Leader-following output-feedback consensus for second order multiagent systems with arbitrary convergence time and...

Area of Science:

• Control Theory
• Robotics
• Networked Systems

Background:

• Multiagent systems often face challenges in achieving coordinated behavior, especially under constraints like limited sensor data and predefined timeframes.
• Existing consensus algorithms may struggle with precise convergence times or handling real-world issues such as measurement noise and actuator saturation.

Purpose of the Study:

• To develop a novel prescribed-time output-feedback consensus strategy for second-order multiagent systems lacking velocity measurements.
• To enable offline pre-assignment of transient performance metrics like convergence rate and overshoot.
• To address and resolve singularity issues arising from measurement errors and actuator saturation.

Main Methods:

• Design of prescribed-time state observers using a time-scaling function to estimate agent states without velocity information.
• Proposal of a distributed output-feedback control scheme for leader-following consensus.
• Implementation of piecewise functions to mitigate singularity problems caused by measurement errors.
• Modification of the control strategy with an auxiliary system to manage actuator saturation.

Main Results:

• Successful estimation of second-order states via novel prescribed-time observers.
• Achievement of leader-following consensus with pre-assignable transient performance.
• Effective resolution of singularity issues and saturation problems in the control strategy.
• Validation of the proposed schemes through numerical simulations.

Conclusions:

• The developed output-feedback control strategy effectively achieves prescribed-time leader-following consensus in second-order multiagent systems.
• The method robustly handles systems with missing velocity measurements, measurement errors, and actuator saturation.
• The ability to pre-assign convergence rates and overshoots offers significant advantages for practical applications requiring predictable system dynamics.