Automated solving of control tasks in the production environment based on knowledge-based systemsCopyright: © IAT
Knowledge-based systems are used for knowledge representation and modeling, for example to support or reduce manual operations in the production environment and to solve problems automatically. Furthermore, these systems improve the understanding of control methods. There are several methods that are used to solve control tasks. These methods often have prerequisites that must be fulfilled before use. For this purpose, classification and categorization of problem definitions and methods for solving them are considered in more detail.
The requirements for the knowledge-based system are the interpretation of the problem, the extensibility and changeability of the knowledge base, the identification selection and validation of solution methods. The architecture contains the interpretation, the processing as well as the knowledge base, in which the methods are stored. The storing of methods is done in a modularized way, so that an extension is possible.
The project enables a suitable selection of methods from a modular knowledge base. The goal is not to apply the individual methods from the respective domains, but to counteract the growing complexity from different domains and, if necessary, to reduce the need for lacking experts by identifying suitable methods and thus to achieve automated solving from different disciplines for control tasks. Using a representative example, which reflects a real plant, individual solution methods and approaches from the knowledge-based system are applied for validation and evaluation.