Geri Dön

Integrated physics-based model for low power bio-inspired computing

Başlık çevirisi mevcut değil.

  1. Tez No: 401236
  2. Yazar: DİNÇER ÜNLÜER
  3. Danışmanlar: PROF. AVIK GHOSH
  4. Tez Türü: Doktora
  5. Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
  6. Anahtar Kelimeler: Belirtilmemiş.
  7. Yıl: 2013
  8. Dil: İngilizce
  9. Üniversite: Unıversıty Of Vırgınıa
  10. Enstitü: Yurtdışı Enstitü
  11. Ana Bilim Dalı: Belirtilmemiş.
  12. Bilim Dalı: Belirtilmemiş.
  13. Sayfa Sayısı: Belirtilmemiş.

Özet

Özet yok.

Özet (Çeviri)

In silicon technology, device metrics such as power, delay, and reliability are hitting a brick wall, prompting industry to seek alternative materials, designs, and switching principles. This thesis investigates bio-inspired computing components through a case study that uses signal processing in axonal systems. A physics-based model is developed for each component of the action potential generation in an axon. The two main building blocks of an axon system consist of sequentially activated voltage-gated ion channels (the switch) and adenosine triphosphatase powered ion pumps (the battery). However, the primary aim in this investigation is not to replicate the biological details, but to replace the quantitatively accurate yet empirical operational equations with qualitative, physics-based microscopic models. The primary purposes of focusing on the physics in this study are to deconstruct biological components in terms of solid-state analogues and to understand their potential use for e_cient low-power switches, even within conventional Boolean logic. Accordingly, a nano-electro-mechanical _eld-e_ect transistor, i.e., a relay, is used to model the ion channel in order to explain how these ion channels are similar to cantilevers that can switch e_ciently. The model developed matches the 7:5mV=decade sub-thermal switching of the low-power sodium channels because it captures both the phase transition from metastable state to stable state and the charge multiplication e_ect. The electronic ratchets act as the ion pump that uses an energy source (adenosine triphosphate hydrolysis) to power ion ow against a concentration gradient to recharge the battery. This dissertation develops an electronic analogue of a ratchet, essentially a non-equilibrium diode. Furthermore, the research demonstrates that with the ratchet, universal Boolean logic is e_ciently implemented and in fact, proves analytically why the electronic ratchet is low power due to ratchet being a voltage-controlled current source. The detailed modeling of the nano-mechanical relays and the ratchets using transport theory identify and conclude the underlying physics behind these devices' power e_cient operations and limitations while making connections to the individual components in the axonal networks.

Benzer Tezler

  1. Propulsion-airframe integration for low-boom supersonic aircraft

    Düşük gürültülü sesüstü hava araçlarında itki-gövde entegrasyonu

    RUMED İMRAK

    Yüksek Lisans

    İngilizce

    İngilizce

    2023

    Savunma ve Savunma Teknolojileriİstanbul Teknik Üniversitesi

    Savunma Teknolojileri Ana Bilim Dalı

    PROF. DR. MELİKE NİKBAY

  2. Gan güç transistörlerinin anahtar kayıplarının nöral ağ tabanlı modellenmesi

    Neural network based modeling of switching losses in gan power transistors

    ALİİMZA DAŞDEMİR

    Yüksek Lisans

    Türkçe

    Türkçe

    2026

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    Elektrik Mühendisliği Ana Bilim Dalı

    DOÇ. DR. MEHMET ONUR GÜLBAHÇE

  3. Hidroelektrik santrallerde modal özelliklerin derin öğrenme esaslı analizi ve deprem kayıtlarında anomali tahmini

    Deep learning-based analysis of modal properties in hydroelectric power plants and anomaly prediction in earthquake records

    GÜRHAN TOKGÖZ

    Doktora

    Türkçe

    Türkçe

    2026

    İnşaat MühendisliğiGazi Üniversitesi

    İnşaat Mühendisliği Ana Bilim Dalı

    DR. ÖĞR. ÜYESİ EDA AVANOĞLU SICACIK

  4. Test verilerine dayalı, makine öğrenmesi ve derin öğrenme yöntemleri ile batarya sağlık durumu tahmini

    Battery state of health estimation based on test data using machine learning and deep learning methods

    MEHMET ALİ ARSLANTAŞ

    Yüksek Lisans

    Türkçe

    Türkçe

    2025

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    Mekatronik Mühendisliği Ana Bilim Dalı

    PROF. DR. FİKRET ÇALIŞKAN

  5. Yansıtıcı dizi anten birim hücrelerinin pca ve bayesçi optimizasyon destekli yapay sinir ağları ile modellenmesi

    Modeling of reflectarray antenna unit cells using pca and bayesian optimization-aided artificial neural networks

    MUHAMMET TARIK ERKAN YILDIZ

    Yüksek Lisans

    Türkçe

    Türkçe

    2026

    Elektrik ve Elektronik MühendisliğiYıldız Teknik Üniversitesi

    Elektronik ve Haberleşme Mühendisliği Ana Bilim Dalı

    DOÇ. DR. PEYMAN MAHOUTİ