Search

In order to improve the accuracy of fault diagnosis and accelerate the speed of fault identification, this paper proposes a fault diagnosis method based on long short-term memory (LSTM) neural network, and constructs its architecture and model framework. Taking the gear failure data as the experimental object, the performance of the model is analyzed by adjusting the number of hidden layers, the number of hidden layer neurons, the learning rate, and the training times. In addition, LSTM is compared with support vector machine (SVM), convolutional neural network (CNN) and recurrent neural network (RNN) to verify that the LSTM method has a better classification effect, and the accuracy of fault diagnosis can be increased to 99.80%.

Antibacterial stainless steels have versatile applications, therefore they are of great research importance. The corrosion resistance and antibacterial properties of cerium bearing 1Cr17 stainless steel was studied by means of conventional electrochemical corrosion and antibacterial tests, and compared with those of an original batch 1Cr17 samples. The results show 0.1% of cerium alloying may enable 1Cr17 stainless steel shows better corrosion resistance and antibacterial properties than its original 1Cr17 counterpart.

This paper presents a research and design method of plug-type IP information monitoring equipment based on parallel algorithm. It mainly uses the Ethernet hub and field-programmable gate array (FPGA) technology to implement the detection and management of full-duplex IP datagram. Meanwhile using time interval as a judge to analyze and process IP datagram. By detecting, the device can meet the design requirements, with the characteristic of real-time, not to delay the network speed, not take up IP addresses, and low cost.

The carbon isotopic composition of nonmethane hydrocarbons (NMHCs) in anthropogenic sources, such as fuel volatiles, vehicle exhaust, coal-combustion exhaust has been determined by Thermal Desorption-Gas Chromatography-Isotope Ratio Mass Chromatography. And it can be concluded that different pollution sources have different δ13C values, the δ13C values of vehicle exhaust are higher than those of diesel oil volatile, gasoline volatile and air in underground parking lots, and lower than those of coal-combustion exhaust. On average alkene is enriched in 13C relative to the average of alkane by about 2%. Furthermore, differences among the pollution sources are of significance in constraining the sources of pollutants in environmental air.

Nowadays electronic banking services are widely provided all over the world, but the adoption of such services has been slower than anticipated especially in developing countries. This paper reviewed the main factors that potential users may consider when evaluating an e-commerce web site and summarized these factors into three categories: security, easy of use and transaction cost. Moreover, the paper applies fuzzy set, entropy theory and gray relational analysis in combination to evaluate the performance of e-banking web sites. In addition, a numerical case is illustrated to test the model. The results show that the potential consumers' attention toward security, transaction cost and understandability are the main factors which have significant effect on their decisions making. Finally, several suggestion and future research are discussed.

Along with popularity of electronic commerce and online purchasing, electronic payment systems have been developed and expected to the main payment means of e-commerce. But many of electronic payment systems have failed with a short-lived time. From the standpoint of buyers, this paper discusses the influential factors of electronic payment system adoption and classifies these influential factors into six dimensions. Moreover, a numerical case is illustrated for presenting rough set approach (RSA) to quantize the influence degree of these factors on electronic payment system adoption. The results derived from the case reveal that only 20 percent of buyers take electronic payment system as their first choice, security and trust have more significant impact on users' attitude than other influential factors. Finally, the paper concludes with the limitations of the study and several suggestion for future research.

Many enterprises have been aware of the severity of risks in supply chains and taken great efforts to supply chain risk management. This paper develops an evaluating index system from the standpoint of the core enterprise in supply chain, and utilizes an evaluation approach based on the modified technique for order performance by similarity to ideal solution (TOPSIS) and grey relational analysis (GRA) in fuzzy environment for supply chain overall risk assessment. In the evaluation process, the paper applies triangular fuzzy numbers to express linguistic variables that consider the subjective judgments of evaluators and applies fuzzy multiple criteria decision-making approach to synthesize the group decision. Furthermore, the paper replaces the Euclidean distance with grey relation coefficient to find a compromise solution, so both the positive ideal solution and negative ideal solution are considered, and the evaluation process ensures the objectivity. Finally, several suggestion and future research are discussed.

Shape, as one of the intrinsic features of weapons and the other military equipments, becomes the most important key factor of object recognition upon the different techniques of image formation, and the researches on techniques of image retrieval using contour, the important shape feature, have been a focus of attention in the domain of image engineering. The paper proposes a hierarchical image retrieval model based on the wavelet descriptor indexed by half-axes-angle using R-tree, which hierarchically filter the similar templates with the object pattern via the multi-resolution description techniques of the contour, and the experiments prove the efficiency and effectiveness by the vehicle image retrieval.

This work presents an approach to the starting point location of the closed contour based on the half-axes-angle and the shape matching problems using the method. It is considered that the problem core of the translation, scaling and rotation invariance of the shape recognition can be changed into the problem of the starting point location, and then be solved by the coordinates transformation. First, the invariance problems of the translation, scaling and rotation of the wavelet descriptor are analyzed, and the relevant description of the half-axes-angles is given. Then the starting point location of the curve and the shape matching strategy based on half-axes-angles are discussed. The experiments are conducted to prove the effectiveness of the method.

A novel approved algorithm based on template matching is proposed, which could be applied to the real-time tracking system. It composes of correlation matching algorithm, TD filter and ESO filter. This algorithm is used to predict the movement and to track the target. The position of target in the first image is supposed to be known. The problem is then to estimate the position of this template in subsequent images. By TD filter and ESO filter, it is possible to obtain the estimate position of the target in the current image. Then the correlation template matching algorithm is used to search target in the estimation position's neighbor in order to get target's accurate position. The simulation results of the algorithm are presented. The comparative analysis has revealed better performance characteristics of the proposed algorithm in comparison with the traditional template matching algorithm. Experimental results demonstrate the effectiveness of this method.

This paper proposes a fuzzy classification system and its application in tobacco leaves grading. The mathematic description of the fuzzy classification model is given out, and we discuss the fuzzy membership function to calculate the membership of the feature of the pattern and how to obtain the confidence of the feature vector. We also discuss the technique to standardize the pattern space and the optimization of the class spaces. And then the fuzzy reasoning technique in the classification system is provided. Finally, we apply this fuzzy classification model in the tobacco leaves grading system (TGS) to show the efficiency, and conduct experiments and comparison with different methods to prove that the TGS have the similar grading ability as a human expert in practice.

The CAC algorithm of a mobile network is responsible for reducing the handoff drop ratio (HDR) to an acceptable low level as well as achieving a high acceptance ratio for locally generated calls. Algorithms based on dynamic resource reservation depend on complex measurement, calculation and information exchange between cells. Recently, an adaptive service has been proposed to protect the QoS of mobile users and increase the capacity of a mobile network. We introduce a new sliding window based CAC algorithm especially designed for the adaptive service. This algorithm can avoid the complex estimation of the handoff probability in the dynamic algorithm, and simulation results show that this algorithm can reduce the HDR to nearly zero and can also get a higher acceptance ratio than the dynamic algorithm.