基于异质数据的复杂网络数据驱动控制  

Data-driven Control of Complex Networks Based on Heterogeneous Data

作　　者：刘斌[1] 马宇杰 LIU Bin;MA Yujie(School of Electrical Information Engineering,Northeast Petroleum University,Daqing 163319)

机构地区：[1]东北石油大学电气信息工程学院,大庆163319

出　　处：《计算机与数字工程》2024年第9期2710-2715,共6页Computer & Digital Engineering

基　　金：黑龙江省自然科学基金优秀青年项目(编号:YQ2019D001);黑龙江省省属本科高校基本科研业务费项目(编号:SJQHB201901);中国石油科技创新基金项目(编号:2021DQ02-1103)资助。

摘　　要：针对复杂网络难以控制,模型难以构建的问题,论文提出基于异质数据的复杂网络数据驱动控制策略。通过对不同时间、任意初始状态、不同最终状态输入输出和状态的异质数据进行重构,建立了基于异质数据集的采样轨迹表示方法,并依据该方法推导出适用于多种复杂网络模型的最小控制能量输入的两种计算方法,对计算方法需要的最少数据量进行了讨论,实现多种复杂网络的数据驱动控制,为复杂网络控制研究提供了新的方法。Aiming at the problems that complex networks are difficult to control and models are difficult to construct,this paper proposes a data-driven control strategy for complex networks based on heterogeneous data.By reconstructing heterogeneous data of different time,arbitrary initial state,and different final state input and output and state,a sampling trajectory representation method based on heterogeneous data set is established.Based on this method,two methods of the minimum control energy input suitable for a variety of complex network models are deduced,and the minimum amount of data required by the calculation method is discussed,and the data-driven control of various complex networks is realized,which provides a new method for the research of complex network control.

关 键 词：复杂网络 数据驱动控制 异质数据 最小控制能量 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]