Non-integer order system identification with time delays and disturbance rejection

In recent years, system identification for nonlinear systems has been a focal point of research due to its significant implications in control theory and applications. Challenges such as slow convergence, low accuracy, time delays, disturbances, and noise persist in this field. This paper's primary motivation is to enhance the precision and efficiency of system identification for complex nonlinear systems, particularly those represented by non-integer orders. Unlike existing methods, this paper presents a novel method that combines a non-integer-order Hammerstein model with an evolutionary optimization technique and recursive least squares algorithm to accurately estimate system parameters and non-integer orders, while mitigating unknown noise effects. Simulation results demonstrate the effectiveness of this innovative strategy, showcasing its potential applicability in real-world scenarios.