Enhanced Adaptive Control of Nonlinear Systems Based on a Lyapunov‐Stability‐Based Modified Context‐Layered Elman Recurrent Neural Network With Adaptive Learning Rate

ABSTRACT This paper proposes a modified context‐layered Elman recurrent neural network (MCL‐ERNN) integrated with a Lyapunov‐based adaptive weight‐update law and an adaptive learning‐rate (ALR) strategy for nonlinear dynamical‐system control. The proposed architecture modifies the conventional Elman recurrent neural network by applying nonlinear processing to delayed hidden‐state information and incorporating an additional delayed output‐feedback pathway, thereby enhancing temporal information representation while retaining a compact recurrent structure. A Lyapunov‐based learning law is formulated for online parameter adaptation, while a bounded ALR mechanism regulates the magnitude of parameter updates according to the evolution of the learning error. The theoretical analysis establishes a local first‐order Lyapunov descent condition under the prescribed learning‐rate bounds and sufficiently small parameter variations; it does not constitute an unconditional guarantee of global closed‐loop stability. The proposed framework is evaluated on three nonlinear dynamical systems and compared with conventional Elman and Jordan recurrent neural network controllers through independent simulation studies. The results show improved tracking accuracy and lower learning errors for the proposed MCL‐ERNN across the considered cases. Robustness investigations further show satisfactory tracking under the investigated external disturbances, measurement noise and plant‐parameter variations. An ablation study indicates that the modified context‐layer architecture, Lyapunov‐based learning mechanism, and ALR strategy provide complementary contributions to the observed performance improvements. A structural operation‐count comparison further characterizes the architectural requirements of the proposed MCL‐ERNN relative to the benchmark recurrent architectures. Overall, the results provide numerical evidence of the effectiveness of the proposed framework for the considered nonlinear control scenarios.

Authors

Institutions

Publication Details

Journal
International Journal of Adaptive Control and Signal Processing
Published
2026-11-17
DOI
https://doi.org/10.1002/acs.70172
Primary Topic
Adaptive Control of Nonlinear Systems
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Enhanced Adaptive Control of Nonlinear Systems Based on a Lyapunov‐Stability‐Based Modified Context‐Layered Elman Recurrent Neural Network With Adaptive Learning Rate

Rajesh Kumar
2 links
International Journal of Adaptive Control and Signal Processing
Adaptive Control of Nonlinear Systems
article

Enhanced Adaptive Control of Nonlinear Systems Based on a Lyapunov‐Stability‐Based Modified Context‐Layered Elman Recurrent Neural Network With Adaptive Learning Rate

Rajesh Kumar
article en

Abstract

ABSTRACT This paper proposes a modified context‐layered Elman recurrent neural network (MCL‐ERNN) integrated with a Lyapunov‐based adaptive weight‐update law and an adaptive learning‐rate (ALR) strategy for nonlinear dynamical‐system control. The proposed architecture modifies the conventional Elman recurrent neural network by applying nonlinear processing to delayed hidden‐state information and incorporating an additional delayed output‐feedback pathway, thereby enhancing temporal information representation while retaining a compact recurrent structure. A Lyapunov‐based learning law is formulated for online parameter adaptation, while a bounded ALR mechanism regulates the magnitude of parameter updates according to the evolution of the learning error. The theoretical analysis establishes a local first‐order Lyapunov descent condition under the prescribed learning‐rate bounds and sufficiently small parameter variations; it does not constitute an unconditional guarantee of global closed‐loop stability. The proposed framework is evaluated on three nonlinear dynamical systems and compared with conventional Elman and Jordan recurrent neural network controllers through independent simulation studies. The results show improved tracking accuracy and lower learning errors for the proposed MCL‐ERNN across the considered cases. Robustness investigations further show satisfactory tracking under the investigated external disturbances, measurement noise and plant‐parameter variations. An ablation study indicates that the modified context‐layer architecture, Lyapunov‐based learning mechanism, and ALR strategy provide complementary contributions to the observed performance improvements. A structural operation‐count comparison further characterizes the architectural requirements of the proposed MCL‐ERNN relative to the benchmark recurrent architectures. Overall, the results provide numerical evidence of the effectiveness of the proposed framework for the considered nonlinear control scenarios.

International Journal of Adaptive Control and Signal Processing
National Institute of Technology Kurukshetra (IN)
Adaptive Control of Nonlinear Systems
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon