基于多源不确定信息融合的机械系统高精度可靠性分析方法

Reliability has been an important index to measure the quality of mechanical systems. It is critical to enhance the core competitiveness of mechanical systems by developing advanced reliability analysis and design methods. During the reliability analysis of practical complex mechanical systems, it is always confronted with bottleneck issues such as the severe lack of reliability experiment data and the resulting low precision of reliability analysis results. This project intends to develop multi-source uncertain information fusion mechanisms to improve the integrity of mechanical system reliability data and reduce its uncertainty degree, based on which a high-precision reliability analysis method is further proposed. Firstly, a classification criterion of uncertain information of mechanical systems is established based on information entropy, and a unified model is constructed for multi-source heterogeneous uncertain information. Secondly, considering the conflict between the multi-source information in spatial domain and the belief decay in time domain, the fusion mechanisms of multi-source uncertain information are developed. Thirdly, the reliability analysis model of mechanical systems is established, and a high-precision solution method is proposed on the basis of multi-source uncertain information fusion. Finally, the developed models and algorithms are integrated in a system, and they are verified using the reliability analysis of a multi-joint mechanical arm of the concrete pump truck. The completion of the project will provide a novel research idea and technological approach for the reliability analysis of mechanical systems under the severe lack of reliability experiment data.

可靠性是衡量机械系统质量的重要指标，发展先进的可靠性分析与设计方法对于提升机械系统的核心竞争力至关重要。针对实际工程问题中机械系统面临的可靠性试验数据严重缺乏及其导致的可靠性分析结果精度低的瓶颈问题，本项目拟研究多源不确定信息融合机制以提升机械系统可靠性数据的完整性、削减可靠性数据的不确定程度，并在此基础上发展高精度的可靠性分析方法。首先，建立基于信息熵的机械系统不确定信息分类准则，并构建多源不确定信息的统一度量模型；其次，考虑多源不确定信息在空间域上的相互冲突性与时间域上的信度衰减性，发展多源不确定信息的融合机制；再次，建立机械系统可靠性分析模型，并在多源不确定信息融合的基础上实现其高精度求解；最后，将相关模型与算法进行系统集成，并通过混凝土泵车多关节臂架的可靠性分析进行验证。项目的顺利完成将为试验数据缺乏的机械系统可靠性分析提供一条新的研究思路和技术途径。

可靠性是衡量机械系统质量的重要指标，发展先进的可靠性分析与设计方法对于提升机械系统的核心竞争力至关重要。实际工程问题中机械系统通常面临多源不确定性难以度量、多源信息冲突与零部件多状态等导致的可靠性分析结果可信度不高、嵌套可靠性优化计算效率低的瓶颈问题。本项目针对上述关键技术难点，紧紧围绕项目预定研究内容和目标，按照时间表有计划、有系统地展开研究，较好地完成了各阶段各项预定研究内容和指标，并在若干相关方面进行了扩展。本项目主要在机械系统多源不确定性建模、多源冲突信息融合、可靠性分析与高效求解、序列可靠性优化设计四个方面获得研究进展。在相关理论和方法的基础上，建立了机械系统设计的不确定性建模与可靠性分析及设计的技术，并搭建了相应的软件系统。本项目理论成果经过进一步深化和拓展后，有望应用于核工业、航空航天、汽车工程等领域机械装备的评估与设计。项目执行期间，以第一作者或通讯作者共发表学术论文8篇，其中SCI收录6篇，EI收录2篇；论文发表在本领域的一系列权威学术刊物上，如Computer Methods in Applied Mechanics and Engineering（1篇），Mechanical Systems and Signal Processing（1篇），Structural and Multidisciplinary Optimization（3篇），中国科学：技术科学(1篇)等；授权发明专利5项、软件著作权2项。项目负责人担任中国交通运输协会青年科技工作者工作委员会委员，国际学术会议分会场主席2次。