A Comparison of Kalman Filter and Extended Kalman Filter in State Estimation

Kalman filtering has been extensively used technique to estimate the states of discrete-time dynamic systems in a wide range of applications such as control system, signal processing and communication systems etc. A correct and accurate state estimation of linear or non-linear system can be improved by selecting the proper estimation technique. State estimation can be improved by linearizing the nonlinear system by means on different mathematical concepts. Kalman filter algorithms are often used technique that provides linear, unbiased and minimum variance estimates of an unknown state vectors for non-linear systems. In this paper we tried to bridge the gap between the Kalman filter and its variant i.e., Extended Kalman Filter (EKF) with their algorithm and performance, when applied to a non-linear system. To demonstrate the effectiveness, this paper also present the state estimation of the motor driven cart model with inverted pendulum through Kalman filter and Extended Kalman filtering which is moving with a constant force. The algorithms discussed here, has been proved more effective when only noisy observation data is available.