State and parameter estimation via minimum distortion filtering with application to Chemical Process Control

State and parameter estimation are cornerstone problems in Chemical Process Control. When the problem is linear and gaussian, the celebrated Kalman Filter provides a simple and elegant solution to the recursive filtering problem. However, many practical systems (including most Chemical Processes) are nonlinear. In this case, the Kalman Filter cannot be directly applied and other methods are necessary. In this paper, we describe a new approach to Nonlinear Filtering known as Minimum Distortion Filtering (MDF). We show that this method is computationally tractable for typical Chemical Process Control problems including estimation of unmeasured states and unknown parameters such as activation energy or frequency factor constants. We illustrate by a simulation study of a Continuous Stirred-Tank Reactor (CSTR).