吳維文 Wei-Wen Wu
國立中正大學 電機工程研究所 碩士生M.S. Student, Dept. of Electrical Engineering, National Chung Cheng University
結合 AI 與全端開發的軟體工程師。研究上，以 AI 自動設計毫米波貼片天線，成果已獲 URSI GASS 2026 接收；工程上，以 Python 為主力、輔以 C/Rust 提升效能，並用 ROS 串接機器人與 AI。A software engineer who combines AI with full-stack development. On the research side, I use AI to automatically design millimeter-wave patch antennas — work accepted to URSI GASS 2026. On the engineering side, I build mainly in Python, reach for C/Rust where performance matters, and use ROS to bridge robotics and AI.
重點發表Featured Publication
A Data-Free Patch Antenna Generation Framework via Two-Stage Gradient Exploration and BiScaleNorm
URSI General Assembly and Scientific Symposium (URSI GASS 2026) · Kraków, Poland · 2026/08
針對電磁模擬器不可微分的特性，提出以類神經網路代理模型估計梯度，解決二值化設計的梯度飽和、局部解與孤島效應，優化貼片天線金屬圖樣。Addresses the non-differentiability of EM simulators with a neural surrogate that estimates gradients — mitigating gradient saturation, local optima, and island effects in binarized patch-antenna design.
三項主要能力Three Core Strengths
以 AI 技術設計貼片天線;熟悉電腦視覺、強化學習與深度學習。AI-driven antenna design; computer vision, reinforcement & deep learning.
ROS 六軸機械手臂控制與模擬，整合視覺抓取。ROS-based 6-axis arm control & simulation with vision-guided grasping.
客製化 ERP、物品管理、專案排程系統，獨立走完開發生命週期。Custom ERP, inventory & project-scheduling systems, owned end-to-end.