Flight Control System Design and Test for Unmanned Rotorcraft

Transcription

1 IEEE Santa Clara Valley Control Systems Society Technical Meeting September 2, 2 Flight Control System Design and Test for Unmanned Rotorcraft Chad R. Frost Flight Control and Cockpit Integration Branch Army/NASA Rotorcraft Division NASA Ames Research Center Moffett Field, CA

2 Overview Background Design Tools Design Methods UAV programs Example

3 Background A UAV is an uninhabited, reusable aircraft that is controlled: Remotely, Autonomously by pre-programmed on-board equipment, Or a combination of both methods Currently >24 UAV systems developed by 3 countries are operational or in test Numerous missions, current and proposed: Military Civilian Space Source: Bob Keith, NATO UAV C2 Workshop 999 (Unclassified).

4 Background

5 Background Many vehicle configurations, but rotary-winged vehicles form a significant and growing portion Hover-capable UAVs offer unique capabilities, but come with unique challenges

6 Background Significant industrial and military expertise exists in fixedwing UAV development. Initial work on rotary-wing UAVs did not exploit the capabilities of the configuration: Lack of familiarity with rotorcraft issues Inability to foresee problem areas NASA involvement in rotorcraft UAV development sought to take performance to a new level.

7 Background Ames is NASA rotorcraft center: Army / NASA Rotorcraft Division NASA: Aerospace Directorate Army: Aviation & Missile RD&E Center Flight Control and Cockpit Integration Branch Expertise in rotorcraft: Flight control Modeling Simulation Design tools developed in-house

8 Design Tools CIFER Comprehensive Identification from Frequency Responses CONDUIT CONtrol Designer s Unified InTerface RIPTIDE Real-time Interactive Prototype Technology Integration/Development Environment

9 Design Tools CIFER Extraction of mathematical description of vehicle dynamics from test data Inverse of simulation Robust software, widely used in aerospace industry

10 Design Tools CONDUIT Evaluation and analysis of any modeled system Linear model from CIFER Non-linear simulation code Control system design Simulink or SystemBuild blockdiagram modeling Control system optimization User-selected specifications Multi-variable, multi-objective FSQP

11

12 Design Tools RIPTIDE Real-time simulation environment Can use models from CIFER, CONDUIT, or stand-alone code Can use control system designs from CONDUIT Hardware-in-the-loop capability

13 Design Tools Elements of RIPTIDE real-time simulation environment Cockpit Displays Cockpit Display Development Math Model Development RIPTIDE SHARED MEMORY Out-the-window Displays A/C Model Controller Pilot Inceptors Matlab SIMULINK Auto Code CONDUIT

14 Design Methods Specifications Flight Test Flight Vehicle Development Cycle Design Development Simulation

15 Design Methods Typical sequence of control system development: Collect data from vehicle Extract linear math model using CIFER Design control system using CONDUIT Optimize control system gains using CONDUIT Shakedown tests in RIPTIDE Fly control system on vehicle If modeling done correctly, vehicle response should match model predictions.

16 UAV Programs Ames participation in: VTUAV LADF R-5/R-MAX BURRO

17 Example: BURRO USMC demonstration program Broad-area Unmanned Responsive Re-supply Operations UAV to pick up loads from moving ship, deliver autonomously to inland troops

18 Introduction Kaman Aerospace K-MAX In production Designed for load-lifting 6, lb vehicle 6, lb slung load capacity Synchropter configuration Servo-flap rotor

19 Introduction Army/NASA CRDA with Kaman to support FCS development Three integrated tools CIFER : System identification CONDUIT: Control system modeling, analysis, optimization RIPTIDE: Desktop real-time simulation

20 Design Manual Little or no optimization Stop when "good enough" Introduction Analysis Manual Subjective Guesswork Traditional Simulation Not real time Not "pilot in the loop" Flight Test Slow Expensive Risky Design CONDUIT Multi-objective optimization Analysis CIFER Automated Accurate Objective State-of-the-Art Simulation CONDUIT (NRT) RIPTIDE (RT, piloted) Flight Test Fewer hours required Higher initial confidence

