物理-数据-知识混合驱动的人机混合增强智能系统管控方法

doi:10.11959/j.issn.2096-6652.202237

智能科学与技术学报 ›› 2022, Vol. 4 ›› Issue (4): 571-583.doi: 10.11959/j.issn.2096-6652.202237

• 专栏：人工智能3.0中的机器学习方法 • 上一篇下一篇

物理-数据-知识混合驱动的人机混合增强智能系统管控方法

张俊, 许沛东, 陈思远, 高天露, 戴宇欣, 张科, 赵杭, 高杰迈, 白昱阳, 李金星, 张浩然, 李湘, 陈玖香

武汉大学电气与自动化学院，湖北武汉 430072

修回日期:2022-07-18 出版日期:2022-12-15 发布日期:2022-12-01
作者简介:张俊（1981− ），男，博士，武汉大学电气与自动化学院教授、博士生导师，中国自动化学会副秘书长，IEEE 射频识别理事会副主席。主要研究方向为大数据、人工智能、物联网等新一代信息技术及其在电力能源等复杂系统中的应用
许沛东（1995− ），男，武汉大学电气与自动化学院博士生，主要研究方向为人工智能、智能电网
陈思远（1993− ），男，武汉大学电气与自动化学院博士生，主要研究方向为电力系统运行与控制、人工智能和机器学习
高天露（1994− ），男，武汉大学电气与自动化学院博士生，主要研究方向为自然语言处理、区块链和人机混合增强智能在电力系统运行分析和控制决策中的应用
戴宇欣（1996− ），男，武汉大学电气与自动化学院博士生，主要研究方向为人工智能技术及其在电力系统中的应用
张科（1993− ），女，武汉大学电气与自动化学院博士生，主要研究方向为人工智能技术在电力系统中的应用、可信任的人工智能系统以及人机混合智能
赵杭（1995− ），男，武汉大学电气与自动化学院博士生，主要研究方向为因果推理在电力系统中的应用
高杰迈（1995− ），女，武汉大学电气与自动化学院博士生，主要研究方向为综合能源系统、智能电网和物理深度学习
白昱阳（1997− ），男，武汉大学电气与自动化学院硕士生，主要研究方向为深度强化学习在电力系统调度中的应用
李金星（1996− ），男，武汉大学电气与自动化学院硕士生，主要研究方向为人工智能技术在电网故障定位及故障处置中的应用
张浩然（1999− ），男，武汉大学电气与自动化学院硕士生，主要研究方向为人工智能在电力系统中的应用
李湘（2000− ），女，武汉大学电气与自动化学院硕士生，主要研究方向为人工智能在电力系统中的应用
陈玖香（2001− ），女，武汉大学电气与自动化学院硕士生，主要研究方向为人工智能在电力系统中的应用
基金资助:
科技创新2030—“新一代人工智能”重大项目(2021ZD0112700)

A hybrid physics-data-knowledge driven approach for human-machine hybrid-augmented intelligence-based system management and control

Jun ZHANG, Peidong XU, Siyuan CHEN, Tianlu GAO, Yuxin DAI, Ke ZHANG, Hang ZHAO, Jiemai GAO, Yuyang BAI, Jinxing LI, Haoran ZHANG, Xiang LI, Jiuxiang CHEN

School of Electrical Engineering and Automation, Wuhan University, Wuhan 430072, China

Revised:2022-07-18 Online:2022-12-15 Published:2022-12-01
Supported by:
The National Key Research and Development Program of China(2021ZD0112700)

