Applied Mechanics and Materials Vols. 494-495

Abstract: In order to improve the beer bottle mouth defect testing speed on production line, in this paper a new fast detection method based on machine vision was presented. The method made full use of the circle center symmetry of the annular region on bottle mouth, in which only pixels position in 1/8 bottle mouth area were needed to obtain the whole pixels location on the circular area. By this way much a lot time were saved. The implementation steps were described in detail. Experiments and analysis shows: the method could recognize the defect on beer bottle mouth accurately and quickly.

Abstract: Due to continuous metal-metal contacts in heavy and high-speed operating conditions, sorting machine inductions roller bearing easily occurs malfunctions. Therefore, its crucial to make incipient fault diagnosis. This paper presents a novel diagnosis algorithm using vibration analysis and information fusion. In the algorithm, vibration signal for roller bearing is firstly analyzed. Then extracted fault features are used as input eigenvector of constructed neural network classifier. In order to improve diagnosis accuracy, take output information of each single classifier as independent evidence, and aggregate them using improved Dempsters combination rule. Experiment results show that proposed algorithm has high accuracy of 99.5%.

Abstract: With the rapid development of electronic technology, the system reliability and economic requirements of the importance of the analog circuit fault diagnosis has become increasingly prominent. Aiming at the shortcomings of traditional diagnosis method, the paper presents an analog circuit fault diagnosis method based on principal component analysis of pretreatment and particle swarm hybrid neural network. The method adopts hybrid algorithm to adjust the network weights and thresholds to avoid falling into the local minimum value, which uses principal component pretreatment effectively reduce the complexity of calculation. Simulation results show that the diagnostic method can be used for tolerance analog circuit fault diagnosis, has better application prospect.

Abstract: Fault diagnosis can insure the power transformer safety and economic operation, and the data mining is the key technology of fault diagnosis for power transformer. In order to achieve the fast parallel fault diagnosis for power transformer, we need to put cloud computing technology into the smart grid. We give a parallel method of K-means based on MapReduce framework on the Hadoop distributed systems cluster to diagnose operation state of power transformer. Finally, through transformer fault diagnosis experimentations of massive DGA data, the results indicate closely linear speedup with an increasing number of node computers.

Abstract: Because of the complexity for the frequency measurement about echo signal of ultrasonic Doppler flow, the acquisition methods directly affect the accuracy of Doppler frequency shift. The algorithm of power spectrum extraction has accurate to detection echo Doppler frequency of stationary signal for the flow-velocity measurement. The key is to obtain relative useful frequency information from complex echo velocity. This study is based on the principle of Doppler velocity measurement, by studying the average frequency algorithm of echoes, and the Fourier transform method is adopted to identify Doppler signal frequency and Doppler frequency shift which was polluted by aliasing noise. The MATLAB simulation results show: the useful frequency to obtain reliable, accurate, and also laid the foundation for the measurement of multiphase flow non-stationary signals.

Abstract: A three degree of freedom vehicle dynamic model of a tractor semi trailer is established and a simulation mode is set up with the Matlab/simulink software. Then, design a Kalman state estimator, using easily measured parameters to estimate the tractor side slip angle, trailer yaw rate those are difficult measured or measuring in high cost; the simulation show that: the estimated value coming from estimator agree well with the simulation measurements, and can improve the estimation accuracy by adjusting the process noise covariance matrix and measurement noise covariance matrix.

Abstract: In order to master the safety status of the large port crane and diagnosis and predict the fault in advance for alerting, the condition monitoring and diagnosis & sensing system is used in this paper. By using sorts of sensors and accelerate meters installing on the crane, acquire all kind of signals of the machine, then put the data to the signal processing and diagnosis sensing system, the operation status of the crane can be found out, and the potential fault or failure can be diagnosis and detected in advance. The condition monitoring and diagnosis & sensing sys-tem can acquired the condition status from the large port crane and diagnosis the potential fault or failure for the machine. Thus, the problem how to master the condition of the large crane port crane and provided the safety alerting for the huge machine is solved.

Abstract: A Fault Detection scheme for T-S fuzzy systems with unknown inputs is discussed based on the LMI approach, adopting the H control theory. Firstly the reference model is introduced, a fault observer is constructed to reflect the residual of system fault. Then the proposed filter design provides sufficient conditions for the existence of a solution to the detection of faults. By means of the Projection Lemma, a quasi-convex formulation of the problem is obtained via LMI. The filter can guarantee the prescribed performance index, which minimizes the error between the residual and the real residual signal and has robust performance to unknown inputs. The effectiveness and feasibility of the proposed approach is illustrated via a numerical simulation.

Abstract: Considering the scattering difference in resistance of the strain gauge which will cause the bridge unbalance, the designers of strain measurement instrument have to add bridge-balance circuit into the design of measuring bridge in order to eliminate the problem of a DC component output caused by the unbalance. Thus, in the design of the bridge-balance circuit, the selection of balance resistances is of great significance. In this paper, through calculation analysis of the balance bridge, we obtained a couple of ideal values of balance resistances and tested them by experiments. In this condition, the experimental results have shown that the circuit has a great ability of regulation equilibrium. In brief, this paper provides the designers an effective method of choosing the balance resistances.

Abstract: Transformer partial discharge ultrasonic array detection method is a new method with high signal-to-noise ratio and high reliability. But the influence of array number to the direction finding accuracy needs to be further studied. Simulation of direction finding angle with the 15 kinds of sensor array (-arrangement) under two conditions of the acute azimuth and obtuse azimuth , study of change rule of direction finding accuracy of the different array number, for example of the square planar array sensor, based on an improved MUSIC(Multiple Signal Classification) algorithm. The results show that the direction finding accuracy increases as the array number rises, and The specific error ranges of array sensors of different array number by simulating have some degree of guiding significance in the selection of transformer partial discharge ultrasonic array sensor in the field environment.