Modül Yapay Sinir Ağları ile Doğrusal Olmayan Sistemlerin Denetimi

Yapay sinir ağı (YSA) tarafından gerçekleştirilen hesaplama, birbiriyle iletişim kurmadan girdi uzayı üzerinde çalışan iki veya daha fazla modüle (alt sistemler) ayrıştırılabiliyorsa, sinir ağı modülerdir (MYSA). Modülerlik, karmaşık bir hesaplama görevini daha basit görevlere bölerek girdi uzayının farklı bölgelerini öğrenip uzmanlaşma eğilimindeki modüllerin bireysel çözümlerini birleştirme yaparak çözüme izin veren böl ve fethet ilkesinin bir tezahürüdür. Bu çalışmada, doğrusal olmayan iki sistemin MYSA ile modellenmesi ve denetim başarıları incelenerek elde edilen sonuçlar YSA ile karşılaştırılmıştır. Sistemlerin modelleme ve denetiminde yapılan karşılaştırma sonuçlarına bakıldığında MYSA performansının YSA’ ya göre iyi olduğu tespit edilmiştir.

Anahtar Kelimeler:

Modül yapay sinir ağları, doğrusal olmayan sistemler, modelleme ve denetim

Modular Neural Network Control of Nonlinear Systems

A neural network is modular (MNN) if the computation performed by an artificial neural network (ANN) can be decomposed into two or more modules (subsystems) operating on the input space without communicating with each other. Modularity is a manifestation of the divide-and-conquer principle, which allows a solution by dividing a complex computational task into simpler tasks, combining individual solutions of modules that tend to learn and specialize in different regions of the input space. In this study, the modeling of two nonlinear systems with MCA and the audit successes were examined and the results obtained were compared with ANN. When the comparison results made in the modeling and inspection of the systems are examined, it has been determined that the MYSA performance is better than the ANN.

Keywords:

Module neural networks, nonlinear systems, modeling and control,

PDF

___

Uğuz, S., Makine Öğrenmesi Teorik Yönleri ve Python Uygulamaları ile Bir Yapay Zeka Ekolü. 2 ed. 2021: Nobel Akademik Yayıncılık. 300.
Chollet, F., Python ile Derin Öğrenme. 1. Baskı ed. Buzdağı Yayınevi, Ankara. 2019.
Prasad, N., et al., A reinforcement learning approach to weaning of mechanical ventilation in intensive care units. arXiv preprint arXiv:1704.06300, 2017.
Perchiazzi, G., et al., Monitoring of total positive end-expiratory pressure during mechanical ventilation by artificial neural networks. Journal of clinical monitoring and computing, 2017. 31: p. 551-559.
Özyurt, F., et al., Brain tumor detection based on Convolutional Neural Network with neutrosophic expert maximum fuzzy sure entropy. Measurement, 2019. 147: p. 106830.
Liu, Y.-J., et al., Adaptive neural network control for a class of nonlinear systems with function constraints on states. IEEE Transactions on Neural Networks and Learning Systems, 2021.
Kubat, P.D.C., Matlab Yapay Zeka ve Mühendislik Uygulamaları. 2019: Abaküs.
Izzo, D., M. Märtens, and B. Pan, A survey on artificial intelligence trends in spacecraft guidance dynamics and control. Astrodynamics, 2019. 3: p. 287-299.
Kamalapurkar, R., et al., Reinforcement learning for optimal feedback control. 2018: Springer.
Wen, T., et al., Hop reachable domain on irregularly shaped asteroids. Journal of Guidance, Control, and Dynamics, 2020. 43(7): p. 1269-1283.
Yin, S., J. Li, and L. Cheng, Low-thrust spacecraft trajectory optimization via a DNN-based method. Advances in Space Research, 2020. 66(7): p. 1635-1646.
Cheng, L., et al., Fast solution continuation of time-optimal asteroid landing trajectories using deep neural networks. Acta Astronautica, 2020. 167: p. 63-72.
Huang, Y., S. Li, and J. Sun, Mars entry fault-tolerant control via neural network and structure adaptive model inversion. Advances in Space Research, 2019. 63(1): p. 557-571.
Yang, T., et al., Neural network-based adaptive antiswing control of an underactuated ship-mounted crane with roll motions and input dead zones. IEEE Transactions on Neural Networks and Learning Systems, 2019. 31(3): p. 901-914.
Zhou, N., Y. Kawano, and M. Cao, Neural network-based adaptive control for spacecraft under actuator failures and input saturations. IEEE transactions on neural networks and learning systems, 2019. 31(9): p. 3696-3710.
He, D., Z. Liu, and Y. Jiang, An intuitive model for on-axis pulse evolution of ultrashort pulsed Gaussian beams diffracted from a circular aperture. Journal of Modern Optics, 2015. 62(8): p. 620-625.
Qiao, J., X. Guo, and W. Li, An online self-organizing modular neural network for nonlinear system modeling. Applied Soft Computing, 2020. 97: p. 106777.
Aytaş, G., Sözlü çeviri eğitiminde bilişsel incelemeler: SAÜ çeviribilim bölümü hazırlık, 2. ve 4. sınıflar örneği, in Sosyal Bilimler Enstitüsü, Ph.D. 2019, Sakarya Universitesi.
Baloğlu, Ş., Modül yapay sinir ağları ile doğrusal olmayan sistemlerin denetimi, Fen Bilimleri Enstitüsü. 2003, Fırat Üniversitesi.
Intisar, C.M. and Q. Zhao. A selective modular neural network framework. 2019 IEEE 10th International Conference on Awareness Science and Technology (iCAST). 2019. IEEE.
Ma, J., et al., Feature guided Gaussian mixture model with semi-supervised EM and local geometric constraint for retinal image registration. Information Sciences, 2017. 417: p. 128-142.
Khanmohammadi, S. and C.-A. Chou, A Gaussian mixture model based discretization algorithm for associative classification of medical data. Expert Systems with Applications, 2016. 58: p. 119-129.
Chowdhury, M.I., et al., CMNN: Coupled modular neural network. IEEE Access, 2021. 9: p. 93871-93891.