The European Physical Journal D ( IF 1.5 ) Pub Date : 2021-06-02 , DOI: 10.1140/epjd/s10053-021-00188-3 Nasr Saeed , Cyrille Ainamon , Serdar Çiçek , Sifeu Takougang Kingni , Zhouchao Wei
Abstract
The analytical and numerical analyses as well as prediction with artificial neural network (ANN) for chaos-based artificial intelligence applications of quantum dot light emitting diode (QDLED) are investigated in this paper. The system of equations describing QDLED has three, or one equilibrium points depending on the capture rate from wetting layer into the dot and the injection current. The stability analysis of the equilibrium points reveals the existence of Pitchfork and Hopf bifurcations. The different dynamical behaviors (including steady state, periodic and chaotic behaviors) found in QDLED are illustrated in two parameters bifurcation diagrams, phase portraits and time series. Finaly, the QDLED system is predicted using ANN for chaos-based artificial intelligence applications.
Graphic abstract
中文翻译:
量子点发光二极管中的干草叉和霍普夫分岔:使用人工神经网络进行分析和预测
摘要
本文研究了基于混沌的量子点发光二极管 (QDLED) 人工智能应用的人工神经网络 (ANN) 的解析和数值分析以及预测。描述 QDLED 的方程组具有三个或一个平衡点,具体取决于从润湿层到点的捕获率和注入电流。平衡点的稳定性分析揭示了 Pitchfork 和 Hopf 分岔的存在。在 QDLED 中发现的不同动力学行为(包括稳态、周期和混沌行为)在两个参数分叉图、相图和时间序列中进行了说明。最后,使用 ANN 预测 QDLED 系统用于基于混沌的人工智能应用。