Transcript Slide 1

Robotics Intensive:
Day 3
Gui Cavalcanti
1/17/2012
Class Goals
1. Discuss open loop vs. closed loop control
2. Discuss PID
3. Review control project goals
4. Work on control systems
5. Demo control systems!
Open Loop vs. Closed Loop Control
Open Loop Control
• Issuing commands based on how you
think a system will react, without taking
into consideration how it is actually
reacting
• Common examples:
– Sprinkler Systems
– Washing Machines
– 3pi spelling your name
Open Loop Pitfalls
• Easy to develop, and thus easy to assume
it’s a good solution
• You have no idea what’s actually
happening, or even if the robot is
successful
Closed Loop Control
• Issuing commands based on both how you think
a system will react and the current state of the
system
• Common examples
– Aircraft Autopilot
– Car Cruise Control
– 3pi following a line
Cruise Control
• System Parts:
– Microcontroller
– Speed Sensor
– Engine Control Actuator
• Decision-Making:
– Microcontroller says “I want 50 MPH”
– Speed Sensor says “We’re driving 45 MPH”
– Microcontroller says “Actuator, make us go
faster”
Cruise Control
5 mph
50 mph
TOO
SLOW!
GO
FASTER!
45 mph
Repeat again and again.
Increase
Throttle
PID Control
PID Control
• Most common type of closed loop control,
commonly used in robotics
• PID:
– Proportional
– Integral
– Derivative
• What you control around matters!
Proportional Control
• Equivalent to a big spring.
• Amount system responds is proportional
to how large the error is.
Integral Control
• Equivalent to filling a bucket over time.
• Amount system responds is proportional
to how long the error has existed.
Derivative Control
• Equivalent to running through water.
• Amount system responds is proportional
to how fast the system is moving.
Controls Project Goals
The Mission
Project Description
1. Implement the basic line following
algorithm from the mbed Cookbook.
2. Implement the PID line following
algorithm from the mbed Cookbook.
3. Demonstrate identification of and use of
intersections.
4. Demonstrate travel to every branch of the
map.
Bonus Points
• Fastest time
• Ability to travel to a specific intersection
• Ability to travel to a specific intersection
by more than one route
• Demonstrate (software-based) object
detection
Work on demos!