Liu Xu刘旭

Research work研究工作

Online 3DGS and a past photo condition an eight-channel Restoration Field for object and anchor readout. Nav2 and Contact-GraspNet connect this spatial information to search, grasping and placement.在线 3DGS 与旧照片共同生成八通道恢复场，提供物体和原始锚点的空间信息；通过 Nav2 与 Contact-GraspNet 接入搜索、抓取和放置。

The original location is empty, so object appearance alone cannot identify it. Removing anchor affinity reduces simulated ordered localization from 34.8% to 8.4%.原位置已空置，物体外观不足以定位它；移除锚点关联后，仿真顺序定位成功率从 34.8% 降至 8.4%。

Evaluation实验验证

Simulated ordered localization: 34.8%, without grasping or placement. Real restoration: 5/10, requiring stable placement on the correct support with horizontal error <0.25 m; orientation is not evaluated.仿真顺序定位 34.8%（不含抓放）；实机复原 5/10：正确支撑面上稳定放置，水平误差 <0.25 m，朝向不作评估。

Method & results方法与结果

Research work研究工作

Current point clouds guide docking, then the visual goal conditions WAM. Two refinement rounds precede arm actions; the first 24 actions are executed before observing and replanning.先根据当前点云选择停靠位置，再以视觉目标驱动 WAM；两轮预测细化后生成机械臂动作，执行前 24 步并重新观测规划。

A reachable pose may be poor for restoration, so frozen-WAM outcomes supervise docking utility. Removing utility reduces mean simulation success across three training seeds from 71.1% to 47.3%.可达位姿未必有利于恢复，故以冻结 WAM 的恢复结果监督停靠效用；移除效用后，三训练种子的仿真平均成功率由 71.1% 降至 47.3%。

Evaluation实验验证

Real success: 19/30 at ≤30 mm. Final position error: 30.7 ± 12.7 mm, mean ± sample SD across all 30 trials, including failures.实机 19/30（≤30 mm）；全部 30 次含失败的位置误差 30.7 ± 12.7 mm（均值 ± 样本标准差）。

Method & results方法与结果

Simulation to flight从仿真到实飞

I led planning and task-interface adaptation for the ten-UAV simulation, covering narrow openings, corners, takeoff sequencing and inter-agent path conflicts. For three real UAVs, I contributed to FAST-LIO localization debugging and onboard hardware integration.我负责十机仿真的规划与任务接口适配，处理窄门、转角、起飞顺序和多机路径冲突；在三机实飞阶段，参与 FAST-LIO 定位联调与机载硬件集成。

Tilting LiDAR improves ground-feature coverage but introduces a blind zone. The VR1 team material records the resulting need to retune planning parameters—a concrete interaction between sensor mounting and obstacle avoidance.激光雷达前倾有利于观察地面特征，也会引入盲区。VR1 团队材料记录了由此带来的规划参数重调需求：传感器安装方式会直接影响避障表现。

Team validation团队验证

Simulation acceptance: 10/10 UAVs, 166 s, fastest of eight teams. Real acceptance: 100% success (3/3) in each sequence; 50 / 60 / 72 s, second of eight teams by completion speed.仿真验收：10/10 架完成，166 s，八支队伍中完成速度第 1；实机验收：各项成功率均为 100%（3/3），50 / 60 / 72 s，完成速度第 2。

Livox Mid-360 · Jetson Orin NX · FAST-LIO · EGO-Planner

System & flight tests系统与实飞测试