| 基于神经网络的线性电路故障诊断非线性L1范数优化方法 |
| |
| 引用本文： | 何怡刚,罗先觉.基于神经网络的线性电路故障诊断非线性L1范数优化方法[J].电子测量与仪器学报,1998,12(1):18-22. |
| |
| 作者姓名： | 何怡刚 罗先觉 |
| |
| 作者单位： | 湖南大学电气系!长沙，410082，西安交通大学电气学院!西安，710049，西安交通大学电气学院!西安，710049 |
| |
| 基金项目： | 国家自然科学青年基金!59707002，高校博士点基金，国家自然科学基金 |
| |
| 摘 要： | 本文提出了一处新的模拟电路故障诊断非线性Ｌ１范数优化方法。在Ｂａｎｄｅｒ Ｌ１范数优化方法基础上增加辅助变量即将电路不可及节点电压增量做为辅助优化变量，构造一个新的故障诊断非线性约束Ｌ１范数优化问题，由一次是优化过程理到的解定位最可能故障元件；
|
| 关 键 词： | 故障诊断 神经网络 L1范数优化 线性电路 |
A Neural-based Nonlinear L1 Optimization Approach for Fault Diagnosis of Linear Circuits |
| |
| He Yigang.A Neural-based Nonlinear L1 Optimization Approach for Fault Diagnosis of Linear Circuits[J].Journal of Electronic Measurement and Instrument,1998,12(1):18-22. |
| |
| Authors: | He Yigang |
| |
| Abstract: | In this paper, a new nonlinear L1 optimization approach for fault diagnosis ofanalog circuits is presented. Based on Bander L1-norm approach 1-3] and taking the inaccessi-ble nodal voltage increment as optimized variable, a new nonlinear constrained L1-norm prob-lem is construced. The most likely faulty elements can be isolated according to the L1-norm so-lutions derived from one optimization process. The proposed modified Hpfield neural networkis used to resolve the L1-norm problem. Furthermore the neural network hardware implemen-tation using Multiple input, Multiple output Operational-Transconductance-Amplifiers(MO-MIOTA)-Capacitors is developed. The given results of illustrative examples of fault diagnosisand computer simulation show the proposed methed is feasible- |
| |
| Keywords: | Fault diagnosis Neural networks L1-norm Transconductance-capacitor |
| 本文献已被 CNKI 维普 等数据库收录！ |
