Embedded Systems Dictionary

n. A Kalman-like filter that is more robust in the face of modeling errors and noise uncertainty. The goal is to find a state estimate, J < 1/ g, where g is an arbitrary value chosen to increase the accuracy of the estimate. The estimate can be calculated by the following equations.

where I is the identity matrix, Kk is the filter gain matrix, initial state estimate is a best guess, and P0 is set to give acceptable filter performance. The ratio of the estimation error to noise will always be less than 1/ g, regardless of the amount and type of noise in the actual environment!
The Kalman filter was very successfully applied to aerospace control problems in the 1960s. However, when it was later applied to problems in industrial control, it was found to be problematic. Accurate system models are not as readily available for industrial problems. In addition, engineers rarely understand the statistical nature of the noise processes that impinge on such systems. The H-infinity filter is more robust in the face of such uncertainties.
(hardware) abbr. See hardware.
(as letters) abbr. See high-availability.
1. v. To write software, typically in an unprofessional manner.
2. v. To attempt access to or control of a computer system in an unauthorized manner.
3. n. A clever programming trick or workaround.
adj. Said of a communications link over which data can be sent in one direction at...