FPGA implemenation of extended Kalman filter for speed-sensorless control of induction motors

This presentation describes a hardware-in-the-loop (HIL) system including the implementation of an extended Kalman filter (EKF)-based estimator on a field-programmable gate array (FPGA) for speed-sensorless control of induction motors (IM). The implemented EKF algorithm simultaneously estimates stator currents (i_sα and i_sβ), stator fluxes (ψ_sα and ψ_sβ), rotor angular velocity (ω_m) and load torque (t_L) by assuming that stator voltages and currents are available. The HIL system also includes a stator-current- and flux-based IM model which provides actual stator currents to the EKF algorithm and is also utilized to validate the flux, speed and load torque estimations of the implemented EKF algorithm. A Virtex-5 VSX110T FPGA evolution board is used for this real-time application. The FPGA board is programmed via VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) in order to develop both the IM model and the EKF algorithm. ISE Design Suit Interface is used as the debugger and compiler. The results obtained from the EKF and IM model developed on the FPGA are graphically compared to verify the sufficiency of estimation performance of the EKF algorithm and demonstrate that the EKF algorithm is implemented successfully with less computational time (less sampling time for each recursive operation) due to the inherent parallel signal processing ability of the FPGA.