21 Introduction Scope: Hover / low-speed Unloaded Ground operator control

22 Aircraft Modeling Start with piloted frequency sweeps of unaugmented K-MAX 8-DOF (rigid-body + 2 rotor states) linear state-space model identified from flight data using CIFER Imag Axis Pitch Lateral-directional Eigenvalues, hover Unstable roll Heave Unstable pitch Real Axis

23 Aircraft Modeling Verified in time domain using CIFER Lateral Control Deflection Roll Rate Response Math model Flight data Ref: Jason Colbourne, et al., American Helicopter Society Forum, May Pitch Rate Response Time (Sec)

24 Aircraft Modeling Sensor dynamics Equivalent delays estimated from manufacturer specs (25ms) 2nd-order Padé approximations Actuator dynamics Identified from bench-test frequency sweeps 2nd-order systems (ω=2 r/s, ζ=.5) Rate- and position-limiting

25 Aircraft Modeling Aircraft, actuator and sensor models implemented in Simulink blockdiagram Control input Flight control laws Actuator models Linearized 8DOF Aircraft dynamics Aircraft response Sensor models

26 Control Law Development Inner Loops Attitude Command / Attitude Hold PID controller Heading Command PD controller Altitude Rate Command PD controller Outer Loops Translational Rate Command PI controller (or position feedback) Modeled in Simulink Velocity Translational Rate Control input PI Translational Rate controller PID Attitude controller Actuator model Actuator Attitude input Attitude rate Linearized 8DOF Aircraft dynamics 2 Attitude 3 Attitude Rate Sensor models

27 Control Law Development Lateral controller shown 6 Command_Mode [ = Velocity 2 = Attitude ] FCC_State 4 Trim_Right 3 Trim_Left 2 Stick Input [stu] 5 sensors.5 trim sensitivity (right).5 trim sensitivity (left) Roll angle m Roll rate Vd FCC_State Trim + Trim - Direct Total Control Control Mixer stick sensitivity deg per stu /57.3 deg > rad discrete latch Holds input value when FCC_State = Active () (Attitude hold) trigger out in Complete Simulink model: Attitude Command Mode Velocity Command Mode 22 inputs 32 outputs 33 states (continuous and discrete) 27 tunable gains ( design parameters ) Attitude limit +/.523 rad (3 deg) dpp_atp Feedback time constant rad/(rad/sec) dpp_akp Proportional gain stu/rad dpp_aki Integrator gain Integrator limited windup (5%) s Lateral_Servo_Command [stu] stick sensitivity ft/sec per stu discrete latch Holds input value when FCC_State = Active () (Velocity hold) trigger out in dpp_akp_v Proportional gain rad/(ft/sec) dpp_aki_v Integrator gain Integrator limited windup.523 rad (3 deg) s Includes nonlinear elements: Limited integrators Authority limits Mode switching

28 Control Law Development CONDUIT Optimization Engine: Multi-objective optimization using FSQP Gain/Phase Margins Adjusts design parameters (system gains) to meet requirements of specifications Specifications represented graphically, 3 regions based on level of performance Categorizes specifications: Hard must be met Soft should be met, without violating Hard specs Objectives minimized after all specs satisfied PM [deg] GM [db] Level Level 2 Level 3

29 Control Law Development Specification selection Stability Performance and Handling Qualities Objectives Rationale Airframe originally designed as a manned vehicle Safety pilot on board demonstrator vehicle Ground operator control will be VFR / simple tasks

30 Control Law Development Stability Specifications (Hard constraints) Gain/Phase Margins (rigid-body freq. range) Eigenvalue Location 8 PM [deg] GM [db] Real Axis

31 Control Law Development Performance Specifications (Soft constraints) Pitch Attitude Change in Second [deg] d_theta [deg] Normalized Attitude Hold (Disturbance Rejection) Time (sec) Actuator Position Saturation Actuator Saturation.5 Actuator Rate Saturation

