ISM 101Gabriel Hugh Elkaim 1 ISM 101 Guest Lecture on Robotics and Control 24.Feb.2005 Gabriel Hugh Elkaim.

1ISM 101 Gabriel Hugh Elkaim

ISM 101Guest Lecture on Robotics and Control

24.Feb.2005

Gabriel Hugh Elkaim


Assistant Professor

Computer Engineering

353B Baskin Engineering

[email protected]

(831) 459-3054

Gabriel Hugh ElkaimGabriel Hugh Elkaim•Background: Aerospace Engineering

•Interest: Robotics/Embedded Systems


ASL LABASL LAB

• Santa Cruz Santa Cruz Autonomous Systems Autonomous Systems LabLab

• Robotics and Robotics and Embedded SystemsEmbedded Systems

• Sensor FusionSensor Fusion• Robust Software Robust Software

DesignDesign


Relevant ExpertiseRelevant Expertise

• Feedback Control SystemsFeedback Control Systems• Embedded System Embedded System

Software/HardwareSoftware/Hardware• Mechatronic DesignMechatronic Design• Microcontroller/DSP Microcontroller/DSP

projectsprojects• Navigation/Guidance Navigation/Guidance

SystemsSystems• Global Positioning SystemGlobal Positioning System


Relevance to ISMRelevance to ISM

Autonomous Mobile Platforms depend on: Autonomous Mobile Platforms depend on:

• Sensing – environment, position, pose or Sensing – environment, position, pose or attitude, obstacles, etc.attitude, obstacles, etc.

• Path Planning (traditional A/I) – given the Path Planning (traditional A/I) – given the environment, get to objectiveenvironment, get to objective

• Control – How do you track the trajectory Control – How do you track the trajectory that you have generatedthat you have generated


OutlineOutline

• Robotics in generalRobotics in general

• Sensors in generalSensors in general– Types of SensorsTypes of Sensors– Filtering IssuesFiltering Issues

• Control in generalControl in general– PID (Proportional Integral Derivative Control)PID (Proportional Integral Derivative Control)– Example, 3 wheeled ground vehicleExample, 3 wheeled ground vehicle


Robotics

• Czech word Robota means compulsory labor.

• “Rosum’s Universal Robots” written in 1920 by Czechoslovakian author Karel Capeck

• Robotics: technology dealing with the design, construction, and operation of robots.


RobotsRobots

According to Merriam-Webster:According to Merriam-Webster:1 a1 a :: a machine that looks like a human being and a machine that looks like a human being and performs various complex acts (as walking or performs various complex acts (as walking or talking) of a human being; talking) of a human being; alsoalso :: a similar but a similar but fictional machine whose lack of capacity for fictional machine whose lack of capacity for human emotions is often emphasized human emotions is often emphasized bb :: an an efficient insensitive person who functions efficient insensitive person who functions automaticallyautomatically22 :: a device that automatically performs a device that automatically performs complicated often repetitive taskscomplicated often repetitive tasks33 :: a mechanism guided by automatic controls a mechanism guided by automatic controls


My DefinitionMy Definition

• Look at a Venn Look at a Venn diagram of diagram of Mechanical and/or Mechanical and/or Electrical Hardware, Electrical Hardware, Software, and Control Software, and Control Systems.Systems.

• Robotics is the Robotics is the overlapping area at the overlapping area at the center of the threecenter of the three

GNCGNC

MechanicalMechanical

SoftwareSoftware

RoboticsRobotics


What are some of the Issues?What are some of the Issues?

• ConfigurationConfiguration– What mechanical scheme do you need to What mechanical scheme do you need to

complete the missioncomplete the mission– Example: UAV that deploys from a type “A” Example: UAV that deploys from a type “A”

sonobuoy (36” long x 4.875” in diameter)sonobuoy (36” long x 4.875” in diameter)– Example: Pipe Inspection must negotiate 90 Example: Pipe Inspection must negotiate 90

degree bends, self contained, etc.degree bends, self contained, etc.


NavigationNavigation

• How do you know where you are?How do you know where you are?– OutdoorsOutdoors– UnderwaterUnderwater– In SpaceIn Space– IndoorsIndoors– UndergroundUnderground


GuidanceGuidance

• Where do you want to go?Where do you want to go?

• How fast do you need to get there?How fast do you need to get there?

• Is there anything in the way?Is there anything in the way?


ControlControl

• How do you get from where you are, to How do you get from where you are, to where you want to go?where you want to go?

