Kalman Filter Based State Estimation for Linearized Twin Rotor System

In this paper, states estimation for MIMO Twin Rotor System (TRS) is performed. In practical, often, system states are unknown or immeasurable. In applications like state feedback control design, fault diagnostics or system monitoring, the states information is needed. The precise estimation of states can be done and verified using Kalman filter as state observer. For generation and confirmation of correct states estimate, DC inputs (resembling practical inputs) are developed and outputs from TRS model are collected. The I/O data is invoked in Kalman filter and resulting states estimates are verified for correctness first by examining the state error covariance and then by comparing the evolution of actual and estimated states. The results show that the Kalman state estimates are highly precise and fast convergent to the actual states. The extracted TRS states information is usable for fault diagnostics, control design, system monitoring or as an alternate to costly instruments used to measure system states in industries.