摘要/Abstract

摘要：

当代系统认知、管理与控制的核心理论、方法与技术已经转移到大数据和人工智能技术上，这导致当前人工智能技术条件局限与复杂系统认知、管理、控制的需求之间形成了一道鸿沟。因此，现实的需求催生了人工智能的一种新型形态——人机混合增强智能形态，即人类智能与机器智能协同贯穿于系统认知、管理、控制等过程的始终，人类的认知和机器智能认知互相混合，形成增强型的智能形态，这种形态是人工智能或机器智能可行的、重要的成长模式。提出了一种物理-数据-知识混合驱动的人机混合增强智能系统管控方法。从可信分布式数据、计算和算法，物理深度学习，融合系统运行规则的混合型深度强化学习，因果分析，可解释性AI与数字人5个方面详细阐述了所提方法。最后，以电力系统调控为背景，以3个应用为例分析了所提方法的应用方式和技术路径。

关键词: 物理-数据-知识混合驱动方法, 人机混合增强智能, 系统管理与控制, 物理深度学习, 因果分析, 可解释AI, 虚拟数字人

Abstract:

The core theories, methods and technologies of contemporary system cognition, management, and control have been transferred to big data and artificial intelligence technology, resulting in a gap between the limitations of current artificial intelligence technology and the needs of complex system cognition, management, and control.As a result, a real need has spawned a new form of artificial intelligence: human-machine hybrid-augmented intelligence form, that is, the cooperation of human intelligence and machine intelligence runs through the process of system cognition, management, control, and so on.Human cognition and machine intelligence cognition are mixed together to form enhanced intelligence form.This form is a feasible and important growth mode of artificial intelligence or machine intelligence.A hybrid physics-data-knowledge (PDK) driven approach for human-machine hybrid-augmented intelligence-based system management and control was proposed.The proposed approach was illustrated by the following: trustworthy distributed data, computing, and algorithm, physics-informed deep learning, hybrid deep reinforce learning incorporating system operation rules, causal analysis, and interpretable AI and virtual digital human.In the context of power system dispatch and control, three examples were used for explaining the applications and technical pathways of the proposed PDK approach.

Key words: hybrid physics-data-knowledge driven approach, human-machine hybrid-augmented intelligence, system management and control, physics-informed deep learning, causal analysis, interpretable AI, virtual digital human

中图分类号:

TP319

张俊, 许沛东, 陈思远, 等. 物理-数据-知识混合驱动的人机混合增强智能系统管控方法[J]. 智能科学与技术学报, 2022, 4(4): 571-583.

Jun ZHANG, Peidong XU, Siyuan CHEN, et al. A hybrid physics-data-knowledge driven approach for human-machine hybrid-augmented intelligence-based system management and control[J]. Chinese Journal of Intelligent Science and Technology, 2022, 4(4): 571-583.

