Design and Multi-Objective Optimization of Efficient UAV Battery Thermal Management System Using PCM–Air Synergistic Cooling Strategy (Advisor: Hongda Du)

Project · 2025–present

Developed a PCM–air hybrid battery thermal management system (BTMS) for UAV lithium-ion batteries, achieving 43.3% lower peak temperature (below 46.8 °C) with only 22.2 wt% mass increase. Multi-factor optimization using an entropy–TOPSIS framework ensures a balance between cooling performance and lightweight design.

提出了一种用于无人机锂离子电池的相变材料–风冷混合热管理系统（BTMS），在总质量仅增加约 22 wt% 的情况下，使电池最高温度降低 43.3% 且保持低于 46.8 °C；通过 TOPSIS 优化实现轻量化与散热性能的平衡。

WM_encoder_decoder for Copyright Protection in Image-to-Image Tasks (Diffusion) · PPT (Advisor: Feng Zheng)

Project / Code

Dual-protection scheme combining digital watermarking and adversarial perturbations in the latent space; precise embed/extract via a pre-trained encoder–decoder; robust, transferable, and high-quality without model fine-tuning.

提出数字水印 + 对抗扰动的双重保护方案，在隐空间抑制篡改与生成操控；无需微调、具有良好迁移性与鲁棒性。

Privacy-Preserving Secure Shared Nearest Neighbor Clustering Scheme in Internet of Things (Advisor: Hanlin Zhang)

Research Project

Cloud-assisted, privacy-preserving SNN clustering using orthogonal matrix transformation and lightweight verification; cuts the most complex step from 10.859 s → 1.183 s while preserving accuracy.

云辅助的隐私保护 SNN 聚类，外包前以正交矩阵加密并可验证返回结果；在保持精度下显著降低本地计算开销。

Secure and Efficient Image Denoising Algorithm in Internet of Things (Advisor: Hanlin Zhang)

Degree Thesis

Secure outsourcing for image denoising using Householder transformation and optimal-probability verification; maintains accuracy (PSNR ≈ 20–21 dB) while greatly reducing client computation.

面向物联网图像去噪的安全外包方案，以豪斯霍尔德变换保护隐私并可验证结果；在 PSNR 约 20–21 dB 下显著降低端侧计算。