Date of Award

Spring 2019

Document Type


Degree Name

Master of Science (MS)


Electrical and Computer Engineering

First Advisor

Yaz, Edwin E.

Second Advisor

Schneider, Susan C.

Third Advisor

Jeong, Chung S.


This thesis is on state estimation for various system and measurement models with uncertain dynamics. The uncertain dynamics may be due to imprecise system modeling or change in parameters due to varying environmental conditions. Uncertainty may be a result of malicious acts such as hacking of sensors or actuators in the system. Uncertainty may also be a result of external disturbances whose waveforms, magnitudes and arrival times may not be known. These types of model uncertainties will be considered and different estimators will be implemented to deal with such uncertainties in state estimation. In this thesis, for linear stochastic systems with additive noise, the measurements and input are available and noises statistics are known, Kalman filter is used to estimate the state. However, for nonlinear systems, Extended Kalman filter is used under the same conditions. When noise statistics are unknown, H-infinity filter is used to estimate the state of the system if the noises are assumed to be of finite energy. For identification of parameters, coefficients in transfer functions are identified by using Kalman and H-infinity filters. By using Extended Kalman and H-infinity filters, unknown parameters of the state-space model can be estimated. For parameters whose range of values is available, a bank of Kalman filters is used to find the actual values. For detection of an intrusion signal that attacks a sensor or actuator of a system, there are several methods considered in this thesis, including the sample mean method, Kalman filter method and stochastic parameter estimation method.The simulation results of the various applications of these filters will be presented and the performance of these filters will be discussed.