METHOD FOR CONTROLLING A DISTRIBUTED CONTROL SYSTEM

The invention relates to a method for controlling a distributed control system (13), which distributed control system (13) is configured for controlling a chemical plant (1), wherein the chemical plant (1) is configured for a chemical production process, wherein a training set (9) of data points is generated, wherein generating the training set (9) of data points comprises applying at least one analytic function, which at least one analytic function represents a steady-state model (8) and links variables of the chemical production process, wherein the data points correspond to variable values of the linked variables (14), wherein a black box model (10) is trained with the training set (9) of data points, wherein an objective function is defined on optimization variables, which optimization variables are at least partially comprised by the linked variables (14), which objective function defines an optimum for the optimization variables, wherein variable constraints (12a, b) are defined which apply to the linked variables (14), wherein the following iteration steps are repeated to arrive at input variables (16) for controlling the distributed control system (13): applying input variables (16) to a distributed control system (13), obtaining process values (11) for at least some of the linked variables (14) from the distributed control system (13), determining a mismatch between the obtained process values (11) and the values for the linked variables (14) according to the black box model (10) based on the same input variables (16), modifying the objective function based on the determined mismatch, modifying the variable constraints based on the determined mismatch, generating new input parameters based on the modified objective function and the modified variable constraints. L'invention concerne un procédé de commande d'un système de commande distribué (13), lequel système de commande distribué (13) est configuré pour commander une usine chimique (1), l'usine chimique (1) étant configurée pour effectuer un processus de production chimique, dans lequel un ensemble d'apprentissage (9) de points de données est généré, la génération de l'ensemble d'apprentissage (9) de points de données consistant à appliquer au moins une fonction analytique, cette fonction analytique représentant un modèle d'état stable (8) et liant des variables du processus de production chimique, les points de données correspondant à des valeurs de variable des variables liées (14), un modèle