32 Control Law Development Handling Qualities Specifications (Soft constraints) Eq. Rise Time (T x-dot, T y-dot ) [sec] Translational Rate Rise Time Phase delay [sec] Bandwidth & Time Delay (roll) Other MTEs;UCE=;Fully Att Bandwidth [rad/sec] Damping Ratio (Zeta) Damping Ratio Attitude Hold time delay, tau_hdot, sec Heave Response Hover/LowSpeed.5 heave mode, invthdot, [rad/sec]

33 Control Law Development Objective Specifications Spec selection reduces actuator sizing, component fatigue, and noise sensitivity Actuator RMS Crossover Freq. (linear scale).5 Actuator RMS Crossover Frequency [rad/sec]

34 Control Law Development Control system gains tuned using CONDUIT Initial tuning: 27 parameters 33 specifications All specifications satisfied (Level ) RMS actuator position and crossover frequency minimized EigLcG: Eigenvalues (All) Real Axis PM [deg] Phase delay [sec] Damping Ratio (Zeta) StbMgG: Gain/Phase Margins (rigid-body freq. range) GM [db] PED Bandwidth [rad/sec] BnwPiH2:BW & T.D. (pitch) Other MTEs;UCE=;Fully Att OvsPiH:Damping Ratio Attitude Hold ROLL YAW PITCH.5 PM [deg] time delay, tau_hdot, sec Phase delay [sec] StbMgG: Gain/Phase Margins (rigid-body freq. range) H 8 COL GM [db] FrqHeH2:Heave response shipboard landing.5 heave mode, invthdot, [rad/sec] Bandwidth [rad/sec] BnwRoH2:BW & T.D. (roll) Other MTEs;UCE=;Fully Att RisPiV:Pitch Attitude Change in sec d_theta [deg] PM [deg] Actuator Position Saturation Phase delay [sec] StbMgG: Gain/Phase Margins (rigid-body freq. range) GM [db] LON (ACAH) SatAcG: Actuator Saturation COL, LON.5 Actuator Rate Saturation BnwYaH2:BW & T.D. Other MTEs (Yaw) Bandwidth [rad/sec] RisRoV:Roll Attitude Change in Second [deg] d_phi [deg] PM [deg] Actuator Position Saturation StbMgG: Gain/Phase Margins (rigid-body freq. range) GM [db] LAT (ACAH) SatAcG: Actuator Saturation LAT, PED.5 Actuator Rate Saturation -.5 HldNmH:Normalized Attitude Hold Time (sec) PITCH ROLL YAW RisYaV:Yaw Attitude Change in Second [deg] d_psi [deg] 2 CrsLnG:Crossover Freq. (linear scale) RmsAcG:Actuator RMS COL LON LAT PED Crossover Frequency [rad/sec] COL LON LAT PED.5 Actuator RMS

35 Control Law Development 2 Lateral Stability Margins (Initial): PM = 46.8 deg. (ω c = 3.75 rad/sec) GM = 9.7 db, (ω 8 =.23 rad/sec) Conditionally stable lat & lon Model predicts stable, welldamped responses Gain (db) -2-4 Roll response Phase (deg) Attitude (deg) -4 Frequency (rad/sec) 6 Time (sec)

36 Flight Test First flight test with CONDUIT-tuned gains Aircraft responses did not agree with model (lon and lat) 4 3 Roll attitude command Roll Attitude (degrees) Roll attitude Time (sec)

37 Flight Test Looking for source of discrepancy: Lon and lat doublets flown closed-loop CIFER used to extract frequency responses Actual sensor and actuator dynamics identified Equivalent time delay greater than originally estimated

38 Flight Test Component Estimated Delay (ms) Actual Delay (ms) Actuators 5 7 Sensors Computer 2 6 Filters 7 TOTAL

39 Flight Test Updated Simulink model with identified delays Added delay results in highly constrained system Phase (degrees) XV-5 Tilt Rotor K-MAX BURRO Narrowed range of stability Frequency (rad/sec)

