Non-linear State Estimation of a DC Series Motor: A Review and a Novel Tuning Method

Abstract

This paper computes different non-linear estimation techniques for the state variables of a DC series motor. The measured variable is the rotor speed, and the estimated variable is the current. Adding a null dynamic, the load torque can be estimated too. Each estimator is based on mathematical structures found in the literature. It does a development and comparison among the extended Luenberger observer (ELO), the extended Kalman filter (EKF) and the extended sliding mode observer (ESMO). The last estimator is designed using the static poles assignment, and a non-static pole assignment is proposed with the Ackerman’s formula as a novel and easy method to recalculated the constants for the estimator in each iteration. Each estimator is simulated under sensor noise, plant-model mismatch, changes in the input and disturbances. The simulation results are contrasted using integral square error (ISE) and making conclusions about it.