复杂应力下电子系统剩余寿命定性定量混合智能预测方法

Electronic system has become the key of modern industrial equipments' safe and reliable operation, and its reliability study gains more attention now. In this project, in order to integrate tightly with the balanced objectives of accurate and fast prediction of remaining useful life (RUL), we will aim at the characters of complex, fast change and uncertainty during the operation of electronic system under complex stress, and then develop a new fast intelligent prediction technique combined with qualitative and quantitative methods via comprehensive analysis the uncertainty in RUL prediction. This project will study those aspects which include the action mechanism analytical method of uncertainty to electronic system; knowledge unified representation method; new mechanism combined with qualitative method and quantitative method, and then establish a new RUL prediction model of combining quantitative and qualitative methods with network structure. In addition, we also study the prediction schemes using artificial intelligence methods and the fast reasoning method based on the advantages of network-based new model such as open, extensible and parallel processing. Moreover, we will develop simulation verification platform and tool to assess the proposed method. This research will realize the breakthrough of RUL prediction theory of electronic system and provide theoretical and technical support to improve the capability of reliability, safety, long term working stability and economic affordability of electronic system under complex stress. The research has significant contributions to theoretical and practical values, especially, to enrichment and perfection of the fault and RUL prediction theory. It also plays an important role in improving the current equipment management and maintenance policy.

电子系统已成为影响现代工业系统/设备可靠安全运行的关键，其可靠性研究被日益重视。项目以综合均衡地提高电子系统剩余寿命预测准确性和快速性为目标，针对复杂应力下系统运行中复杂快变不确定的特点，综合分析其剩余寿命预测的不确定性，研究不确定性对电子系统的作用机理，不确定性信息/知识的统一表示方法以及定性定量方法结合的新机制，建立具有网络化结构的定性定量混合模型；利用网络化结构开放性、可扩展和可并行处理的优势，融合人工智能方法，探索快速推理方法，形成定性定量相结合的快速智能预测新方法；开发仿真验证工具。研究将丰富电子系统剩余寿命预测理论与方法，为提高复杂应力下电子系统的可靠性、安全性、长期工作稳定性以及经济可承受性提供技术支撑，具有积极的科学意义和推广应用价值，并对完善故障与寿命预测理论具有重要理论意义，对改革现行设备管理和维修保障模式具有重要推动作用。

项目以综合均衡地提高电子系统剩余寿命预测准确性和快速性为目标，选取典型的复杂电子系统，针对复杂应力下系统运行中复杂快变不确定的特点，综合分析其剩余寿命预测的不确定性，研究了不确定性对电子系统的作用机理，不确定性信息/知识的统一表示方法以及定性定量方法结合的新机制，建立了具有网络化结构的定性定量混合模型；融合人工智能方法，探索快速推理方法，形成了定性定量相结合的快速智能预测新方法，并开发了仿真验证工具。研究丰富了电子系统剩余寿命预测理论与方法，为提高复杂应力下电子系统的可靠性、安全性、长期工作稳定性以及经济可承受性提供技术支撑，具有积极的科学意义和推广应用价值，并对完善故障与寿命预测理论具有重要理论意义，对改革现行设备管理和维修保障模式具有重要推动作用。