H-2/H-infinity-Neural-Based FOPID Controller Applied for Radar-Guided Missile


Yaghi M., EFE M. Ö.

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, cilt.67, sa.6, ss.4806-4814, 2020 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 67 Sayı: 6
  • Basım Tarihi: 2020
  • Doi Numarası: 10.1109/tie.2019.2927196
  • Dergi Adı: IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Aerospace Database, Applied Science & Technology Source, Business Source Elite, Business Source Premier, Communication Abstracts, Compendex, Computer & Applied Sciences, INSPEC, Metadex, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.4806-4814
  • Hacettepe Üniversitesi Adresli: Evet

Özet

In this paper, a neural tuning technique is proposed and applied to a fractional-order proportional-integral-derivative (FOPID) controller. The proposed controller is applied to a radar-guided missile which is used for tracking high-speed moving targets in defense systems. The proposed neural tuning along with optimization method is intended to improve the tracking performance of the proportional navigation (PN) system and the stability of the missile trajectory during flight time. Due to the coupled and nonlinear dynamics of the considered missile, we propose a new control structure that integrates the FOPID controller into the PN system of the missile which results in better stability properties for the missile. In order to tune the FOPID, we propose a neural tuning technique that starts with a genetic algorithm-based optimizer and continues with a neural network-based scheme. This speeds up finding a near-optimal solution and refining it effectively. Then, optimization process is applied within the neural tuning technique to achieve better stability and tracking performance for the missile during the whole flight time. The proposed controller is compared with the standard PID controller tuned by the conventional Ziegler-Nichols (ZN) tuning method as well as particle swarm optimization (PSO) method, and the simulation results proved the superiority of the proposed tuning method over ZN- and PSO-based tuning approaches.