工业云环境下的制造服务协作机理与可靠性分析理论研究

With the gradual implementations and applications of a series of industrial cloud platforms, it has become an inevitable trend for manufacturing enterprises to meet the needs of personalized and complex tasks by conducting manufacturing service collaboration so as to realize value co-creation and sharing. However, due to the complex environment of industrial cloud platforms and the social characteristics of its manufacturing services and tasks, the enough social collaborative relationships between manufacturing services are unknown, the reliability of distributed collaboration is not high, and the utility of group collaboration is also not balanced. In response to the above problems, this project further summarizes a scientific problem, namely the principle of manufacturing service collaboration, under the industrial cloud platform environment. Introducing the theory of complex networks with community structure, the modeling as well as evolutionary and statistical analysis of a manufacturing service collaboration network are conducted in turn. As the four constituent parts of the systematical principle of manufacturing service collaboration, the constitutive principle, the evolutionary principle, the failure principle and the behavior principle are ascertained sequentially. Finally, the reliable and efficient collaboration mechanisms are correspondingly put forward. It would provide manufacturing enterprises the theoretical guidance to answer the following questions, such as (a) which manufacturing service is available to collaborate with, (b) is the manufacturing service reliable to collaborate with, (c) is the manufacturing enterprise who the manufacturing service belong to willing to collaborate with, and so on.

随着一系列工业云平台的逐步实施与应用，制造企业通过制造服务协作来满足个性化复杂任务需求并实现价值共创与共享已成为必然发展趋势。然而，由于工业云平台的复杂环境及其所汇聚的制造服务与任务需求的社会性特征，造成制造服务社会化协作关系不明、分布式协作可靠性不高、多主体协作效用不均衡等难题。针对上述问题，基于前期承担的青年科学基金和面上项目分别在制造服务组合协作、供需匹配等方面的工作基础，本项目进一步归纳工业云平台环境下的制造服务协作机理这一科学问题。考虑制造服务群体协作特征，基于群落结构复杂网络模型及理论，通过制造服务协作网络的构建、演化与统计分析，依次探明制造服务协作的构成机理、演化机理、失效机理和行为机理，并相应提出制造服务可靠与高效协作机制。为能够提供制造服务的制造企业存在的①面临协作需求时不知与谁协作、②评估能否开展协作时不敢协作、③决策是否开展协作时不愿协作等问题，提供理论指导依据。

本项目将群落结构复杂网络理论引入到社会化制造服务动态协作研究中，探究了工业云环境下的制造服务协作机理，研究了制造服务按需动态协作理论及其高效可靠协作机制，并对所提出的理论和方法进行仿真分析与验证。主要成果如下：.（1）探明了制造服务协作要素构成；从协作关系分类与挖掘、协作模式分类与特征提取、群体协作动态行为描述等方面揭示了制造服务协作构成机理，构建了基于复杂网络群落结构的制造服务协作网络模型。.（2）提出了针对制造服务协作网络6类失效的检测方法；揭示了不同类型失效在不同基本网络结构中的级联传播特性；设计了考虑失效影响的容错控制机制，提出了基于按需备份机制的制造服务可靠协作方法。.（3）建立了考虑双边耦合性以及单边非理性的提供者长期效用模型和短期效用模型，提出了考虑协作者长/短期效用的制造服务协作共识机制，并探明了协作行为机理，实现了协作激励。.（4）设计了基于群落结构复杂网络的制造服务协作仿真软件，解决了可用测试数据集获取、协作动态调度问题求解效果难以保证和基准不一等问题，为制造服务协作运营理论模型与方法提供了统一的测试环境。.依托上述工作：①获国家科技进步二等奖（排名第一）；②培养副教授2名，博士后4名，博士生5名，硕士生4名；③获国家发明专利授权7项，软著2项；完成英文专著1部；发表论文19篇，其中在IEEE Trans.等期刊发表SCI论文15篇，EI4篇，累计SCI引用1447次，谷歌学术引用2764次，2篇入选ESI高被引用论文，单篇最高SCI引用670次；④获教育部长江学者特聘教授、科睿唯安全球高被引科学家、美国制造工程师协会SME智能制造领域20 Most Influential Academics等人才奖项6项；⑤被聘为国家制造强国建设战略咨询委员会智能制造专家委员会委员、中国机械工程学会理事等学术兼职；⑥在CIRP/IEEE等系列会议做大会/特邀报告16次。