Multirate Data Assimilation in a Cultivation Process

Functional Modeling View on Product and Process Engineering in Design and Operations

“…106 Through Monte Carlo simulations, the uncertainty of the necessary parameter estimation is feasible, based on suitably designed experiments. 107 However, an important consideration is to use online filtering methods, e.g., various forms of Kalman filtering 108 in order to combine online data with off-line data 109 in a suitable manner. Often, the MPC solution is used to develop a simpler but perhaps more robust solution, based on the stabilizing control layer with a few interactions with control of additional degrees of freedom.…”

High-gain extended Kalman filter for continuous-discrete systems with asynchronous measurements

“…Let us consider a boat 4 moving in an area delimited by two beacons (denoted by A and B). The state of this system is x = (x 1 , x 2 , γ) ∈ R 3 , where as it is schematised in Figure 1a, (x 1 , x 2 ) ∈ R 2 is the position of the boat w.r.t.…”

“…Such systems are often met in practice, for instance in global positioning problems, as in [33], or in the field of drone control [10]. Likewise, one can be confronted with such asynchronous systems in the fields of submarine robotics, as can be seen from [3,31,6], chemical engineering [41], or cultivation engineering [4].…”

“…As emphasised in [43], this state estimation problem can be tackled by considering one of the three following options. First, model the state dynamics as discrete and apply a known estimator for discrete state systems -see for example [4] in the linear setting. Second, lift the measurements into the space of continuous functions, e.g.…”

“…In [28], the authors built upon the ideas of [25,1] where an already designed continuous-time Luenberger-like observer was coupled with asynchronous intersamples predictors. Finally, the problem under consideration was also addressed by using multirate versions of the Kalman filter, see for instance [27,4,24,17]. In particular, in [17], the authors studied the exponential convergence of the proposed observer and the preservation of observability for multirate systems.…”