图/表 10

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

参考文献 38

[1]	SHLESINGER M F . Complex adaptive systems:an introduction to computational models of social life[J]. Journal of Statistical Physics, 2007,129(2): 409-410.
[2]	NIAZI M . Introduction to the modeling and analysis of complex systems:a review[J]. Complex Adaptive Systems Modeling, 2016,4: 1-3.
[3]	MANOJ B S , CHAKRABORTY A , SINGH R . Complex networks:a networking and signal processing perspective[M].[S.l.:s.n.], 2018.
[4]	THURNER S , KLIMEK P , HANEL R . Introduction to the theory of complex systems[M]. Oxford: Oxford University Press, 2018.
[5]	WANG F Y , CHEN J L . Intelligent control method and application[M].[S.l.]: China Science and Technology Press, 2020.
[6]	ZHENG N N , LIU Z Y , REN P J ,et al. Hybrid-augmented intelligence:collaboration and cognition[J]. Frontiers of Information Technology ＆ Electronic Engineering, 2017,18(2): 153-179.
[7]	白昱阳, 黄彦浩, 陈思远 ,等. 云边智能:电力系统运行控制的边缘计算方法及其应用现状与展望[J]. 自动化学报, 2020,46(3): 397-410.
	BAI Y Y , HUANG Y H , CHEN S Y ,et al. Cloud-edge intelligence:status quo and future prospective of edge computing approaches and applications in power system operation and control[J]. Acta Automatica Sinica, 2020,46(3): 397-410.
[8]	CHEN S Y , ZHANG J , BAI Y Y ,et al. Blockchain enabled intelligence of federated systems (BELIEFS):an attack-tolerant trustable distributed intelligence paradigm[J]. Energy Reports, 2021,7: 8900-8911.
[9]	HESTER T , STONE P . TEXPLORE:real-time sample-efficient reinforcement learning for robots[J]. Machine Learning, 2013,90(3): 385-429.
[10]	XIONG Y H , ZHENG G J , XU K ,et al. Learning traffic signal control from demonstrations[C]// Proceedings of the 28th ACM International Conference on Information and Knowledge Management. New York:ACM Press, 2019: 2289-2292.
[11]	ROSS S , GORDON G , BAGNELL D . A reduction of imitation learning and structured prediction to no-regret online learning[C]// Proceedings of the 14th International Conference on Artificial Intelligence and Statistics.[S.l.:s.n.], 2011: 627-635.
[12]	HESTER T , VECERIK M , PIETQUIN O ,et al. Deep Q-learning from demonstrations[J]. arXiv preprint, 2017,arXiv:1704.03732.
[13]	JING M X , MA X J , HUANG W B ,et al. Reinforcement learning from imperfect demonstrations under soft expert guidance[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2020,34(4): 5109-5116.
[14]	NAIR A , MCGREW B , ANDRYCHOWICZ M ,et al. Overcoming exploration in reinforcement learning with demonstrations[C]// Proceedings of 2018 IEEE International Conference on Robotics and Automation. Piscataway:IEEE Press, 2018: 6292-6299.
[15]	GAO Y , XU H Z , LIN J ,et al. Reinforcement learning from imperfect demonstrations[J]. arXiv preprint, 2018,arXiv:1802.05313.
[16]	RAMíREZ J , YU W , PERRUSQUíA A , . Model-free reinforcement learning from expert demonstrations:a survey[J]. Artificial Intelligence Review, 2022,55(4): 3213-3241.
[17]	KILINC O , MONTANA G . Reinforcement learning for robotic manipulation using simulated locomotion demonstrations[J]. Machine Learning, 2022,111(2): 465-486.
[18]	LI X S , WANG X , ZHENG X H ,et al. SADRL:merging human experience with machine intelligence via supervised assisted deep reinforcement learning[J]. Neurocomputing, 2022,467: 300-309.
[19]	LI X S , WANG X , ZHENG X H ,et al. Supervised assisted deep reinforcement learning for emergency voltage control of power systems[J]. Neurocomputing, 2022,475: 69-79.
[20]	XU P D , PEI Y Z , ZHENG X H ,et al. A simulation-constraint graph reinforcement learning method for line flow control[C]// Proceedings of 2020 IEEE 4th Conference on Energy Internet and Energy System Integration. Piscataway:IEEE Press, 2020: 319-324.
[21]	XU P D , DUAN J J , ZHANG J ,et al. Active power correction strategies based on deep reinforcement learning—part I:a simulation-driven solution for robustness[J]. CSEE Journal of Power and Energy Systems, 2022,8(4): 1122-1133.
