Online model maintenance in real-time optimization methods

Simultaneous online model identification and production optimization using modifier adaptation

“…Along the same lines, [5,13] designed a MA variant for cases where several models of the plant are available but only one is used at a time. Here, the cost-minimising control input is computed while identifying the model that best fits the plant from an available set.…”

“…As mentioned, the closest work on the joint identification and control input optimization problem addressed in this paper appears in [5,13]. While modifier adaptation is exploited to find the optimal control, identification is based on a heuristic of choosing the least norm defined on the modifiers across the set of models.…”

“…Although modifier adaptation is traditionally used for control input optimization, the modifiers contain valuable information about the relationship between a model and the plant. This observation has been exploited in [5,13] in order to perform model identification. In particular, the modifiers can be used for discriminating between different models, which can be helpful in cases that there are unknown features of the process that need to be modeled.…”

“…This presents the theoretical possibility of the model updates ending up in a limit cycle. As discussed in [13], unless the models in the available set are distinguishable (i.e. can be uniquely identified given a set of plant measurements and gradient estimation estimates), the model selection conclusions drawn at iteration k can be different than those of iteration k + 1.…”

“…Here, the approaches of Section 5 are implemented on a CSTR, where reactants A and B generate product C. The case study is based on [13].…”