Abstract
Detection of out-of round (OOR) faults of metro vehicle wheels is very important to improve stationarity and stability in metro vehicles and avoid accidents caused by OOR faults. Diagnosis of OOR faults demands extracting useful information accurately from mass of vibration signals with poor signal-to-noise ratio (SNR) of metro vehicle wheels for complex running condition. In this paper, we proposed a diagnosis method on OOR faults of metro vehicle wheels combined with variational mode decomposition (VMD), kernel principal component analysis (KPCA), and deep belief network (DBN) to diagnose the OOR faults of metro wheels. Vibration signals of China metro vehicle wheels collected while the metro vehicle is running are used to train the diagnosis model and adjust parameters of DBN and KPCA based on testing accuracy. The different dimensions of KPCA, epoch number, and node number of DBN are compared, and the better parameters of diagnosis model based on vibration signals are concluded in this paper. The generalization of the diagnosis model is checked nine times by testing the calculation of each group of parameters and using an error declining process. The mean accuracy of diagnosis model proposed in this paper is 0.9136, and the diagnosis model presented in this paper is very significant to detect OOR faults online.
Highlights
Metro vehicles are one of urban transportation modes in most large and medium-sized cities for metro vehicles’ advantages of safety, fast speed, and comfort
We proposed a diagnosis method on OOR faults of metro vehicle wheels combined variational mode decomposition (VMD), kernel principal component analysis (KPCA), and deep belief network (DBN) and match parameters of the diagnosis model to diagnose the OOR faults of metro wheels
We propose a method of KPCA-DBN to diagnose OOR faults rapidly and accurately without artificial feature extraction and selection for effective information of raw data being kept by KPCA [16, 17]
Summary
Metro vehicles are one of urban transportation modes in most large and medium-sized cities for metro vehicles’ advantages of safety, fast speed, and comfort. In order to diagnose accurately OOR faults of metro vehicles online, we may build a system which includes noise reduction, extraction of components of faults, improving diagnosis accuracy, and online speed. Deep learning is often used to predict passenger flow and safety prediction of rail transit system [15,16,17], and we try to combine variational mode decomposition (VMD), kernel principal component analysis (KPCA), and deep belief network (DBN) to detect OOR faults of metro wheels in this paper. We proposed a diagnosis method on OOR faults of metro vehicle wheels combined variational mode decomposition (VMD), kernel principal component analysis (KPCA), and deep belief network (DBN) and match parameters of the diagnosis model to diagnose the OOR faults of metro wheels
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
