The real-time digital implementation of the Kalman filter requires a very fast signal processor specialised and optimised to perform complex mathematical calculations and manipulate a large amount of data. In fact, the algorithm is computationally intensive, and all of the steps involved require vector and/or matrix operation. In this paper the analysis of an actual fixed point implementation of the Kalman filter is discussed based on a last generation μC DSP (TMS320F240). The considered case study refers to rotor speed and position estimation in AC drives when using low resolution position transducers. It is shown that the adopted processor is suitable for an efficient and relatively simple implementation of the filter allowing to increase the resolution in speed calculation with respect to the classical differentiation solution.
Fixed point implementation of Kalman filtering for AC drives: a case study using TMS320F24x DSP
TURSINI, MARCO
2000-01-01
Abstract
The real-time digital implementation of the Kalman filter requires a very fast signal processor specialised and optimised to perform complex mathematical calculations and manipulate a large amount of data. In fact, the algorithm is computationally intensive, and all of the steps involved require vector and/or matrix operation. In this paper the analysis of an actual fixed point implementation of the Kalman filter is discussed based on a last generation μC DSP (TMS320F240). The considered case study refers to rotor speed and position estimation in AC drives when using low resolution position transducers. It is shown that the adopted processor is suitable for an efficient and relatively simple implementation of the filter allowing to increase the resolution in speed calculation with respect to the classical differentiation solution.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
