融合人工智能实验室（Hybrid Artificial Intelligence Lab ，HAI LAB）主要从事人工智能和数据科学的理论和应用研究，核心目标是实现人工智能在基础行业中的落地应用，研究方向是计算机科学、数学、通信科学等领域的交叉融合。具体包括：
1、无线智能。将机器学习应用到无线通信和网络，处理传统无线通信和网络中存在的非理想、高复杂度难题；
2、机器学习理论。关注知识驱动的机器学习理论和联邦学习理论，研究机器学习模型和算法的轻量级实现；
3、集成电路智能制造。基于数据科学理论和人工智能技术，研究大规模集成电路工程的智能制造，提高良率。
期刊文章
A Model-Data Dual-Driven Resource Allocation Scheme for IREE Oriented 6G Networks
Tao Yu, Simin Wang, Shunqing Zhang, Xiaojing Chen, Zi Xu, Xin Wang, Jiandong Li, Junyu Liu, Sihai Zhang, IEEE Internet of Things Journal, DOI: 10.1109/JIOT.2025.3577760
Physics-Informed Machine Learning based Edge Detection for SEM Images
Yi Fang, Chun Wang, Sihai Zhang. in IEEE Transactions on Semiconductor Manufacturing, DOI: 10.1109/TSM.2025.3576269
WAE: An evaluation metric for attribution-based XAI on time series forecasting
Yueshan Chen, Sihai Zhang, Neurocomputing, 2025, 622: 129379.
Why Contour Averaging Works for SEM Metrology: Analysis and Validation
Jingxian Wei, Chenyu Xu, Sihai Zhang, IEEE Transactions on Semiconductor Manufacturing, DOI: 10.1109/TSM.2024.3471635.
The Quantitative evaluation of carbon reduction enabled by mobile communication in other industries: indicators, methods, and cases
Sihai Zhang, Chan Cui, Hongyu Sun, Yongmei Zeng, Jiang Zhu, Jinkang Zhu, SCIENTIA SINICA Informationis, 2024, 54(4): 777–787. DOI: 10.1360/SSI-2023-0295
Optimal Two-Timescale Configuration of Mobile Edge Computing with Mixed Energy Supply
Xiaojing Chen, Si Chen, Wei Ni, Xin Wang, Sihai Zhang, Shunqing Zhang, Yanzan Sun, Shugong Xu, Abbas Jamalipour, IEEE Transactions on Smart Grid, doi: 10.1109/TSG.2024.3390772.
The Predictability of Stock Price: Empirical Study on Tick Data in Chinese Stock Market
Yueshan Chen, Xingyu Xu, Tian Lan, Sihai Zhang, Big Data Research, 2024, 35: 100414.
DGL: Device Generic Latency Model for Neural Architecture Search on Mobile Devices
Qinsi Wang, Sihai Zhang, IEEE Transactions on Mobile Computing, 2023, 23(2): 1954-1967.
FedSeq: A Hybrid Federated Learning Framework Based on Sequential In-Cluster Training
Zhikun Chen, Daofeng Li, Rui Ni, Jinkang Zhu, Sihai Zhang, IEEE Systems Journal, 2023, 17(3): 4038-4049.
Wireless knowledge driven channel estimation in massive MIMO
Sihai Zhang, Jiashu Lin, Yamei Xu, Ming Zhao, Jinkang Zhu, SCIENTIA SINICA Informationis, Volume 53, Issue 4, 2023, Pages 758-, ISSN 1674-7267.
会议文章
Knowledge-enhanced N2N: Learning SEM Images Denoising with No Clean Data
Yi Fang, Chun Wang, Tian Lan and Sihai Zhang, 2025 International Joint Conference on Neural Networks, Rome, Italy, 2025.
LoS Identifcation Based Transmit Power Minimization in Cell-free ISAC System
Jingwen Wu, Chao Hu, Sihai Zhang , Ling Qiu, 2025 IEEE 101st Vehicular Technology Conference (VTC2025-Spring), Oslo, Norway, 2025.
Two-timescale Transmit Power Minimization in Semi-Passive IRS-Aided ISAC Systems
Jingwen Wu, Qilue Huang, Sihai Zhang and Ling Qiu, 2024 the 10th International Conference on Computer and Communications (ICCC), Chengdu, China, 2024.
A User-driven Sleep and wake-up Technology for Energy-Efficient 5G-Advanced Base Stations
Xiangzhi Guo, Xiaohui Chen, Weidong Wang, Sihai Zhang, 2024 16th International Conference on Wireless Communications and Signal Processing, Hefei, China, 2024.
Performance Analysis of Machine Learning on Homomorphically Encrypted Data
Morogo Freshina Mahatho, Yueshan Chen, Sihai Zhang, 2024 16th International Conference on Wireless Communications and Signal Processing, Hefei, China, 2024.
Multi-Point Channel Charting Aided Resource Management in Cell-Free Networks
Yuanqing Ye, Chenyu Xu, Zhen Wang, Lie Shan, Sihai Zhang, International Conference on Future Communications and Networks, Valletta, Malta, November, 2024.
Feature Fusion based HotspotDetection with R-EfficientNet
Chun Wang, Yi Fang, and Sihai Zhang, InProceedings of Great Lakes Symposiumon VLSI (GLSVLSI ’24).