40 Flight Test Added lead filter to lon & lat attitude feedback 2 CONDUIT-tuned result acceptable range of crossover frequency FCS gains re-tuned with CONDUIT Gain (db) -2-4 CONDUIT successfully traded off phase margin for gain margin Phase (degrees) Frequency (rad/sec)

41 Flight Test 8 Gain/Phase Margins (rigid-body freq. range) 8 Gain/Phase Margins (rigid-body freq. range) PM [deg] 6 4 PM [deg] 6 4 After CONDUIT tuning Baseline gains CONDUIT tuning results Actuator Position Saturation GM [db] LON (ACAH) Actuator Saturation LON.5 Actuator Rate Saturation Actuator Position Saturation GM [db] LAT (ACAH) Actuator Saturation LAT.5 Actuator Rate Saturation Bandwidth & Time Delay (pitch).4 Bandwidth & Time Delay (roll).4 Phase delay [sec].3.2. PITCH Phase delay [sec].3.2. ROLL Bandwidth [rad/sec] Bandwidth [rad/sec]

42 Flight Test Eigenvalue Location PM [deg] Gain/Phase Margins (rigid-body freq. range) COL PM [deg] Gain/Phase Margins (rigid-body freq. range) LON (TRC) LON (ACAH) PM [deg] Gain/Phase Margins (rigid-body freq. range) LAT (TRC) LAT (ACAH) CONDUIT results: Level 2 (8 specs) PM [deg].5.5 Real Axis Gain/Phase Margins (rigid-body freq. range) PED GM [db] time delay, tau_hdot, sec GM [db] Heave Response Hover/LowSpeed.5 heave mode, invthdot, [rad/sec] Actuator Position Saturation GM [db] Actuator Saturation LON.5 Actuator Rate Saturation Actuator Position Saturation GM [db] Actuator Saturation LAT.5 Actuator Rate Saturation Reduced bandwidth Phase delay [sec] Bandwidth & Time Delay (pitch) Other MTEs;UCE=;Fully Att Bandwidth [rad/sec] Phase delay [sec].4.2. Bandwidth & Time Delay (roll) Other MTEs;UCE=;Fully Att.3.3 PITCH ROLL YAW Bandwidth [rad/sec] Phase delay [sec] Bandwidth & Time Delay (yaw) Other MTEs Bandwidth [rad/sec] Normalized Attitude Hold (Disturbance Rejection) PITCH, ROLL YAW Time (sec) Damping Ratio (Zeta) Damping Ratio Attitude Hold ROLL PITCH Pitch Attitude Change in Second [deg] Roll Attitude Change in Second [deg] Yaw Attitude Change in Second [deg] d_theta [deg] d_phi [deg] d_psi [deg] Eq. Rise Time (T x-dot, T y-dot ) [sec] Translational Rate Rise Time LON LAT Actuator RMS COL LON LAT PED.5 Actuator RMS Crossover Freq. (linear scale) COL LON LAT PED Crossover Frequency [rad/sec]

43 Flight Test Roll response much improved; model responses agree well with flight results Roll Attitude (degress) Roll attitude command Roll attitude Time (sec)

44 Flight Test BURRO successfully demonstrated to USMC nine months after start of development

45 Conclusions Design space is very limited Aircraft dynamics Control system hardware CONDUIT was able to extract the best achievable performance within design limitations High frequency dynamics were key driver of closed- loop performance CIFER was useful in identifying system elements Advanced design tools allowed rapid development of a successful UAV 9 month time span Recovery from added delay

46 Current and Future Work Build of K-MAX BURRO UAV successfully demonstrated to USMC Build 2 now in development 2-lb loaded hover -DOF EOM and CIFER ident complete FCS design complete 5-lb case in development Envelope expansion to 7 KTAS in progress

47 Questions? For additional information: