Neural-Network-Based Adaptive Fixed-Time Control for a 2-DOF Helicopter System With Input Quantization and Output...

Area of Science:

• Robotics and Control Systems
• Nonlinear System Dynamics
• Artificial Intelligence in Engineering

Background:

• Nonlinear helicopter systems present significant control challenges due to inherent complexities.
• Input quantization and output constraints can degrade control performance and system stability.
• Existing control methods may struggle to guarantee fixed-time convergence under such conditions.

Purpose of the Study:

• To develop a novel adaptive fixed-time control strategy for a 2-DOF nonlinear helicopter.
• To address challenges posed by input quantization and output constraints.
• To ensure system signals remain bounded within a predetermined fixed time.

Main Methods:

• A hysteresis quantizer and adaptive variables were used to manage input quantization and associated errors.
• Radial basis function neural networks (NNs) approximated system uncertainties.
• A logarithmic barrier Lyapunov function (BLF) enforced output constraints.
• Lyapunov stability analysis and fixed-time stability criteria were employed.

Main Results:

• The proposed NN-based adaptive control method effectively mitigated chattering caused by quantization.
• System uncertainties were accurately approximated by the radial basis function NN.
• The logarithmic barrier Lyapunov function successfully prevented output constraint violations.
• Closed-loop system signals were rigorously proven to be bounded within a fixed time.

Conclusions:

• The developed control strategy ensures robust and stable operation of the 2-DOF nonlinear helicopter system.
• The method demonstrates effectiveness in handling input quantization and output constraints.
• Numerical simulations and experimental validation confirm the practical feasibility of the proposed approach.