研究了一类考虑错分代价的故障预测技术安全性能评估方法。鉴于概率论不精确的特性，针对两类故障类别的统计数据构建了错分代价模型，在此基础上确立了错分代价最小的故障趋势判别定理；从两类故障类别出发，构建了多类故障类别预测的错分代价模型以及错分代价最小的故障趋势判别规则，并给出了故障预测技术安全性能评估的基本流程；用火箭控制系统故障检测的实例验证了错分代价模型的有效性，并对故障预测技术未来的安全性能评估的未来发展趋势进行了探讨。
Abstract
A new safety performance evaluation of the fault-prediction technology was researched based on misclassfication cost. First of all, a misclassification cost model for the statistical data of two fault-classificatory sets was constructed based on the inaccurate characteristic of probability theory, and the fault-direction decision rules of the running state were established by considering the minimum misclassification cost. Secondly, a misclassification cost model of numerous fault-classificatory sets was constructed, and the fault-direction decision rules were established by considering the minimum misclassification cost. And a basic flow of safety performance evaluation was provided for the fault prediction technology. Finally, the effectiveness of the model was verified by taking the fault detection of launch vehicle control system as an example, and the further development trend of the fault prediction technology was discussed.
关键词
系统评估与可行性分析 /
概率论 /
故障预测 /
性能评估 /
错分代价
{{custom_keyword}} /
Key words
systematic evaluation and feasibility analysis /
probability /
fault prediction /
performance evaluation /
misclassification cost
{{custom_keyword}} /
基金
国家自然科学基金项目（61004118）； 中国国家博士后科学基金项目（20060401018）; 重庆市自然科学基金项目（CSTC2006BB2422）； 重庆市教委科学技术研究项目（KJ060414）；重庆交通大学博士启动基金项目（07-01-12）
{{custom_fund}}
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］ LazaroJesus, Arias Jagoba, Astarloa Armando, et al. Hardware architecture for a general regression neural network coprocessor [J]. Neurocomputing, Dedicated Hardware Architectures for Intelligent Systems, 2007, 71(1-3): 78-87.
［2］ UbonratSiripatrawan, Pantipa Jantatwat. A novel method for shelf life prediction of a packaged moisture sensitive snack using multilayer perceptron neural network [J]. Experts Systems with Applucations: An International Journal, 2008,34（2）:1562-1567.
［3］ ZemouriRyad, Racoceanu Daniel, Zerhouni Noureddine, et al. Training the recurrent neural network by the fuzzy min-max algorithm for fault prediction[J]. AIP Conference Proceedings, 2009, 1107(1): 85-90.
［4］ WangHong-jiang, Leung Chi-sing,Sum Pui-Fai, et al. Kernel width optimization for faulty RBF neural networks with multi-node open fault[J].Neural Process Lett.,2010,(32):97-107.
［5］ HuangDa-rong, Huang Xi-yue, Xiang Chang-cheng,et al. The covariance constraint control method of fault detect for complicated systems [J]. Lecture Notes in Artificial Intelligence, 2006, (8):580-588.
［6］ AdelF Iskander, Akmal A Younis. A proactive fault tolerance management algorithm for mobile Ad hoc networks [J]. IEEE, 2007,(6): 571-575.
［7］ 刘东，张春元，李瑞，等.软件容错模型中的容错实时调度算法[J].计算机研究与发展，2007, 44(9): 1495-1500.
LIU Dong, ZHANG Chun-yuan, LI Rui,et al. Fault-tolerant real-time scheduling algorithm in software fault-tolerant module [J]. Journal of Computer Research and Development,2007, 44(9): 1495-1500. (in Chinese)
［8］ ZhouZhi-jie, Hu Chang-hua, Fan Hong-dong, et al. Fault prediction of the nonlinear systems with uncertainty [J]. Simulation Modeling Practice & Theory, 2008, 16(6): 690-70.
［9］ 黄大荣，黄丽芬.灰色系统理论在故障预测中的应用现状及其发展趋势[J].火炮发射与控制学报，2009，(3):88-92.
HUANG Da-rong, HUANG Li-fen. Present situation and development tendency of grey system theory in fault forecast application [J]. Journal of Gun Launch & Control,2009,（3）：88-92. (in Chinese)
［10］ MhaskarP, Gani A，Christofides P D. Fault-toleran control of nonlinear processes: performance-based reconfiguration and robustness[J]. International Journal of Robust and Nonlinear Control, 2006, 16(3):91-111.
［11］ HilgertNadine, Rossi Vivien, Vila Jean-Pierre, et al. Identification, estimation, and control of uncertain dynamic systems: a nonparametric approach [J]. Communications in Statistics-Theory and Methods, 2007, 36(14): 2509-2525.
［12］ CatalCagatay, Diri Banu. A systematic review of software fault prediction studies [J]. Expert Systems with Applications, 2009, 36(4): 7346-7354.
［13］ CatalCagatay, Diri Banu. Unlabelled extra data do not always mean extra performance for semi-supervised fault prediction [J]. Expert Systems, 2009, 26(5): 458-471.
［14］ 赵刚，黄大荣，黄席樾.火箭漏电故障诊断技术研究[J].兵工学报，2010,31(7):916-921.
ZHAO Gang, HUANG Da-rong, HUANG Xi-yue. Research on the leak current fault diagnostic technology for launch vehicle [J].Acta Armamentarii, 2010, 31(7):916-921. (in Chinese)
［15］ 黄大荣，黄丽芬.基于集对分析联系数故障树的BA系统可靠性分析[J].计算机应用研究,2010，27(1)：111-113.
HUANG Da-rong, HUANG Li-fen. Reliability analysis of BA system based on connection number of set pair analysis and FTA[J]. Application Research of Computer, 2010, 27(1)：111-113. (in Chinese)
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