• What if something is not as predictedWhat if something is not as predicted


OdometryOdometry


GPS – Global Positioning SystemGPS – Global Positioning System


InertialsInertials


AttitudeAttitude


Control Issues

• Get the device to do what it is commandedGet the device to do what it is commanded

• Open Loop ControlOpen Loop Control

• Feedback ControlFeedback Control– Must have a sensorMust have a sensor– Increases Disturbance RejectionIncreases Disturbance Rejection– Decreases Sensitivity to parameter variationDecreases Sensitivity to parameter variation

• Entire specialty of engineeringEntire specialty of engineering


Examples of Control SystemsExamples of Control Systems

• Toilet BowlToilet Bowl

• Cruise ControlCruise Control

• Thermostat on HouseThermostat on House

• Missile Guidance SystemMissile Guidance System

• Mobile Robot Obstacle AvoidanceMobile Robot Obstacle Avoidance

• Many, many moreMany, many more


Cruise Control in DetailCruise Control in Detail


Cruise Control – Open LoopCruise Control – Open Loop


Cruise Control – Closed LoopCruise Control – Closed Loop


Generic Control System Block DiagramGeneric Control System Block Diagram


Sensor IssuesSensor Issues

• Dynamic RangeDynamic Range

• LinearityLinearity

• HysteresisHysteresis

• QuantizationQuantization

• Temperature EffectsTemperature Effects

• BandwidthBandwidth


Sensors – LinearitySensors – Linearity


Sensors – Dynamic RangeSensors – Dynamic Range


Sensors – HysteresisSensors – Hysteresis


Sensors – QuantizationSensors – Quantization


Sensors – Temperature EffectsSensors – Temperature Effects


Sensors – BandwidthSensors – Bandwidth


Actuator IssuesActuator Issues

• Power / StrengthPower / Strength

• LinearityLinearity

• HysteresisHysteresis

• QuantizationQuantization

• Temperature EffectsTemperature Effects

• BandwidthBandwidth


Control System – PIDControl System – PID • ProportionalProportional• IntegralIntegral• DerivativeDerivative


Control System – Motor DriveControl System – Motor Drive


Control System – Motor DriveControl System – Motor Drive


Control System – Voice CoilControl System – Voice Coil


Control System – Voice CoilControl System – Voice Coil


Control System – HeaterControl System – Heater


Control System – HeaterControl System – Heater


Control System – Motor Drive w/PControl System – Motor Drive w/P


Control System – Voice Coil w/PControl System – Voice Coil w/P


Control System – Heater w/PControl System – Heater w/P


Control System – Motor Drive w/IControl System – Motor Drive w/I


Control System – Heater w/IControl System – Heater w/I


Control System – Motor Drive w/PIControl System – Motor Drive w/PI


Control System – Heater w/PIControl System – Heater w/PI


Integrator Windup – Motor Drive w/PIIntegrator Windup – Motor Drive w/PI


Integrator Limit – Motor Drive w/PIIntegrator Limit – Motor Drive w/PI


Control System – Voice Coil w/PDControl System – Voice Coil w/PD


Control System – Heater w/PIDControl System – Heater w/PID


PID ControllersPID Controllers

• Proportional gain increases response speed, Proportional gain increases response speed, to much gain causes system to ring.to much gain causes system to ring.

• Integral gain kills steady-state error, wind-Integral gain kills steady-state error, wind-up and/or too much gain can cause system up and/or too much gain can cause system to go unstable.to go unstable.

• Derivative gain adds damping and stability, Derivative gain adds damping and stability, but is sensitive to jitter and noise.but is sensitive to jitter and noise.


Tuning PID ControllersTuning PID Controllers

• Don’t need to understand Controls or Don’t need to understand Controls or System to use PID.System to use PID.

• Start with pure Derivative control.Start with pure Derivative control.

• Increase gain until system oscillates or you Increase gain until system oscillates or you see over 50% overshoot.see over 50% overshoot.

• Go up to verge of ringing, back off by a Go up to verge of ringing, back off by a factor of 2 or 4.factor of 2 or 4.



• Start with Proportional gain, increase by Start with Proportional gain, increase by factor of 8 to 10 until oscillation.factor of 8 to 10 until oscillation.

• If it is already oscillating, decrease by factor If it is already oscillating, decrease by factor of 8 to 10.of 8 to 10.

• From verge of oscillation, back off by From verge of oscillation, back off by factor of 2 to 4.factor of 2 to 4.



• Start with Integral gain very small, 0.0001 Start with Integral gain very small, 0.0001 to 0.01.to 0.01.

• Increase until you get response you like.Increase until you get response you like.

• Be sure to implement anti-windup.Be sure to implement anti-windup.

• If you have problems, play with sample If you have problems, play with sample rate.rate.


A 3-Wheeled VehicleA 3-Wheeled Vehicle









ISM 101Gabriel Hugh Elkaim 1 ISM 101 Guest Lecture on Robotics and Control 24.Feb.2005 Gabriel Hugh Elkaim.

Documents

gabriel hugh elkaim

gabriel hugh elkaim

issues control

roboticsembedded systems

objective control

outline robotics

wheeled ground vehicle

sensing environment