Self-Learning Control to Attitude Stabilization with Integral Event-Triggered Mechanism

Conventional event-triggered attitude controllers effectively reduce communication frequency but usually employ a zero-order-hold strategy between consecutive triggering instants, resulting in abrupt control torque variations that may degrade actuator performance. To address this issue, this paper proposes a self-learning-based integral event-triggered control strategy for spacecraft attitude stabilization. The proposed method uses stored historical commands and elapsed inter-event time to construct a bounded time-varying compensation signal without increasing the communication or controller-update rate. Lyapunov-based analysis establishes uniform ultimate boundedness under bounded disturbances, actuator faults, and inertia uncertainties, while Zeno behavior is excluded. In the reported comparison, the proposed self-learning integral event-triggered controller achieves lower mean attitude and angular-velocity errors together with reduced error variances, while increasing the average inter-event interval by approximately 43% and reducing the number of triggering events by approximately 30%.