In4073 Embedded Real-Time Systems

Electrical Model Quad Rotor UAV

In4073 Emb RT Sys (2017-2018)

System Setup

(source: assignment.pdf)

user I/O(pilot)

joystick

drone

PC link

PC

flight control(ES)

2

HW view

In4073 Emb RT Sys (2017-2018)

ARMCortex M0

115k2bps

BLE

PC

BLE1 Mbps

Quadrupel drone

gyro/accel (6)motors (4)barometer

ms intervals

Drone: Frames & Main Variables

X, u

L, p

Z, w

earth framex

y

z

N, r

M, q

Y, v

body frame

In4073 Emb RT Sys (2017-2018)

Drone attitude: Euler Angles

z (3)

x (2,3)

y (0)

x (0)

x(1)

z (2)

z (0,1)

y (3)z (0,1)

ϕ(1) (2)θ

(3)

ϕ(3)

(1)

y (1,2)

θ (2)body frame

earth frame

ψ

ψ

In4073 Emb RT Sys (2017-2018)

rotation order ψ, θ, Φ

Drone: Forces

X, u

L, p

Z, w

earth framex

y

z

N, r

M, q

Y, v

body frame

In4073 Emb RT Sys (2017-2018)

Forces: X, Y, ZMoments: L,M,N

Drone: Actuators

rotor 1 – rotor 4through RPM, denoted by Ω

driven by ES signals ae1 – ae4

ae = 0 -> Ω = 0ae = 1000 -> Ω = max

xz

y

engine 4

engine 3

engine 2

engine 1

In4073 Emb RT Sys (2017-2018)

Drone: Dynamics (in hover)

dq/dt = M / IYTi = rotor thrust = f(Ωi)mg = gravityh = rotor distance ref. center of gravityIY = heli rotation inertia in Y-axis

h

mg

T1

M = T h

T2

T4T3 T

du/dt = X / maccelerated rotation & xlation!

mg

Tmg

mg

mg

X

X = Θ mgΘΣ Ti = mg

In4073 Emb RT Sys (2017-2018)

Drone: Rotor Actuators

In general

Z = -b(Ω12+Ω2

2+Ω32+Ω4

2)L = b(Ω4

2-Ω22)

M = b(Ω12- Ω3

2)N = d(Ω2

2+Ω42-Ω1

2-Ω32)

So compute Ωi (i.e., aei) fromdesired lift (Z), roll rate (L),pitch rate (M), and yaw rate (N)

xz

y

engine 4

engine 3

engine 2

engine 1

In4073 Emb RT Sys (2017-2018)

Drone: Sensors (angles)

In4073 Emb RT Sys (2017-2018)

ax = sinΘ g ~ Θ g

p, q, r.Gyro:

ax, ay, az.Accelerometer: .

Drone: Sensors (angles)

In4073 Emb RT Sys (2017-2018)

Φ, θ, ψ.DMP:

SW view

joystick

drone

PC link

PC

liftrollpitchyaw

sensorsESjoystick

keyboard actuators

ae1ae2ae3ae4

phithetapqr

saxsaysazspsqsr

drone

In4073 Emb RT Sys (2017-2018)

Communication protocol (lab 1)

PC -> Drone (send) periodic: pilot control ad hoc: mode changing, param tuning

Drone -> PC (receive) periodic: telemetry (for visualization) ad hoc: logging (for post-mortem analysis)

Dependable, robust to data loss header synch

In4073 Emb RT Sys (2017-2018)

Drone: Control Circuit

control loop example (yaw rate):eps = yaw - sr; // measure deviationN_needed = P * eps; // compute ctl actionae1 .. ae4 = f(N_needed); // actuate, see slide 9

liftrollpitchyaw

sensorsESjoystick

keyboard actuators

ae1ae2ae3ae4

phithetapqr

saxsaysazspsqsr

drone

In4073 Emb RT Sys (2017-2018)

Calibration

• real p, q, r, .. are sensed in terms of sp, sq, sr, ..• sp, sq, .. have a (voltage) bias (are not zero at rest)• so need to calibrate all 6 sensors at rest:

• let sr0 be sensor output at rest• real estimate of r are given by (z for zeroed)

zr = sr – sr0

sr

sr0-

+

cal

zr

mode

In4073 Emb RT Sys (2017-2018)

Filtering

• signals also need to be filtered to remove noise• filtered signal input to embedded controller

filterzr

controlleryaw ae1-ae4

In4073 Emb RT Sys (2017-2018)

Controller Modes

• controller mode: manual• controller model: calibrate• controller mode: control (yaw, pitch, roll)

filter

controlleryaw ae1-4

sr

sr0-

+

cal

zr

mode

In4073 Emb RT Sys (2017-2018)

Before you go

Safety first:• goggles• common sense

In4073 Emb RT Sys (2017-2018)