Chengdu Electromechanical College
Chengdu Electromechanical College
Mu D.,Sichuan University |
Mu D.,Chengdu Electromechanical College |
Shen B.-L.,Sichuan University |
Zhao X.,Civil Aviation Flight University of China
Materials and Design | Year: 2010
In this study, CoCrMo alloy was boronized at 950°C for 2, 4, 6 and 8h, respectively. The boronized samples were characterized by scanning electron microscopy, X-ray diffraction, microhardness tester and ring-on-block wear tester. X-ray diffraction studies showed the boride layer formed at 950°C for 2-8h consisted of the phases Co2B and CrB. A large number of pores formed in diffusion zone were probably attributed to the Kirkendall effect. Depending on boronizing time, the thickness of boride layer ranged from 4 to 11μm. The excellent wear resistance of the boronized CoCrMo alloy was attributed to the high surface hardness of the Co2B and CrB under dry-sliding conditions when compared to the as-received state. © 2010 Elsevier Ltd.
Di X.,Tsinghua University |
Fan Y.,Tsinghua University |
Shen Y.,Chengdu Electromechanical College
Computers and Mathematics with Applications | Year: 2011
Users in Service-oriented architecture (SOA) seek the best Quality of service (QoS) by service selection from the candidates responding in succession. In case the QoS changes dynamically, choosing one service and stop the searching is problematic for a service user who makes the choice online. Lack of accurate knowledge of service distribution, the user is unable to make a good decision. The Local Martingale Difference (LMD) approach is developed in this paper to help users to achieve optimal results, in the sense of probability. The stopping time is proved to be bounded to ensure the existence of an optimal solution first. Then, a global estimation over the time horizon is transformed to a local determination based on current martingale difference to make the algorithm feasible. Independent of any predetermined threshold or manual intervention, LMD enables users to stop around the optimal time, based on the information collected during the stochastic process. Verified to be efficient by comparison with three traditional methods, LMD is adaptable in vast applications with dynamic QoS. © 2011 Elsevier Ltd. All rights reserved.
Gao Y.,Chengdu Electromechanical College
Advanced Materials Research | Year: 2011
For poor accuracy of detection of the urban traffic network classification, the Support Vector Machine (SVM)is applied to classification of traffic incidents.This paper presents a traffic pattern classification method based on multi-class support vector machine, and design the network structure of the detection system. Simulation results show that: compared to other algorithms, the network, which is valid for urban transportation classification, has the advantages of high detection rate and low false alarm rate for small samples.
Bai S.,Chengdu Electromechanical College
Taiyangneng Xuebao/Acta Energiae Solaris Sinica | Year: 2014
A new structure of photovoltaic inverter with Luo-Boost circuit and three-phase four-switch units was proposed to improve the efficiency of photovoltaic power generation and reduce the cost of photovoltaic systems effectively. Accordingly, a new feedforward power predictive control strategy was brought out, which would determine the related predictive value of feedforward power through the partition and arrangement of the natural environment conditions, and then the optimization of its historical largest photovoltaic power. With small amount of calculation, fast speed, simple, high precision and reliability, the new control strategy can leave out the traditional inverter DC voltage control loop and complete the MPPT function of the photovoltaic inverter itself, so as to enhance system response speed and reliability. Both simulation and experimental results both verified the feasibility and advantage of proposed structure and control method.
Fu S.,Chengdu Electromechanical College |
Xu H.,Chengdu Electromechanical College
Journal of Materials Engineering and Performance | Year: 2010
A (Ti,V)C/Fe composite was produced by a process which combines in situ synthesis with powder metallurgy technique using Ti, Fe, FeV and carbon powder. The microstructure of the Fe-(Ti,V)C composite was studied by scanning electron microscope and x-ray diffraction. The results show that the production of an iron matrix composite reinforced by (Ti,V)C particulates using the process is feasible. Fine (Ti,V)C particles are uniformly dispersed in the pearlite. The Fe-(Ti,V)C composite possesses excellent wear-resistance under the condition of dry sliding with heavy loads. © 2009 ASM International.
Li G.,Chengdu Electromechanical College
Applied Mechanics and Materials | Year: 2011
The friction model with the effects of friction and transmission flexibility of harmonic drives system is used for adaptive impulse control design in precision control system. Experimental response based on a new pulse shape which is consisted of two harmonic expansions is studied, the first harmonic is the main pulse to drive the arm to move a displacement, the second harmonic has two functions: it enforces the main harmonic pulse in its first half that will help it to eliminate the dead zone, and provides a minus pulse in its second half to stop the arm motion quickly. An adaptive impulse feedback controller is developed based on it. Simulation and experiments show the effectiveness of the proposed controller, robustness of dead zone elimination to friction variation can be guaranteed. © 2011 Trans Tech Publications.
Wang F.,Chengdu Electromechanical College
Journal of Computational Information Systems | Year: 2012
In this paper the dynamic selection normalized subband adaptive filter (DS-NSAF) is described. It dynamically selects the subband filters based on the largest decrease of the successive mean square deviation (MSD) at every iteration. Following this, the mean-square performance of this algorithm is studied. This analysis is based on energy conservation arguments and does not restrict the inputs to specific models or to being Gaussian. Simulation results illustrate the good agreement between theory and practice. © 2012 Binary Information Press.
Zhu C.,Chengdu Electromechanical College
Applied Mechanics and Materials | Year: 2012
The present work reported the preparation of V(C,N)-reinforced iron matrix composite via in situ synthesis. The phases and morphologies of the sintered composite were investigated using X-ray diffraction (XRD) and scanning electron microscopy(SEM). Dry sliding wear test had been carried out to study the wear behavior of V(C,N)-reinforced iron matrix composite. The results show that a uniform distribution of the V(C,N) reinforcing phase throughout the iron matrix is successfully obtained. Compared with hardened 45 steel, the wear resistance of the V(C,N)-reinforced composite is better. © (2012) Trans Tech Publications, Switzerland.
Wang Q.,Chengdu Electromechanical College
2011 IEEE 3rd International Conference on Communication Software and Networks, ICCSN 2011 | Year: 2011
The complexity and the nonlinear characteristics of Highway traffic incident, this paper presents a detection method based on Artificial Neural Network, designs AID algorithm based on testing facilities, simulation results show that BP algorithm has the advantages of high detection rate, low false alarm rate and short detection time after comparing with traditional algorithm. But there are also shortcomings to make further research. © 2011 IEEE.
Tang T.,Chengdu Electromechanical College
Proceedings - International Conference on Natural Computation | Year: 2012
Image-based and minutiae-based are two major methods of fingerprint recognition. In this work, we presented an image-based fingerprint recognition method by using wavelet transformation and this method is efficient even for low quality fingerprint. The features extraction of the proposed method differing with previous wavelet methods is based on the blocks of enhanced region of interest (ROI). The alignment is required to build ROI including location the reference point and rotation alignment. Fingerprint matching was performed on simply Euclidian distance of feature vector extracted from wavelet domain. These features consist of mean energy, standard deviation and Shannon entropy for the purpose of making these features more discriminative. The good recognition accuracy was achieved on the FVC2002 database. © 2012 IEEE.