Study Notes on Robot Reference Frames and Pose Representation

Introduction to Robot Localization and Reference Frames

  • End Effector and Goal
    • The gripper operates to pick up and move objects accurately in space.
    • The problem involves determining the position and trajectory of the gripper.

Understanding Reference Frames

  • Concept of Frames
    • A reference frame describes the position of a body in relation to another frame.
    • Necessary to track the current configuration, especially when reaching objects.
  • Examples of Reference Frames:
    • Two reference frames: a robotic frame and a world frame (e.g., the room).

Example Scenario: Wheeled Robot

  • Robot Description
    • Imagine a wheeled robot depicted from a bird's eye view.
    • Moves from one location to another within its operating environment.
Description of Robot’s Position
  • Global Reference Frame
    • It’s essential to describe the robot’s location concerning a stable world reference (e.g., GPS coordinates for mapping).
    • Use of a fixed world frame allows for consistent positional context.

Coordinate Systems

  • Axes Definitions
    • The description utilizes x, y, and z axes as unit vectors indicated by hats (C6x, C6y, C6z).
    • Each unit vector has a length of 1, e.g.,
      ||C6x|| = ||C6y|| = ||C6z|| = 1
    • The world frame denoted by subscript w (e.g., x w , y w , z w ).
    • The frame is static and fixed relative to the environment.

Rigid Body Dynamics

  • Rigid Body Assumption
    • The robot is assumed to be rigid, meaning distances between any two points do not change regardless of the robot's movement.
    • One reference frame is sufficient for describing all points on the robot due to constant distances between points.
  • Reference Frame Positioning
    • The frame can be attached to various points, commonly the center of mass.
    • Conventionally, x points forward, y to the left, and z upward.

Pose Representation

  • Pose Definition
    • Pose consists of two main components: position and orientation.
    • Position: location of the robot’s body frame origin with respect to the world frame.
    • Orientation: the alignment of the body frame relative to the world frame.
Mathematical Notations
  • Denote the pose of the robot as a vector
    ext{Pose} = egin{pmatrix} ext{Position} \ ext{Orientation} ext{ } ext{(Attitude)}\ ext{ } ext{ } ext{ } ext{} ext{(includes three rotation parameters)} ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{} ext{1+ |K| + |K| |K| |K|} ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ }\ ext{ } ext{ } ext{ } ext{} ext{(where |K| is all K} ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ } ext{ }\ ext{ } ext{ } ext{ } ext{z = }
Importance of Orientation
  • Orientation Relevance
    • Knowing the orientation (or attitude) is crucial, especially for robots that are not planar (e.g., aerial robots).
    • Incorrect orientation awareness can lead to unintended movements.

Rotation Representations

  • Methods of Orientation Representation:
    • Various mathematical techniques exist, including:
    • Euler angles: represent orientation as a sequence of rotations about the axes.
    • Quaternion representation: useful for avoiding gimbal lock in three-dimensional space.
    • Exponential coordinates: alternative way of describing rotations.
  • Need for Consistency
    • Always ensure understanding of which reference frame is being used in calculations.

Validating Reference Frames

Right-Hand Rule in Coordinate Systems
  • Right-Handed System Validation:
    • The cross product of vectors can determine if a coordinate system is right-handed.
    • For vectors C6x, C6y, if C6z = C6x imes C6y holds true, the system is right-handed.
    • Example:
    • If C6x = (1, 0, 0), C6y = (0, 1, 0), C6z = (0, 0, 1), then
      C6z = C6x imes C6y will equal (0, 0, 1), confirming a right-handed system.

Applying the Knowledge of Reference Frames

  • Frame Relationships:
    • Points need location descriptions using appropriate reference frames.
    • Vectors drawn from frame origins provide positional context for objects within robotic movements.

Practical Example: Camera and Sensors

  • Sensor Integration
    • Link different frames for multiple sensors to understand their positions and orientations in the world frame.
    • Example: A camera's position needs to be known concerning the robot and target (e.g., a fire scene).
Summary of Key Points

  • Rigid Body Reference Frames:

    • Each rigid body must have its reference frame, typically placed at the center of mass.
    • The body frame defines the robot's position and orientation concerning the world frame systematically using vectors.

  • Constant Vectors in Rigid Bodies:

    • Despite frame movements or rotations, the vector between points on a rigid body does not change, ensuring stable positional information.

  • Pose Calculation:

    • Essential for successful navigation and operation of robots by merging positional data with orientation tracking.