Expert System for Industrial Environment Based on Self-Learning Rules
Expert system, industrial environment, decision support system, self-edited rules.
This work presents a methodology of how to acquire and represent knowledge through automatic logic rules for a simulated industrial plant. The initial knowledge about an industrial process can be acquired through a specialist who interprets situations present in the plant and can describe what is occurring. In the work, a way of acquiring statistical knowledge of the plant during the execution of its processes is presented, using a method of online clustering known as TEDA-Cloud being modified to improve performance. The representation of knowledge is described through the manipulation of a neural network known as CILP and its own symbology is described to represent the logical variables drawn from the process signals. The results show an efficiency in interpreting the rules and acceleration in the process of clustering and classifications of the standards that define the rules.