Xin Chen

Safety-critical networked control requires feasible commands when distributed optimization stops early. We develop a persistent-allocation framework for separable convex programs with affine coupling constraints. Agents directly store zero-sum physical allocations and enforce hard local certificates, preserving the certified state across communication-topology changes within a frozen program. Starting from a certified seed for the current program, atomic bilateral commits preserve feasibility after every accepted update. We establish separate frozen-program convergence results for a full-network proximal reference and the pairwise implementation under their respective regularity and activation assumptions. A value-function counterexample illustrates nonoptimal pairwise stalling outside the regularity conditions and relates it to an established hard-boundary obstruction. For single-integrator control, a zero-order-hold condition guarantees inter-sample safety along convex barrier trajectories. Controlled simulations show how persistent allocations retain optimization progress across topology changes and how edgewise publication exposes feasible improvements under interruption, while identifying a tradeoff between numerical payload and sequential stages. Multi-robot simulations relate the edge-update budget to control cost and verify safety throughout each control-holding interval.

1. In a wide range of practical applications – such as hypersonic aircraft (Xu et al., 2015), spacecraft guidance and control (Lu & Xia, 2013), motion tracking systems (Mannava et al., 2012), and m...

With declining hydrocarbon reserves, optimizing directional wellbore trajectory planning is crucial for accessing complex reservoirs while managing cost and risk. Multi-segment composite trajectories have been widely used in complex formations, but their high-dimensional nature challenges the efficiency of gradient-free multi-objective optimization methods. This paper proposes a conjugate gradient-assisted multi-objective algorithm to accelerate the optimization process. A dual-population update strategy integrates gradient-based search with constraint satisfaction, while adaptive step size and hybrid search mitigate multi-objective gradient conflicts. Gradients for formation-related objectives are approximated via finite differences. The algorithm is validated on a five-segment Bézier trajectory for an extended-reach horizontal wellbore, demonstrating superior normalized hypervolume, convergence, and diversity compared to gradient-based and evolutionary algorithms. Ablation studies verify the effectiveness of the hybrid strategy. The selected trajectory achieves an optimal balance among wellbore length, curvature, and stability under all constraints. To our knowledge, this is the first application of gradient-based multi-objective optimization to wellbore trajectory planning, offering a promising pathway toward real-time drilling trajectory design with enhanced efficiency and robustness.

Concerning the problem of inaccurate target localization and trajectory tracking in robot grasping. This paper presents a multi-time-scale model predictive control method for robot grasping based on visual servoing. First, the depth camera captures the depth information and 2-D images of the target and fuses them to generate a point cloud. Then, the mathematical model of position with depth variation is established by using the characterization of the point cloud reflecting the target position information. Next, considering the multi-time-scale characterization of the trajectory, it is decomposed into several subsequences. The support vector regression (SVR) prediction model is also established to predict future subsequences. Tracking of the target trajectory is realized by applying control to the robot. Finally, some experimental results based on actual industry data show that the method controls the robot to perform grasping tasks accurately in the complex industry field.

This paper addresses the optimal consensus problem in uncertain switching multiagent systems. The inherent uncertainty and time-varying structure of local tracking error system render conventional methods ineffective for deriving optimal control protocols. To overcome these challenges, we introduce a reference model for each agent and construct a modified augmented local tracking error (ALTE) system. This approach transforms the optimal consensus problem into two sub-problems: 1) model reference control (MRC) between agents and their reference models; 2) distributed optimal stabilization of the modified ALTE system. We propose a new control scheme that combines filtered tracking error with equivalent input disturbance method to achieve MRC. To realize distributed optimal stabilization of the modified ALTE, we introduce a deep deterministic policy gradient method based on value iteration. Through theoretical analysis, we demonstrate that the multiagent system achieves a near Nash equilibrium, which is further validated by numerical simulation.

In a two-crane cooperative lift process, unreasonable load distribution on the two cranes may cause one of the cranes to overload, which may cause a dangerous overturn accident. Therefore, the load distribution should be taken as a constraint to yield a safe path for a cooperative lift. Besides, the load distribution on the two cranes varies with the changing postures of the cranes. However, the explicit relationship between the load distribution and the postures has not been reported. Therefore, this article first presents a relationship model between the postures of the two cranes and the load distribution on them. Next, a new path-planning method based on reinforcement learning is explained, which utilizes the load constraint as the optimization object in the cooperative two-crane lift. Simulation results show that the new method yields a short lift path with reasonable load distribution.