State-space modeling and estimation of coupled PDE system: A case study in Sintering process

This paper presents a generic technique of state space modeling of systems described by coupled Partial Differential Equations. The Finite Difference Method has been used to obtain a discretized form of the equations which are then cast into a state space model. An Extended Kalman Filter based algorithm can be used with the obtained state space model to estimate the spatially distributed profile of the dependent variables. This technique has been framed into an algorithm which is discussed in details in this work. Fire line estimation in Sintering process has been carried out using the algorithm, as an example of an industrial process.