This article presents an adaptive fault-tolerant tracking control strategy for unknown affine nonlinear systems subject to actuator faults and external disturbances. To address the hyperparameter initialization challenges inherent in conventional neural network training, an improved self-organizing radial basis function neural network (SRBFNN) with an input-dependent variable structure is developed. Furthermore, a novel self-organizing RBFNN-based observer is introduced to estimate system states across all dimensions. Leveraging the reconstructed states, the proposed adaptive controller effectively compensates for all uncertainties, including estimation errors in the observer, ensuring accurate state tracking with reduced control effort. The uniform ultimate boundedness of all closed-loop signals and tracking errors is rigorously established via Lyapunov stability analysis. Finally, simulations on two different nonlinear systems comprehensively validate the effectiveness and superiority of the proposed control approach.

This article investigates the practical problem of the fault-tolerant tracking control for nonlinear systems subjected to dynamic uncertainties, unknown disturbances, and simultaneous actuator and sensor faults. A novel performance-enhanced intelligent fault-tolerant control (IFTC) framework is developed by integrating the predefined-time stability theory with the optimal control principle. A 1-dimensional convolutional neural network is proposed as a restorer to efficiently detect and isolate sensor faults, upon whose outputs a single radial basis function neural network is constructed to approximate two distinct components of the unknown system dynamics in real time. The proposed fault-tolerant controller ensures the boundedness of all signals within the closed-loop system and guarantees that the tracking errors converge within a user-specified time. Moreover, by incorporating a cost function related to the control effort, the developed controller avoids the excessive control input, achieving an enhanced control performance alongside the predefined-time convergence. Comprehensive simulation results validate the effectiveness and practicality of the proposed performance-enhanced IFTC scheme, highlighting its potential for real-world applications.

Achieving rapid and accurate object classification holds significant importance in various domains. However, conventional vision-based techniques suffer from several limitations, including high data redundancy and strong dependence on image quality. In this work, we present a high-speed, image-free object classification method based on dual-pixel measurement and normalized central moment invariants. Leveraging the complementary modulation capability of a digital micromirror device (DMD), the proposed system requires only five tailored binary illumination patterns to simultaneously extract geometric features and perform classification. The system can achieve a classification update rate of up to 4.44 kHz, offering significant improvements in both efficiency and accuracy compared to traditional image-based approaches. Numerical simulations verify the robustness of the method under similarity transformations—including translation, scaling, and rotation—while experimental validations further demonstrate reliable performance across diverse object types. This approach enables real-time, low-data throughput, and reconstruction-free classification, offering new potential for optical computing and edge intelligence applications.