Basic Robot Topology
Design Requirements
• Performance requirements are critical to all successful designs
– What is the robot expected to do?
• How long? How precisely? With how much guidance?...
• How big/heavy/strong/fast?
– What are the environmental conditions?
– What is the time frame in which it will operate?
– How expensive should it be?
Mathematical Modeling
• Modeling the robot to some degree is important to understand its performance and limitations
• Models can be of varying complexity
• Best technical approach for modeling is to develop a series of models, from simple to complex, as required to satisfy questions
• Always know underlying assumptions on which each model is based
– Respect the limitations of models
– Balance use of simplified models to minimize design cycle time
• Be a critical observer of all model results
Feedback Sensor Selection
• High fidelity sensoring can be a critical element in a successful design
– Balance use of direct sensing versus estimation to meet your performance specifications
• Understand sensor requirements:
– Type and system interface
– Accuracy/Precision/Limits/redundancy management issues
• Always co-locate sensor unless absolutely unavoidable
• Understand sensor signal processing
Actuator Selection
• Response characteristics are critical:
– Speed/strength/motion
– Delay in response?
• Power supply requirements
• Physical attributes
– Size
– Weight
– Cost
• Types:
– Electrical: motors/solenoids etc.
– Mechanical: springs/dampers
– Hydraulic or Pneumatic: pistons
Processing Consideration
• Computer or microprocessor control is inherent
• How fast (sampling rate)
– Delay?
• How precise (how are numbers represented)
– Number of bits
• Fixed or floating point numbers?
Control Design Concepts
• Keep it simple
– Strive to maintain linearity
– Avoid extraneous logic and switchable modes
• Design for transient free mode switching
– Understand failure modes and effects
• Design for robustness to failures where appropriate
• Balance performance with stability concerns
– Always adhere to stability guidelines Sensors Sensors
System Integration
• Ensure input devices do not degrade the system
– Control input interface devices
– Sensors must be mounted in environmentally friendly locations
• Ensure constraints imposed by other subsystems do not compromise system
– Structural load considerations
• Use high fidelity simulation of entire system for validation (hardware in the loop to extent possible)
– Functional tests for system features
– End-to-end tests for system integration
– Stress test system as extensively as possible
– Develop regression tests if upgrades are planned
Robotics Functions
Transducer: Sensors and Actuators
• Transducer
– A device that converts a signal from one physical form to a corresponding signal having a different physical form
• Physical form: mechanical, thermal, magnetic, electric, optical,chemical...
– Transducers are ENERGY CONVERTERS or MODIFIERS
• Sensor
– A device that receives and responds to a signal or stimulus
• This is a broader concept that includes the extension of our perception capabilities to acquire information about physical quantities
• Transducers: sensors and actuators
– Sensor: an input transducer (i.e., a microphone)
– Actuator: an output transducer (i.e., a loudspeaker)
Sensors
Design Consideration
• Range
– Sensors have a limit on upper and lower bounds of the states they measure
• Precision/Accuracy/Resolution/Tolerances
– Precision is how repeatable are measurements when sensing the same state (not always the same as accuracy due to drift etc.)
– Accuracy is how close is the sensed value to the actual value of the sensed state
– Resolution is how finely can the sensor distinguish changes in state
– Tolerances are bounds identified with respect to the above rating a sensorís performance
• Type
– Analog/Digital
• Analog sensors may have a digital interface but still exhibit idiosyncrasies of analog equipment (i.e. airspeed sensor may still be subject to temperature drift)
• Direct digital sensors do not drift (i.e. digital encoders for rotation or translation)
• Noise
– Noise cannot be distinguished from real data so signal to noise ratio for sensor is important
Sensor Types
• Position
– Relative vs. absolute
• GPS provides globally referenced position
• Rangefinders provide local position information with respect to the environment
– Characteristics (where do the sensors work well and where do they not, what important features must be taken into account
• GPS works well in open spaces; may not work at all in most urban areas
• Rangefinders may work well until environment becomes cluttered or uneven
• Velocity
– Groundspeed
– Air or water speed for aerial or marine robots
• Typically a local measurement relative to the robot itself
• Acceleration
– Almost always an absolute measure unless calibrated
• Orientation (pitch/roll/compass heading)
• Rotational rate
Sensor Classification
Sampled data system constraints
• Aliasing occurs if data with a waveform higher than 1/2 the frequency is sampled
– High frequencies are “aliased” to lower frequencies
– Prefiltering is required to eliminate aliasing if input cannot be guaranteed to adhere to max frequency constraint
Sensor Placement
• Co-location
– Always best to place sensor closest to location of importance
• Flexible structures typically require understanding of mode shapes
• Environment
– Heat
– Electromagnetic interference
• Cameras should not be next to motors etc…
Gyroscope
• Several types exist (mechanical and optical).
• Mechanical example: flywheel gyroscope
– Conservation of angular momentum
– Torque on axes depends on T = I.ω.Ω
Light Sensor: Photoelectric
Read More..
