States

In an estimation problem, the state is the set of variables that we want to estimate. In WOLF, the state is splitted in small blocks (state_blocks) and distributed along the WOLF tree. For example, the state block that corresponds to the position of a frame is located at such frame.

State blocks are indexed by single letters or char, which in WOLF we call keys. The following keys are used:

• 'P' position (StatePoint2d/3d)

• 'O' orientation (StateAngle for 2D, StateQuaternion for 3D)

• 'V' velocity (StateVector2d/3d)

• 'I' generic sensor intrinsic parameters (its type depends on the sensor)

New classes requiring other state blocks may define new keys. You are advised not to repeat keys to avoid possible conflicts.

This section contains four main topics:

• Composites: Data structures related to states that are indexed by keys.

• StateBlock: Classes containing the partitions of the state.

• NodeStateBlocks: Common abstract base class for nodes in the WOLF tree that have state blocks.

• Accessing the robot state: How to query state in WOLF and algorithms related with it.