Unscented Kalman Filter-Based Bayesian Filtering of RC Circuit

Abstract

In this paper, we present output voltage estimation of second order resistor capacitor (RC) low pass filter (LPF) circuit using the unscented Kalman filter (UKF). Firstly, with the use of state space depiction we formulated a deterministic model of RC LPF circuit. The deterministic model is then transformed into the stochastic model by introducing noise to circuit components and input source. Finally, the MHE method and UKF method have been simulated to obtain the optimal state estimate when the circuit elements are affected by the Gaussian noise process. The derived mathematical expressions, so obtained, are simulated in MATLAB software. Then the comparison is shown between the estimated output values and PSPICE simulated values (actual output values) and digression of estimated output is expressed in the context of root mean square error (RMSE) to represent a loss function. The RMSE values validates the efficacy of proposed method with respect to the latter method.