大型采煤设备异常追踪检测与诊断的免疫智能方法研究

项目名称： 大型采煤设备异常追踪检测与诊断的免疫智能方法研究

项目编号： No.61472271

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 其他

项目作者： 田玉玲

作者单位： 太原理工大学

项目金额： 82万元

中文摘要： 本项目以综采设备远程监控系统为支撑，以大型煤矿设备故障诊断为目标，在获取设备运行状态数据的基础上，提出基于变化点子空间追踪的异常检测方法和基于免疫理论的智能故障诊断模型。采用多维数据流相关性分析和变化点检测方法对信号进行检测，遴选出能够反映突变状态的关键点数据作为异常活动候选解；设计变化点子空间追踪算法，利用最小的计算资源获得更高的稳定性和检测率；利用B细胞免疫网络理论建立故障传播模型，将设备系统中单元之间的故障传播的因果关系映射为免疫系统中细胞之间的交互识别关系，解决故障单元的影响通过整个系统进行传播的问题。借鉴免疫的自适应特性，建立包含记忆机制、适应性机制和决策机制的机器学习系统，以自适应的方式解决未知及早期故障的识别与诊断，同时克服因故障征兆的混叠导致故障难以辨别的问题，使系统的诊断能力达到最优，提高故障检测与诊断的效率与准确率。本项目的研究为煤矿设备的可靠运行提供了保证。

中文关键词： 人工免疫系统；故障诊断；变化点检测；采煤设备；连续学习

英文摘要： The project is supported by remote monitoring system of coal mine equipment, taking the fault diagnosis of mining equipment as the target, on the basis of the existing real time data, the anomaly detection method based on change point space tracking and the intelligence fault diagnosis model based on immune theory are proposed. The underlying methodology is to perform signal detection via the change point detection and multiple data streams correlation analysis, and the change point data reflecting mutation status as the candidate solution of abnormal is selected. The change point space tracking algorithm is designed to achieve higher stability and detection rate with the minimum computing resource. The fault propagation model is constructed by the B-lymphocyte immune network for the fault localization. The cause-consequent relationship of fault propagation of systems corresponds to the interaction between B-lymphocytes in the immune system to solve the problem of fault propagation among the whole system. Inspired by the adaptive immune system mechanisms we suggest a learning algorithms which includes the memory mechanism, adaptive mechanism and decision-making mechanism.Learning the unknown and incipient fault pattern In adaptive manner, and solves the problems that how to distinguish the faults which aroused by the overlap of the omens. The system can achieve optimal diagnosis ability and improve the efficiency and the accuracy of the detection and fault. The research can provid a guarantee for the reliable operation of mine equipment.

英文关键词： Artificial imme system;Fault diagnosis;Change Point Detecting;mining equipment;Life-long Learning