Expand description
A no-std implementation of the Kalman and Extended Kalman Filter based on nalgebra::SMatrix statically sized matrices.
The base Kalman type is the most configurable, allowing multi-rate sensor measurements. For simpler systems with single measurements, Kalman1M is easier to use.
See this blog post to understand why the library looks the way it does.
§Getting started
See Kalman1M for an example of linear Kalman filter for a 2-state system model with no inputs.
See Kalman for an example of a linear Kalman filter for a 2-state system model with no inputs and multiple sensors.
See the examples for more realistic simulations including systems with inputs, non-linear systems (Extended Kalman Filter)
Re-exports§
pub use kalman::*;
Modules§
- kalman
- Implementation of the Kalman and Extended Kalman Filter. The Kalman type performs state predictions based on a generic System and updates the state based on a Measurement.
- measurement
- A measurement is the input to the Kalman filter during the update stage. It is defined by the measurement vector z (sensor readings), the observation matrix H (or function h(x)) and the measurment noise R.
- system
- A system performs the modelling role in the Kalman filter, predicting the next state based on the current state and any inputs. A system must implement the System trait and either InputSystem or NoInputSystem accordingly.