While there is ample research on hardware design and reconfiguration control for modular self-reconfigurable satellites, relatively few reconfiguration planning algorithms, especially algorithms used in real-world reconfiguration have been developed. Decentralized path planning, which only uses partial observation for each module to make decision is an important problem for real-world task. This article presents partially observable self-reconfiguration path planning, addressing the reconfiguration path planning problem for a single module using partial observations while aiming to maximize the policy learning efficiency. An end-to-end algorithm is proposed by employing a recurrent Q-learning algorithm and a deep neural network, where a Long Short Term Memory network is used to remember useful features from historical observations. Moreover, a 3-D convolutional neural network is used to automatically extract high-level features from observation data and is shown to significantly increase the learning efficiency. Experiments performed on a test rig of electromagnetic self-reconfigurable satellite verified the potency of the proposed algorithm.

Modular self-reconfigurable satellites can adapt to diverse space missions by dynamically rearranging their configurations. Enabling multiple modules to move simultaneously enhances reconfiguration efficiency but remains constrained by existing planning algorithms. This article introduces the Parallel Reconfiguration Planner (PRP), a learning-based approach that ensures safe, concurrent module movements. PRP leverages multi-agent reinforcement learning to generate reconfiguration policies while incorporating safety constraints modeled using Linear Temporal Logic (LTL). To improve learning efficiency, PRP integrates a guarded exploration mechanism based on these safety specifications. Experiments on 3D reconfiguration scenarios demonstrate that PRP achieves more efficient reconfiguration policies in scenarios that have more moving modules compared to the baseline method.

To deal with the challenging problem of unmanned surface vehicle (USV) path following, a method based on model predictive control (MPC) is designed in this paper. An USV is a typical nonholonomic constrained and under-actuated system, and it has huge inertia while there is no effective brake system to control its drift. In order to follow a given path better, prediction based method should be used to take the future states into account. The kinetic model of USV has been established to simulate the prediction horizon while the inertia and hydrodynamic forces are considered. Numerical simulations are carried out and demonstrate that the controller can command the USV to follow a series of waypoints smoothly and accurately.