机电装备主控系统故障智能自愈策略研究

Aimed at the purpose of fault elimination and suppression, this subject which combined with artificial intelligence theory, artificial immune principle as well as fault self-healing thought begins within the system. In this subject, a model of intelligent fault self-healing immune system was established, while construction and implementation methods for the model was also investigated in order to deal with the main fault of the master control system of mechanical-electrical equipment..Specifically, based on the research of typical failure and anti-attenuating performance, the factor of fault characterization was adopt to describe and extract the characteristics of the system failure state, and then establish the fault descriptions model which can meet with the actual conditions. Combining this model with the expert system, a new structural model of fault self-healing immune was proposed and a simulation was also involved..This method is aimed to achieve the master control system's inherent self-healing protective properties by possess it with fault self-recovery function in the design phase and automatically eliminate the fault during the operation phase in stead of the traditional method such as "fault diagnosis-down for maintenance ", which provides a new effective way to promote the development and the engineering applications of the self-healing theory, and establishes a scientific theoretical foundation for the realization of the intelligent operation of the mechanical and electrical equipment as well.

针对机电装备主控系统的主要故障，本课题将人工智能理论、人工免疫原理和故障自愈思想相结合，以故障消除和抑制为目的，从系统内部出发，建立故障智能自愈免疫系统模型，并研究其构建和实现方法。.具体地，在研究系统典型故障和抗衰性能的基础上，采用故障特征描述子合理地描述和提取系统故障状态特征，建立符合实际工况的故障描述模型；并与专家系统相结合提出一种新的故障自愈免疫结构模型，并予以仿真验证。.该方法旨在改变传统"故障诊断＋停机维修"等方法，使主控系统在设计阶段就具备故障自愈功能，在运行中自行消除故障，实现其内在的自愈保护性能。从而为推动机电装备主控状态故障自愈理论的发展和工程化提供一种新的有效途径，为实现机电装备的智能化运转奠定科学理论基础。

为了改变传统“故障诊断＋停机维修”等方法，使机电装备主控系统在设计阶段就具备故障自愈功能，在运行中自行消除故障，实现其内在的自愈保护性能。针对机电装备主控系统的主要故障，本课题将人工智能理论、人工免疫原理和故障自愈思想相结合，以故障消除和抑制为目的，从系统内部出发，建立故障智能自愈免疫系统模型，并研究其构建和实现方法。. 本项目主要开展以下方面研究： “自愈”概念的内涵和外延进行探讨；深入研究机电装备主控系统的功能、特性、典型故障；对机电装备主控系统的有关问题进行定义，分析人工免疫系统和故障自愈系统之间在性质和行为上的相似性，建立其与人工免疫系统之间的概念映射空间，形成自己/非己、抗原/抗体等一系列概念的一一对应关系；建立起完善的自愈系统框架，该系统包括状态检测模块、信息处理模块、状态评估模块、自己/非己识别模块、免疫调节决策模块、自愈策略执行模块等。模块之间的连接形式和每个模块各自功能等均已确定；针对机电装备主控系统的软件系统，建立软件系统故障耐受模型。确定系统错误状态的产生方法。应用免疫系统特征建立具有不同选择机制的非线性系统辨识方法；另外，机电装备主控系统各部分之间的故障信息的耦合关联也会为系统故障自愈带来困难，因此，考虑各子系统之间的故障关联性也是本项目需要解决的难点。对于如何处理多故障、新型故障本项目也开展了基础性研究。在验证和改进自愈系统过程中，先后以磁悬浮系统、汽轮发电机组、电液伺服系统等为研究对象，进一步总结机电装备主控系统在故障自愈中的共性问题，为构建更加具有普适性的故障智能自愈系统提供了实验依据。. 本项目取得的理论研究成果既是对故障自愈理论的补充和发展，更是为机电装备的数字化制造和智能化制造提供一种新的系统设计方法。本项目的研究成果可以扩展应用到指导相关机电装备的设计、制造和维修工程中，具有一定的理论价值和工程意义。