Spacecraft on-orbit inspection provides a prerequisite for the subsequent autonomous proximity operations. To fulfill the mission, robust constrained control algorithms for relative position tracking and attitude adjustment should be employed to handle the nonlinear coupled dynamics, system constraints, and external disturbances. In this paper, a robust nonlinear model predictive control (NMPC) scheme with prescribed performance based on a fixed-time neural network disturbance observer is proposed. Predefined tracking performance requirements are achieved, and, meanwhile, safety and stability are guaranteed. By exploiting the prescribed performance control technique, the proposed NMPC structure is capable of establishing a quantitative relationship between design parameters and certain prespecified performance values. Moreover, the proposed fixed-time neural network disturbance observer shows superiority in the disturbance estimations due to the property of universal approximation and fixed-time convergence. To illustrate the capabilities of our algorithm, comparison simulation tests concerning with prescribed performance and disturbance estimating are presented. Selection of tuning parameters is discussed, and computational load of the proposed algorithm is profiled.

航天器在轨检查为后续自主接近操作提供了先决条件。为了完成任务，应该采用相对位置跟踪和姿态调整的鲁棒约束控制算法来处理非线性耦合动力学、系统约束和外部干扰。在本文中，提出了一种基于固定时间神经网络扰动观测器的具有规定性能的鲁棒非线性模型预测控制（NMPC）方案。达到预定义的跟踪性能要求，同时保证安全性和稳定性。通过利用规定的性能控制技术，所提出的 NMPC 结构能够在设计参数和某些预先指定的性能值之间建立定量关系。而且，所提出的固定时间神经网络扰动观测器由于具有普遍逼近和固定时间收敛的特性，在扰动估计方面显示出优越性。为了说明我们算法的能力，提出了与规定的性能和干扰估计有关的比较模拟测试。讨论了调整参数的选择，并分析了所提出算法的计算负载。