Angular speed is the main variable used to regulate electrical drives, where incremental encoders and tachometers are normally used to provide an effective position and speed measurement using time-fixed (the elapsed time is fixed, and the angle is measured by accumulating encoder pulses) or pulse-fixed (the number of pulses is fixed, and the time is measured by accumulating cycles of a reference clock) estimators. In both cases, errors are made in the speed calculation, which can be particularly large at low (time-fixed estimator) or high (pulse-fixed estimator) rotor speeds. Each method has different sources of errors and optimal ranges for speed measurement, which generate uncertainty in the efficiency of the electric drive. This work proposes a new speed estimation method that uses both frequency and period speed techniques. The implemented algorithm combines the estimations obtained based on the actual rotor speed and the inverse variance weighting method. The obtained results state that the resulting measured speed outperforms the time-fixed and pulse-fixed estimations across any speed range.
Barrero et al. (Tue,) studied this question.