Communication-Efficient Neural Adaptive Sliding-Mode Coordination of Dual Manipulators for Aerospace-Payload Handling

Cooperative payload handling in aerospace applications requires multiple manipulators to maintain coordinated motion under nonlinear coupled dynamics, uncertainties, and communication constraints. However, accurate joint regulation alone cannot guarantee inter-manipulator synchronization and relative payload geometry, while event-triggered control may lose its communication efficiency if continuously measured states remain implicitly available to the controller. This paper proposes a communication-efficient neural adaptive sliding-mode coordination framework for dual-manipulator payload handling. Joint regulation, synchronization, and relative payload geometry are integrated into a unified coordination potential, whose gradient generates the reference dynamics and supports coordination-aware event triggering. An RBF neural network is employed to approximate the coupled nonlinear dynamics, together with adaptive robust compensation for residual uncertainties. A normalized sample-and-hold architecture ensures that all controller-side state-dependent quantities are reconstructed solely from the latest transmitted state. Lyapunov analysis establishes uniform ultimate boundedness and derives an explicit admissible bound on the event threshold, revealing the trade-off between communication reduction and coordination accuracy. Zeno behavior is excluded by proving a positive lower bound on the inter-event interval. The proposed framework is evaluated on a fixed-base dual-PUMA560 benchmark for aerospace-payload handling.