The Kalman filter is an optional estimation algorithm. It can also be thought of as a state observer for stochastic systems. Kalman filters can find estimates for states that cannot be measured directly. Kalman filters only work for linear systems with linear noise. They can also be used for sensor fusion of different noisy sensors. It is commonly used in guidance and navigation systems, computer vision systems, and signal processing algorithms. Kalman filters use noisy plant models with a state observer to determine the expected value of unobservable states that may contain process or measurement noise. It is selecting gains that are an optimal tradeoff between the model error and sensor noise. One of the first applications of the Kalman filter was on the Apollo Project. They all have two main steps, a prediction and an update step. Kalman filters can deal with systems that propagate in time.
At each next timestep, the state of a system has a greater uncertainty. The optimal way to determine the state of a system is to combine the measurement that contains measurement noise with the prediction that contains process noise. This is accomplished by multiplying the probability functions of the measurement and the state estimate.
[[Discrete Kalman Filter Equation]]
[[Modeling the GPS Space Segment]]
[[Active Flutter Suppression]] – traditionally uses Kalman filter.
[[Battery State of Charge]] – can use a Kalman filter for estimation
[[State Observers]]
[[Sensor Fusion with Kalman Filters]]
[[Kalman Filter for Nonlinear Systems]]
[[Extended Kalman Filter]] – for nonlinear systems
[[Unscented Kalman Filter]]
[[Tools for Identifying Wnt pathways]] – Kalman filter can be used as a data analysis method.
[[Cartesian Guidance Filter]] – uses a Kalman filter
[[Range Bearing Tracking Problem]] – Kalman filter can be used to determine the actual location of the tracked object
[[Recursive Least Squares]] – Kalman filter is based on the recursive least squares
OGLI Simulation Program – used Kalman filter for guidance law input
[[Inertial Navigation System]] – can use a Kalman filter
[[Pilot-Vehicle Model]] – uses Kalman filter as an observer
[[X-38 Sideslip Estimation]] – kalman filters were tried but abandoned as it was difficult to tune
[[BQM-167 Software Architecture]] – used a 17-state Kalman filter
- GyroNoiseAllan2021
- UnderstandingKalmanFilters
- UnderstandingKalmanFiltersd
- bevlyMECH4420Lectureb
- naruokaApplicationInertialGNSS2021
Backlinks:
[[Controller Design]]
State Estimation Algorithms
[[State-Space Model]]