基于鲁棒性一类SVM的综导系统信息故障检测方法研究
|
摘要
为了提高舰艇综合导航系统的可靠性,并考虑到系统准确建模和大量故障数据获取的困难性,提出了一种基于鲁棒性一类支持向量机(support vector machine, SVM)的信息故障检测方法。该方法采用鲁棒性一类SVM的分类原理,并根据分类平面方程构造检测统计量,结合主元分析(principal component analysis, PCA)的方法实现实时的故障检测。实测数据试验表明,所提出的算法对含有野点的导航信息阶跃性故障和渐变性故障的检测能力优于普通一类SVM算法,而且该方法对参数的变化具有较低的敏感性。
In order to improve reliability of the integrated navigation system and in consideration of the difficulty in acquiring the accurate system model and a great number of abnormal parameters, this paper presents a message fault detection method based on robust one-class Support Vector Machine(SVM). The method adopts the principle of classification based on robust one-class SVM, and gains the expression of classification plane which combined with the Principal Component Analysis(PCA), is used to formulate the test statistic and realize the real-time fault detection. The new method does not rely on the system model and requires only normal and small sample to train the model, thus is easy in application and convenient for fault detection. The results of the experiment based on the measured data show that the method has a better detection performance for both step fault and gradual fault than the general one-class SVM for navigation message with outliers, and is not sensitive to changes in the radial basis function kernel parameter.
In order to improve reliability of the integrated navigation system and in consideration of the difficulty in acquiring the accurate system model and a great number of abnormal parameters, this paper presents a message fault detection method based on robust one-class Support Vector Machine(SVM). The method adopts the principle of classification based on robust one-class SVM, and gains the expression of classification plane which combined with the Principal Component Analysis(PCA), is used to formulate the test statistic and realize the real-time fault detection. The new method does not rely on the system model and requires only normal and small sample to train the model, thus is easy in application and convenient for fault detection. The results of the experiment based on the measured data show that the method has a better detection performance for both step fault and gradual fault than the general one-class SVM for navigation message with outliers, and is not sensitive to changes in the radial basis function kernel parameter.
引文
[1] JOERGER M,PERVAN B.Kalman filter-based integrity monitoring against sensor faults [J].Journal of Guidance Control & Dynamics,2013,36(2):349-361.
[2] 陈景霞,王建文.基于UPF的INS非高斯噪声故障检测 [J].控制工程,2012,19(5):743-746.CHEN Jing-xia,WANG Jian-wen.UPF-based fault detection method of integrated navigation system with non-gaussian noises [J].Control Engineering of China,2012,19(5):743-746.(in Chinese)
[3] BHATTI U I,OCHIENG W Y,FENG S.Perfor-mance of rate detector algorithms for an integrated GPS/INS system in the presence of slowly growing error [J].GPS Solutions,2012,16(3):293-301.
[4] SUN Y,ZHANG J,XUE R.Leveraged fault identification method for receiver autonomous integrity monitoring [J].Chinese Journal of Aeronautics,2015,28(4):1217-1225.
[5] DING D F,ZHAO X M.Navigation system failure detection based on FM Gauss wavelet tansforms [J].Applied Mechanics & Materials,2014,12(5):1874-1878.
[6] CHERRY G A,QIN S J.Multiblock principal component analysis based on a combined index for semiconductor fault detection and diagnosis [J].IEEE Transactions on Semiconductor Manufacturing,2006,19(2):159-172.
[7] RATO T J,REIS M S.Fault detection in the Tennessee Eastman benchmark process using dynamic principal components analysis based on decorrela-ted residuals (DPCA-DR) [J].Chemometrics & Intelligent Laboratory Systems,2013,125(7):101-108.
[8] 张华强,赵剡,连远峰.基于自适应概率神经网络的SINS/GPS组合导航系统故障诊断方法 [J].中国惯性技术学报,2012,10(6):749-753.ZHANG Hua-qiang,ZHAO Dan,LIAN Yuan-feng.SINS/GPS integrated navigation system fault diagnosis based on adaptive probabilistic neural network [J].Chinese Journal of Inertia Technology,2012,10(6):749-753.(in Chinese)
[9] MAHADEVAN S,SHAH S L.Fault detection and diagnosis in process data using one-class support vector machines [J].Journal of Process Control,2009,19(10):1627-1639.
[10] YIN S,ZHU X,JING C.Fault detection based on a robust one class support vector machine [J].Neurocomputing,2014,145(18):263-268.
[11] WUYTS S,GINS G,VAN DEN KERKHOF P,et al.Fault identification in batch processes using process data or contribution plots:A comparative study [J].IFAC-Papers on Line,2015,48(8):1282-1287.
[12] ALAWI A,SANG W C,MARTIN E,et al.Sensor fault identification using weighted combined contribution plots [J].Fault Detection Supervision & Safety of Technical Processes,2007,39(13):908-913.
[2] 陈景霞,王建文.基于UPF的INS非高斯噪声故障检测 [J].控制工程,2012,19(5):743-746.CHEN Jing-xia,WANG Jian-wen.UPF-based fault detection method of integrated navigation system with non-gaussian noises [J].Control Engineering of China,2012,19(5):743-746.(in Chinese)
[3] BHATTI U I,OCHIENG W Y,FENG S.Perfor-mance of rate detector algorithms for an integrated GPS/INS system in the presence of slowly growing error [J].GPS Solutions,2012,16(3):293-301.
[4] SUN Y,ZHANG J,XUE R.Leveraged fault identification method for receiver autonomous integrity monitoring [J].Chinese Journal of Aeronautics,2015,28(4):1217-1225.
[5] DING D F,ZHAO X M.Navigation system failure detection based on FM Gauss wavelet tansforms [J].Applied Mechanics & Materials,2014,12(5):1874-1878.
[6] CHERRY G A,QIN S J.Multiblock principal component analysis based on a combined index for semiconductor fault detection and diagnosis [J].IEEE Transactions on Semiconductor Manufacturing,2006,19(2):159-172.
[7] RATO T J,REIS M S.Fault detection in the Tennessee Eastman benchmark process using dynamic principal components analysis based on decorrela-ted residuals (DPCA-DR) [J].Chemometrics & Intelligent Laboratory Systems,2013,125(7):101-108.
[8] 张华强,赵剡,连远峰.基于自适应概率神经网络的SINS/GPS组合导航系统故障诊断方法 [J].中国惯性技术学报,2012,10(6):749-753.ZHANG Hua-qiang,ZHAO Dan,LIAN Yuan-feng.SINS/GPS integrated navigation system fault diagnosis based on adaptive probabilistic neural network [J].Chinese Journal of Inertia Technology,2012,10(6):749-753.(in Chinese)
[9] MAHADEVAN S,SHAH S L.Fault detection and diagnosis in process data using one-class support vector machines [J].Journal of Process Control,2009,19(10):1627-1639.
[10] YIN S,ZHU X,JING C.Fault detection based on a robust one class support vector machine [J].Neurocomputing,2014,145(18):263-268.
[11] WUYTS S,GINS G,VAN DEN KERKHOF P,et al.Fault identification in batch processes using process data or contribution plots:A comparative study [J].IFAC-Papers on Line,2015,48(8):1282-1287.
[12] ALAWI A,SANG W C,MARTIN E,et al.Sensor fault identification using weighted combined contribution plots [J].Fault Detection Supervision & Safety of Technical Processes,2007,39(13):908-913.