A framework for designing cognitive trajectory controllers using genetically evolved interval type-2 fuzzy cognitive maps


Amirkhani A., Shirzadeh M., Kumbasar T., Mashadi B.

INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS, cilt.37, sa.1, ss.305-335, 2022 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 37 Sayı: 1
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1002/int.22626
  • Dergi Adı: INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, PASCAL, Aerospace Database, Applied Science & Technology Source, Communication Abstracts, Compendex, Computer & Applied Sciences, INSPEC, Metadex, zbMATH, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.305-335
  • Anahtar Kelimeler: fuzzy cognitive map, genetic algorithm, mobile robot, real-time, type-2 fuzzy sets, uncertainty, NONHOLONOMIC MOBILE ROBOT, SYSTEMS, TRACKING, PERFORMANCE, NETWORK
  • İstanbul Teknik Üniversitesi Adresli: Evet

Özet

In this paper, we provide a complete framework for the design of genetically evolved cognitive tracking controller based on interval type-2 (IT2) fuzzy cognitive map (FCM). We construct the cognitive controller based on a nonlinear controller by transforming its representation into a FCM. This representation gives the opportunity to prove the stability of the cognitive controller in the framework of nonlinear control theory. Moreover, with the deployment of IT2-fuzzy sets which are known to be capable to handle high level of uncertainty, the proposed cognitive controller has the ability to deal with uncertainty that are encountered in real-time world applications. To accomplish the design of the cognitive controller, we present a systematic approach based on genetic algorithm to optimize its parameters and learn fuzzy rules by extracting them from model space (e.g., a set of rules). Within the paper, all steps in constructing and designing the IT2-FCM-based cognitive controller are presented. We first show the performance improvements of the proposed IT2-FCM-based tracking controller with extensive and comparative simulation results and then with experimental results that were collected on real-world mobile robot. The results clearly show the superiority of proposed cognitive control systems when compared to its conventional and fuzzy controller counterparts. We believe that the proposed genetically evolved design approach of the IT2-FCM-based cognitive controller will provide a bridge between the well-developed cognitive sciences and control theory.