基于改进LS-SVM的航空发动机传感器故障诊断与自适应重构控制
Fault diagnosis and adaptive reconfiguration control for sensors in aeroengine based on improved least squares support vector
-
摘要: 提出了一种基于改进LS-SVM的航空发动机传感器故障诊断与自适应重构控制方法.该方法通过给误差变量赋予不同权值因子提高LS-SVM的鲁棒性,采用修剪算法提高LS-SVM的稀疏性;该方法从某涡扇发动机输入输出空间中建立其正常模型,采用阈值判别法对传感器故障进行实时监视与诊断,并用模型输出值代替故障传感器测量值反馈回闭环控制系统,实现对发动机的自适应重构控制.仿真结果表明,该方法能及时准确地定位故障,并进行有效的自适应重构控制.Abstract: Amethod of fault diagnosis and adaptive reconfiguration control based on improved least squares support vector machine(LS-SVM) was presented for sensors in aeroengine control system.LS-SVMrobustness was improved by adding weighed values to errors,and its sparseness was improved by clipping algorithm.Some turbofan engine normal models were set up with improved LS-SVM.The threshold discriminance was used to real-timely watch and diagnose sensors fault.Adaptive reconfiguration control for aeroengine was realized by model output which is fed back to closed loop control system in replacement of fault sensor measurements.The simulation results show that the proposed method can exactly diagnose sensor faults in time,and also effectively realize adaptive reconfiguration control.
-
[1] 钱坤,谢寿生,胡金海,等.并行鲁棒观测器对发动机减小转速信号的故障分离[J].航空动力学报,2005,20(1):136-141.QIAN Kun,XIE Shousheng,HU Jinhin,et al.Fault isolation of aeroengine speed decreasing signal based on parallel multiple robust observers[J].Journal of Aerospace Power,2005,20(1):136-141. [2] 何保成,于达仁,史新兴.应用传感器仿真模分析发动机控制系统故障[J].推进技术,2001,22(5):364-367.HE Baocheng,YU Daren,SHI Xinxing.Application of simulation model of sensors to analysis fault of control system of turbojet engine[J].Journal of Propulsion Technology,2001,22(5):364-367. [3] Boskovic J D,Bergstrom S.Robust integrated flight control design under failures,damage and state-dependent disturbances[J].Journal of Guidance,Control,and Dynamics.2005,28(5):902-917. [4] Duan Zhuohua,Cai Zixing,Yu Jinxia.Unknown fault detection for mobile robots based on particle filters[J].Proceedings of the 6th World Congress on Intelligent Control and Automation,IEEE,2006:5452-5457. [5] Koenig D.Unknown input proportional multiple-integral observer design for linear descriptor systems:application to state and fault estimation[J].Automatic Control,IEEE Transactions on,2005,50(2):212-217. [6] Surkens J A K,Vandewalle J.Least squares support vector machine classifiers[J].Neural Processing Letters,1999,9(3):293-300. [7] 朱家元,杨云,张恒喜,等.基于优化最小二乘支持向量机的小样本预测研究[J].航空学报,2004,25(6):565-568.ZHU Jiayuan,YANG Yun,ZHANG Hengxi,et al.Data prediction with few observations based on optimized least squares support vector machines[J].Acta Aeronautica et Astronautica Sinica,2004,25(6):565-568.
点击查看大图
计量
- 文章访问数: 1492
- HTML浏览量: 3
- PDF量: 531
- 被引次数: 0