tao, Ph.D.) And U. of Wyoming (Douglas Smith, Ph.D.)

Phase II Wind Tunnel Model Design

Adaptive Control of Synthetic Jet Arrays with Unknown Nonlinearities

slide-29
SLIDE 29
  • 29-

VA SAB 01ACGSC Meeting 99, Boulder

INNOVATIVE WIND TUNNEL MODEL WITH INTEGRATED SYNTHETIC JET ACTUATORS AND REAL-TIME CONTROL SYSTEM

  • Real-time adaptive control and data

acquisition system implemented using xPC Target RTOS from Mathworks

  • Algorithms implemented and tested in

Matlab/ Simulink can be rapidly deployed in hardware

  • Parametric models suitable for control

algorithm design were developed for synthetic jet actuators operating at high angles of attack to delay flow separation

  • J. Burkholder / Barron Associates, Inc., D. Smith / University of Wyoming, G. Tao / University of Virginia

Synthetic Jet Actuator

  • Tailless aircraft model with distributed

synthetic jet actuators and pressure ports

  • Wind tunnel testing is being conducted
  • ver a wide range of operating conditions
slide-30
SLIDE 30
  • 30-

ACGSC Meeting 99, Boulder Wed, Feb 28, 2007

NASA SBIR/STTR Technologies

Active Flow Control with Adaptive Design Techniques for Improved Aircraft Safety

PI: Jason Burkholder / Barron Associates, Inc. – Charlottesville, VA

Significa cance ce of Opportunity

  • Potential for low-cost safety improvements for commercial transport aircraft
  • Innovative synthetic jet actuators strategically-located on airfoil could

delay stall and provide “back-up” control power

  • Adaptive control is required due to complex, nonlinear actuator

dynamics

Phase se I Resu sults

  • Designed and implemented adaptive control laws – verified performance

analytically and in simulation

  • Designed wind tunnel model, novel actuators, and comprehensive Phase II

test plan

Ph Phase II II Work rk Tasks

  • Develop fully functional AIFAC tool (Adaptive Inverse For Actuator Compensation)
  • Fabricate wind tunnel models and synthetic jet actuators – optimize actuator

layout

  • Implement real-time adaptive control system and demonstrate in closed-loop

wind tunnel tests

  • Quantify safety improvements and develop V&V Plan to facilitate future flight

tests

Proposal T2.02-9831

Ap Applicati tions

  • AirSTAR Testbed for AvSP/SAAP
  • Complex damage-adaptive FDI & control
  • Operation near edge of flight envelope
  • NASA Intelligent Flight Control System (IFCS)
  • Commercial and military aircraft – especially tailless “stealth” aircraft

Con Conta tacts ts

burkholder@barron-associates.com (434) 973-1215

Phase II Actuator Designs Phase II Actuator Designs Phase II Wind Tunnel Model Design Phase II Wind Tunnel Model Design