FINITE-TIME STABILITY OF FRACTIONAL-ORDER NEURAL NETWORKS WITH PROPORTIONAL DELAY
Yanrong An, Muhammad Aamir Ali, Jarunee Soontharanon, Thanin Sitthiwirattham · Fractals · 2025
Neural networks are valuable tools for modeling complex dynamical systems in various scientific and engineering domains. Fractional-order neural networks, specifically, offer enhanced accuracy in capturing memory and hereditary effects compared to their integer-order counterparts. This paper aims to establish easily verifiable conditions ensuring the finite-time stability of fractional-order neural networks with proportional delay. To accomplish this, we first generalize the existing framework of fractional-order neural networks to encompass a wider range of practical applications. We then derive two novel fractional-order Gronwall inequalities incorporating proportional delay, which are essential analytical tools. Finally, we utilize these inequalities to obtain two rigorous and practical criteria that guarantee the finite-time stability of the proposed network system.