[22]	PEARL J , MACKENZIE D . The book of why:the new science of cause and effect[M].[S.l.:s.n.], 2018.
[23]	HEINDORF S , SCHOLTEN Y , WACHSMUTH H ,et al. CauseNet:towards a causality graph extracted from the web[C]// Proceedings of the 29th ACM International Conference on Information ＆ Knowledge Management. New York:ACM Press, 2020: 3023-3030.
[24]	SPEER R , CHIN J , HAVASI C . ConceptNet 5.5:an open multilingual graph of general knowledge[C]// Proceedings of the 31st AAAI Conference on Artificial Intelligence.[S.l.:s.n.], 2017: 4444-4451.
[25]	JAIMINI U , SHETH A . CausalKG:causal knowledge graph explainability using interventional and counterfactual reasoning[J]. IEEE Internet Computing, 2022,26(1): 43-50.
[26]	PEARL J . Causality[M]. Cambridge: Cambridge University Press, 2009.
[27]	VIALE R . Causal cognition and causal realism[J]. International Studies in the Philosophy of Science, 1999,13(2): 151-167.
[28]	TENENBAUM J B , KEMP C , GRIFFITHS T L ,et al. How to grow a mind:statistics,structure,and abstraction[J]. Science, 2011,331(6022): 1279-1285.
[29]	MARCUS G , DAVIS E . Rebooting AI:building artificial intelligence we can trust[M].[S.l.:s.n.], 2019.
[30]	BAREINBOIM E , BRITO C , PEARL J . Local characterizations of causal Bayesian networks[M]// Graph structures for knowledge representation and reasoning. Heidelberg: Springer, 2012: 1-17.
[31]	SHPITSER I , SHERMAN E . Identification of personalized effects associated with causal pathways[J]. Uncertainty in Artificial Intelligence, 2018:198.
[32]	BLOMQVIST E , ALIREZAIE M , SANTINI M . Towards causal knowledge graphs - position paper[C]// Proceedings of the Knowledge Discovery in Healthcare Data.[S.l.:s.n.], 2020.
[33]	ZHANG K , XU P D , ZHANG J . Explainable AI in deep reinforcement learning models:a SHAP method applied in power system emergency control[C]// Proceedings of 2020 IEEE 4th Conference on Energy Internet and Energy System Integration. Piscataway:IEEE Press, 2020: 711-716.
[34]	ZHANG K , XU P D , GAO T L ,et al. A trustworthy framework of artificial intelligence for power grid dispatching systems[C]// Proceedings of 2021 IEEE 1st International Conference on Digital Twins and Parallel Intelligence. Piscataway:IEEE Press, 2021: 418-421.
[35]	ZHANG K , ZHANG J , XU P D ,et al. Explainable AI in deep reinforcement learning models for power system emergency control[J]. IEEE Transactions on Computational Social Systems, 2022,9(2): 419-427.
[36]	DAI Y X , CHEN Q M , ZHANG J ,et al. Enhanced oblique decision tree enabled policy extraction for deep reinforcement learning in power system emergency control[J]. Electric Power Systems Research, 2022,209:107932.
[37]	李金星, 李湘, 高天露 ,等. 基于电网多元信息知识图谱的故障处置研究及应用[J]. 电力信息与通信技术, 2021,19(11): 30-38.
	LI J X , LI X , GAO T L ,et al. Research and application of fault handling based on power grid multivariate information knowledge graph[J]. Electric Power Information and Communication Technology, 2021,19(11): 30-38.
[38]	戴宇欣, 陈琪美, 高天露 ,等. 基于加权倾斜决策树的电力系统深度强化学习控制策略提取[J]. 电力信息与通信技术, 2021,19(11): 17-23.
	DAI Y X , CHEN Q M , GAO T L ,et al. Deep reinforcement learning control policy extraction based on weighted oblique decision tree[J]. Electric Power Information and Communication Technology, 2021,19(11): 17-23.

物理-数据-知识混合驱动的人机混合增强智能系统管控方法

A hybrid physics-data-knowledge driven approach for human-machine hybrid-augmented intelligence-based system management and control

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 38

相关文章 3

Metrics

推荐阅读 0

[1]	芦玉琦, 陈德旺, 赵钊林. 基于AlphaZero的地铁列车大量速度曲线自动生成算法[J]. 智能科学与技术学报, 2021, 3(2): 179-184.
[2]	吴宇震, 张俊, 高天露, 孙玉健, 刘金旭. 平行港口：智慧绿色时代下港口工业智联网新形态与体系结构[J]. 智能科学与技术学报, 2021, 3(2): 218-227.
[3]	陈德旺,蔡际杰,黄允浒. 面向可解释性人工智能与大数据的模糊系统发展展望[J]. 智能科学与技术学报, 2019, 1(4): 327-334.