A modified HP-adaptive pseudospectral method for multi-UAV formation reconfiguration.

Area of Science:

• Robotics
• Control Systems
• Optimization

Background:

• Multi-unmanned aerial vehicle (UAV) formation reconfiguration presents significant computational challenges due to complex state constraints.
• These constraints typically result in high-dimensional nonlinear programming (NLP) problems, increasing computational burden.

Purpose of the Study:

• To develop a novel computational method for efficient multi-UAV formation reconfiguration.
• To reduce the computational overhead associated with solving state-constrained optimization problems in UAV control.

Main Methods:

• Modification of the hp-adaptive pseudospectral method (PM) combined with an exact penalty function scheme.
• Transformation of the UAV reconfiguration control problem into an optimization problem.
• Utilization of the exact penalty function to eliminate inter-UAV distance state constraints, converting the problem into a standard NLP.

Main Results:

• The proposed method dramatically decreases computational overhead compared to the standard hp-adaptive PM.
• The modified method effectively handles state constraints in UAV formation reconfiguration.
• Numerical simulations demonstrate the scheme's effectiveness.

Conclusions:

• The novel computational method offers an efficient solution for multi-UAV formation reconfiguration.
• This approach simplifies the handling of state constraints, making complex UAV control problems more tractable.
• The technique shows promise for practical applications in autonomous systems.