Objeto de conferencia
A study of prior knowledge insertion in evolutionary fuzzy recurrent controllers design
Registro en:
Autor
Apolloni, Javier
Kavka, Carlos
Roggero, Patricia
Institución
Resumen
A fuzzy controller is usually designed by formulating the knowledge of a human expert into a set of linguistic variables and fuzzy rules. As it is well known, the use of prior knowledge can dramatically improve the performance and quality of the fuzzy system design process. In previous works we have introduced the RFV model, a representation for recurrent fuzzy controllers based on Voronoi diagrams that can represent fuzzy systems with synergistic rules, ful lling the completeness property and providing a simple way to introduce prior knowledge. In this work we present our current approach in the study of the inclusion of prior knowledge in the context of the RFV model Eje: VI Workshop de Agentes y Sistemas Inteligentes (WASI) Red de Universidades con Carreras en Informática (RedUNCI)