A Study on Robust Control Scheme Using Prescribed Performance-Based Time Delay Control and RBF Neural Network

This paper proposes a robust control scheme combining prescribed performance-based time-delay estimation (TDE) with a radial basis function neural network (RBF-NN). The proposed prescribed performance-based time delay control constructs desired error dynamics through transformed errors to achieve prescribed performance. This design preserves the simple control structure of traditional time-delay control (TDC) while maintaining its stability condition and $\bar {M}$ stability analysis, enabling the application of existing design methodologies. RBF-NN is incorporated to compensate residual TDE error, and Lyapunov analysis establishes stability of the overall system. The approach is validated in simulation on a 3-DOF robotic manipulator under cases without and with payload variations and additive joint disturbances, demonstrating satisfaction of the prescribed performance bounds and improved tracking relative to the TDC while retaining simplicity.