Using Simulink to Generate HDL Code for Validating an Embedded Kalman Filter

Filters are tools that can be used for precision control of digital measurements. Due to the filtering problem of tomographic projections of the soil following a nonlinear model during the measurement, the use of unscented Kalman filter with neural networks was useful to ensure an improvement in signal/noise ratio. Embedded systems may have limited accuracy in numerical calculations due to processing time recursive mathematical functions. More accurate values require more iteration to determine new members of the series causing bottlenecks. In this paper we are presenting a validation of a non linear embedded filter based on the Kalman model.