An effective and novel roller bearing fault diagnosis technique based on empirical mode decomposition (EMD) energy entropy and support vector machine (SVM) is put forward in this article. The vibration signal of roller bearing is decomposed by EMD and the first 5 intrinsic mode function (IMF) components are obtained. SVM served as a fault diagnosis classifier and the extracted energy features of the first 5 IMFs are taken as network input vectors, and then the fault bearing and the normal bearing can be distinguished. An technique for fault of roller bearing by SVM is evaluated against a series of fault diagnosis methods that are widely used in machinery, with particular regard to the effect of training set size on fault diagnosis accuracy. We trained the SVM using RBF kernel function. We compare our experimental results with the existing results given by SMO and SVM-light algorithms. It can be seen that the fault diagnosis method based on SVM-light is superior to that based on SMO in diagnosis accuracy of roller bearing. In addition to the SVM, the same datasets were classified using RBF NN and Hopfield NN. The experimental results show that the technique of support vector machine based on EMD energy entropy has higher fault diagnosis ability.
An empirical mode decomposition (EMD) model for BeiDou Navigation Satellite System (BDS) code bias has been established upon the observation model of multiple global navigation satellite systems (multi-GNSS). To validate the correctness and effectiveness of the model, seven days from day of year (DOY) 213-219, 2015 from eight Multi-GNSS Experiment (MGEX) stations were processed. Results show that after code bias correction, the standard deviation of the multipath combination (MP) series on B1 and B2 frequencies decreased by 38.63 % and 17.4 %, respectively. The timespan needed for convergence in BDS precise point positioning (PPP) was improved by 7.9 % after inclined geosynchronous orbit (IGSO) and medium earth orbit (MEO) code bias correction, and another improvement of 11.4 % was generated by applying geostationary orbit (GEO) code bias correction. Despite the improvement of convergence time, the accuracy of the single-day solution barely increased for PPP in multi-GNSS as compared to the single GNSS. A continuous decrease in percentage along with prolonged timespan for PPP convergence was observed with increasing cut-off elevation angle. However, the performance of multi-GNSS PPP, which was superior to that of the single GNSS, shows that it is extremely valuable for practical applications in mountainous or sheltered areas.
Empirical Mode Decomposition (EMD) is suitable to process the nonlinear and non-stationary time series for filtering noise out to extract the signals. The formal errors are provided along with Global Navigation Satellite System (GNSS) position time series, however, not being considered by the traditional EMD. In this contribution, we proposed a modified approach that called weighted Empirical Mode Decomposition (weighted EMD) to extract signals from GNSS position time series, by constructing the weight factors based on the formal errors. The position time series over the period from 2011 to 2018 of six permanent stations (SCBZ, SCJU, SCMN, HLFY, FJPT, SNXY) were analyzed by weighted EMD, as well as the traditional EMD. The results show that weighted EMD can extract more signals than traditional EMD from original GNSS position time series. Additionally, the fitting errors were reduced 14.52 %, 12.25 % and 8.06 % for North, East and Up components for weighted EMD relative to traditional EMD, respectively. Moreover, 100 simulations of four stations are further carried out to validate the performances of weighted EMD and traditional EMD. The mean Root Mean Squared Errors (RMSEs) are reduced from traditional EMD to weighted EMD with the reductions of 9.08 %, 9.63 % and 6.84 % for East, North and Up components, respectively, which highlights the necessity of considering the formal errors. Therefore, it reasonable to conclude that weighted EMD can extract the signals more than traditional EMD, which can be suggested to analyze GNSS position time series with formal errors., Xiaomeng Qiu, Fengwei Wang, Yunqi Zhou and Shijian Zhou., and Obsahuje bibliografii