Simulation and electronic design of a chaotic 5d artificial neural network

This study investigates the complex dynamics of a five-dimensional artificial neural network (ANN) system with a hyperbolic tangent activation function. The objective is to analyze the chaotic behavior and validate multi-platform implementation of the proposed system. The methodology involves numerical simulations in MATLAB-Simulink and LabVIEW, followed by circuit design in Multisim for hardware feasibility verification. Through systematic parameter variation, the system exhibits rich dynamical regimes, including periodic oscillations and chaos. Lyapunov exponent analysis reveals a positive value (LE1 = 0.035) and Kaplan-Yorke dimension DKY = 2.488, confirming chaotic dynamics with fractional attractor geometry. Bifurcation diagrams demonstrate transitions between periodic and chaotic states as parameter b varies. The Multisim circuit simulation results demonstrate excellent agreement with MATLAB-Simulink and LabVIEW outputs, confirming the system's robustness and practical implementability across different platforms for potential applications in secure communications and cryptography.