Journal of the Franklin Institute, Год журнала: 2024, Номер unknown, С. 107470 - 107470
Опубликована: Дек. 1, 2024
Язык: Английский
Journal of the Franklin Institute, Год журнала: 2024, Номер unknown, С. 107470 - 107470
Опубликована: Дек. 1, 2024
Язык: Английский
Fractal and Fractional, Год журнала: 2024, Номер 8(9), С. 548 - 548
Опубликована: Сен. 20, 2024
In this study, a model-free PIφ-sliding mode control ( PIφ-SMC) methodology is proposed to synchronize specific class of chaotic fractional-order memristive neural network systems (FOMNNSs) with delays and input saturation. The Lyapunov stability theory used design two-level PIφ-SMC which can effectively manage the inherent behavior delayed FOMNNSs achieve finite-time synchronization. At outset, an initial sliding surface introduced. Subsequently, robust designed as second surface, based on proportional–integral (PI) rules. asymptotic both surfaces demonstrated. final step involves dynamic-free law that against system uncertainties, saturations, delays. independence rules from functions accomplished through application norm-boundedness property in states. soft actor-critic (SAC) algorithm deep Q-Learning utilized optimally adjust coefficients embedded controller’s structure. By maximizing reward signal, optimal policy found by SAC agent. This approach ensures motion meets reachability condition within finite time. validity protocol subsequently demonstrated extensive simulation results two numerical examples.
Язык: Английский
Процитировано
13IEEE Transactions on Network Science and Engineering, Год журнала: 2024, Номер 11(5), С. 4564 - 4576
Опубликована: Май 29, 2024
Considering the infinite memory property of fractional differential equations, and combining advantages fuzzy logic systems reservoir computing echo state networks, a fractional-order network is proposed for estimating uncertain terms nonlinear systems. In order to improve its learning performance reduce response oscillations, modified output weight law constructed by recording historical data over period time in regression matrix both iterative domains, which prediction error defined without using strict condition, i.e., interval excitation. Furthermore, leveraging Lyapunov stability theory frequency distribution model, suitable integral energy functions are formulated based on backstepping framework, proving that composite control scheme can maintain boundedness all signals within closed-loop system. Finally, Chua-Hartleys chaotic system single-machine-infinite- bus power adopted simulation validate validity scheme.
Язык: Английский
Процитировано
6Mathematics, Год журнала: 2024, Номер 12(16), С. 2536 - 2536
Опубликована: Авг. 16, 2024
This paper explores a general class of singular kernels with the objective designing new families uniformly continuous sliding mode controllers. The proposed controller results from filtering discontinuous switching function by means Sonine integral, producing control signal, preserving finite-time motion and robustness against but unknown not necessarily integer-order differentiable disturbances. principle dynamic memory resetting is considered to demonstrate stability. A set sufficient conditions design kernels, above characteristics, presented, several examples are exposed propose approaches. Simulation studied illustrate feasibility some schemes.
Язык: Английский
Процитировано
0Journal of the Franklin Institute, Год журнала: 2024, Номер unknown, С. 107273 - 107273
Опубликована: Сен. 1, 2024
Язык: Английский
Процитировано
0Nonlinear Dynamics, Год журнала: 2024, Номер unknown
Опубликована: Окт. 6, 2024
Язык: Английский
Процитировано
0Journal of the Franklin Institute, Год журнала: 2024, Номер unknown, С. 107345 - 107345
Опубликована: Окт. 1, 2024
Язык: Английский
Процитировано
0Journal of the Franklin Institute, Год журнала: 2024, Номер unknown, С. 107406 - 107406
Опубликована: Ноя. 1, 2024
Язык: Английский
Процитировано
0Journal of the Franklin Institute, Год журнала: 2024, Номер unknown, С. 107470 - 107470
Опубликована: Дек. 1, 2024
Язык: Английский
Процитировано
0