MathNAS: If Blocks Have a Role in Mathematical Architecture Design
Qinsi Wang, Jinghan Ke, Zhi Liang and Sihai Zhang, NeurIPS, 2023
Energy Consumption Constrained Resource Cell Optimization Based on Multipoint Channel Charting
Yuanqing Ye, Zhikun Chen, Ming Zhao, and Sihai Zhang, 2023 15th International Conference on Wireless Communications and Signal Processing, Hangzhou, China,2023.
Data Collaborative Federated Learning for Non-I.i.d Data in Wireless Networks
Jiaqi Pan, Zhikun Chen, Sihai Zhang, and Ming Zhao, International Conference on Future Communications and Networks, Queenstown, New Zealand, December, 2023.
Adaptive User Selection and Bandwidth Allocation for Fast Convergence of Federated Learning in Wireless Networks
Jiaqi Pan, Zhikun Chen, Ming Zhao, Sihai Zhang, and Jinkang Zhu, International Conference on Future Communications and Networks, Queenstown, New Zealand, December, 2023.
完整论文汇总：
论文汇总.pdf
詹望,方毅,张四海.一种基于噪声模型的扫描电子显微镜图像去噪方法:中国,CN202510219767.1[P]. 2025-2-26.
张四海,方毅,王纯.一种基于物理信息指导的SEM图像边缘检测方法:中国,CN202510338237.9[P]. 2025-3-21.
张四海,孙宏宇. 一种基于休眠控制的离网光伏供电基站节能方法及装置:中国,CN202510242414.3[P]. 2025-5-30.
赵伟,卢志强,张四海,何海辉,吴旭,罗昭锋,胡兵,一种实验设备预约管理系统,201610331323.8.
张四海,周武旸,基于大规模实时轨迹分析的出租车拒载行为检测方法,201610032750.6.
赵明,贾玉琳,卫海超,张四海,周武旸,异构网中基于干扰协调的最大最小公平性资源分配方法,201511025987.3.
张四海,钱岑,秦晓卫,周武旸,基于波束业务量的卫星波束覆盖动态调整方法,201310626926.7.
张四海,钱岑,秦晓卫,周武旸,基于移动社交网络节点社会特性的机会路由行为建模方法,201310626892.1
方彬,范娟,张四海,周武旸,李磊,一种蜂窝网和无线局域网融合的接入控制方法,201310134084.3.
方彬,范娟,张四海,周武旸,李磊,一种分布式的垂直切换和资源分配方法,201310133836.4.
杨非,赵明,张四海,周武旸,一种多信道盲已知干扰消除方法,201310010968.8.
刘畅,秦晓卫,张四海,周武旸,一种正交频分多址接入中继系统的用户公平资源分配方法,201010266609.5.
王煦法 ,曹先彬,罗文坚,马建辉,张四海,一种分层协同的网络病毒和恶意代码识别方法,200310106551.8.
无线大数据-运营商转型启示
殷一平，张四海，王振 著; 人民邮电出版社; 第1版(2018年10月).
托马斯•达文波特(作者), 珍妮•哈里斯 (作者), 罗伯特•莫里森(作者), 杨琪 (译者), 张四海 (译者)，浙江人民出版社; 第1版 (2018年3月1日).
张四海等; 未来移动通信论坛; 第1版(2017年11月).
计算机程序设计(C语言版)
贾伯琪，顾为兵，苏仕华，张四海，何克东编著; 机械工业出版社; 第1版(2011年7月).
工作中的数据分析
无线大数据与智慧5G白皮书
AI驱动的6G无线智能空口传输技术
项目来源：科技部 项目时间： 2022 . 10 - 2025 . 10
针对6G复杂化的网络环境，研究基于网络内生智能的高效有序化无线接入体系。
通感算融合的原生绿色通信组网理论、架构与方法
项目来源：科技部 项目时间： 2022 . 10 - 2026 . 10
揭示组网架构与方法的耗能演变规律，研究原生绿色网络架构与智简资源调度技术，实现能源-通感算资源-个性化服务的低能耗智能匹配。
未来移动通信系统中低能耗个性化联邦学习基础理论研究
项目来源：安徽省自然科学基金委 项目时间： 2022 . 01 - 2024 . 12
设计低能耗联邦学习方案，降低系统通信功耗；设计个性化联邦学习方案，缓解用户数据异构性问题。
面向人机物融合的新型联邦学习理论研究
项目来源：中国科大 项目时间： 2022 . 01 - 2024 . 12
研究基于动态一致平均的无中心联邦学习理论，研究低网络开销的用户分组联邦学习理论。
OPC智能优化
项目来源：长鑫存储 项目时间： 2022 . 06 - 2024. 06
利用机器学习方法，对集成电路研发与生产过程中产生的SEM图像进行分析，获取轮廓、语义等关键工艺信息，提高生产良率。
华为无线使能
项目来源：华为技术 项目时间： 2022 . 07 - 2023. 07
基于各行业节能减排能力提出使能系数，完成验证和分析；基于大数据统计分析方法完成若干关键行业的使能系数评估和保护各方数据隐私的分布式减碳总量评估。