In this paper, two kinds of dynamic models for shaft misalignment of rotating machinery are proposed for vibration analysis and diagnosis of the shaft misalignment state. In order to obtain the solution of the dynamic models and clarify the vibration signal features measured in the shaft misalignment state, the calculation method of vibration forces caused by misalignments is also shown. The results of computer simulation and experiment using the same rotating machine are shown to verify the efficiency of the dynamic analysis method proposed in this paper. Finally, the method for distinguishing structure faults of rotating machines (shaft misalignment state, unbalance state and looseness state) is discussed by using symptom parameters and spectra of the vibration signal measured in these states.

Submitted on January 9, 2017; Revised on May 16, 2017; Accepted on May 24, 2017
References: 21

Liver illness is the most hazardous ailment that influences a large number of individuals consistently and ends man's life. An effective diagnosis model is required in the process of liver disease treatment. This study accordingly aims to employ Case-Based Reasoning (CBR) methodology supported by Genetic Algorithm (GA) to optimize the prediction results of liver disease and to analyze their performances on different datasets. CBR methodology has been implemented to find the prediction results of liver disease for different datasets. We proposed a GA-based CBR framework to compare its performance with CBR in order to observe how effective it is at predicting liver illness. CBR prediction accuracy is very low so it is not very much appreciated. The proposed GA-CBR integrated model outperforms the CBR model by achieving better accuracy for all used datasets of liver disease. In this optimization of weights of features and selection of suitable instances are done simultaneously rather than separately. This leads to better prediction performance as compare to independent models. The outcome of this model illustrates that performance of usual CBR enhances fundamentally by utilizing our integrated GA-CBR model approach.

Submitted on March 17, 2017; First Revised on April 22, 2017; Second Revised on May 22, 2017; Accepted on May 24, 2017
References: 41

This study is intended to highlight the importance of flywheels, in particular reaction wheels and steps that can be taken to ensure that they remain in stable condition for the duration of their mission. While there is an ample amount of work on this topic, this study provides an easy to follow innovative two step approach to tackle the said issue and a starting point for a different kind of analysis based on vibrations. The proposed methodology starts by obtaining the vibration data to be analyzed and applying a data based feature extraction technique known as Stochastic Resonance (SR) to the data. SR is a fairly novel tool which has shown a lot of promise in the context of mechanical systems fault diagnosis. The results obtained from the application of SR to the data is usually in the time domain but is converted to the frequency domain to reveal more information which can be used to take appropriate corrective action. As a safeguard, vibration suppression using nonlinearity which is an emerging tool is applied as a second step, to counteract whatever vibrations that may occur, thus leading to a more stable reaction wheel system.

Submitted on April 9, 2017; Revised on June 12, 2017; Accepted on June 18, 2017
References: 17

Many statistical distributions including Rayleigh, Rician, Nakagami-m, Hoyt, η- μ and κ-μ are proposed in literature to model fading in wireless Communication systems. This paper presents the study of MIMO system in a recently proposed Two Wave with Diffuse Power (TWDP) fading model, having two specular multipath components in the presence of diffusely propagating waves. It has been verified that the TWDP fading model is better suited to represent real world frequency selective fading on the basis of data collected from wireless sensor networks. This paper studies the performance of MIMO system in TWDP fading scenario. Upper and lower bounds of the capacity for MIMO system have been derived in TWDP fading scenario.

Submitted on March 22, 2017; Revised on May 27, 2017; Accepted on June 15, 2017
References: 11

Gene’s expression data contains hundreds to thousands of features. It is challenging for machine learning algorithms to find the relevant information from such huge and correlated data. Irrelevant and redundant features are computationally costly and decrease the accuracy of machine learning algorithms. Feature selection plays important role to solve the problem of dimensionality. But most of the traditional feature selection algorithms fail to scale on high dimensionality problems. In this paper Clustering based Feature Selection Framework named as (CFSF) is proposed. CFSF produces optimal feature subset by eliminating irrelevant features using clustering algorithm and redundant features by applying filter measure on each cluster. Extensive experiments are carried out to compare proposed framework and other representative methods with respect to two classifiers namely Naive Bayes and Instance Based on microarray datasets. The empirical study demonstrates that the proposed framework is very efficient and effective for producing optimal feature subset and improves classifier performance.

Submitted on December 4, 2016; Revised on May 7, 2017; Accepted on June 18, 2017
References: 25

Exploded view of a product is used in the applications for understanding the intricate details of the assembly. The problem with all the present-day CAD software packages is that the reconciliation of the product assembly from the exploded view is performed inaccurately wherein the parts collide with one another and the original assembly is not regained back through collision free path thereby leading to non-optimal performability of the software package. This present work proposes an efficient process for successfully generating the exploded view for assemblies through collision free paths, so that the reconciliation of product assembly from exploded view is accurate. The technique adopted here is based on the assembly coherence data and disassembly feasibility testing and is proposed in the context of a computer aided geometric assembly of a knuckle joint and internal combustion engine connecting rod assembly.

Submitted on March 11, 2017; Revised on July 4, 2017; Accepted on July 6, 2017
References: 18

Dielectric electro-active polymer (DEAP) is a new smart material named “artificial muscles”, which has a remarkable potential in the field of biomimetic robots. However, hysteresis nonlinearity widely exists in this material, which will reduce the performance of tracking precision and system stability. To deal with this situation, a radial basis function (RBF) neural network combined with sliding mode control algorithm is presented for a second-order DEAP linear actuator. Firstly, an inverse hysteresis operator based on Prandtl-Ishlinskii (P-I) model is used to eliminate hysteresis behavior. Secondly, an adaptive RBF neural network sliding mode controller is designed to obtain high tracking accuracy and keep system stability. The proposed algorithm makes the tracking error converge to zero and keeps the system globally stable in the case of external disturbances and parameter variations. Simulation results demonstrate that the proposed controller has the superiority to a pure sliding mode controller.

Submitted on January 14, 2017; Revised on May 14, 2017; Accepted on June 17, 2017
References: 18

This paper presents the combination of pulse position modulation (PPM) and pulsed fiber laser in order to improve the problem that the pulsed fiber laser with high power but low repetition frequency results in lower data transmission rate and limits its application. We compare the bandwidth requirement, power requirement, bandwidth utilization and error performance of several schemes of LPPM, DPPM, MPPM and OOK. Also, the modulation rate of three PPM based on fiber laser is obtained. It is shown that DPPM performs best in bandwidth requirement, bandwidth utilization and modulation rate. Moreover, DPPM doesn’t require symbol synchronization, which is of great importance in optical communication system. Then the optimal value for DPPM and modulation characteristic of fiber laser are presented. Finally, through simulation and test, we get the results, which shows the feasibility of the scheme.

Submitted on February 8, 2017; Revised on April 25, 2017; Accepted on June 19, 2017
References: 10

The importance of the supply of electricity to economic growth is self-evident. However, the existing research mostly focuses on the linear relationship between electricity consumption and economic growth, while ignoring the nonlinear relationship between them. What’s more, the existing research also ignores the spatial correlation of power consumption and its spillover effect on economic growth. In order to address the deficiency of the existing research, this study empirically analyzes the effects of Chinese provincial electricity of economic growth based on the panel threshold model and panel spatial model with 30 provincial samples from 1995 to 2014. The results show that China's power consumption causes threshold effect and spatial spillover on economic growth. When per capita GDP, per capita consumption expenditure, and per capita investment exceed a certain threshold, the positive role in stimulating economic growth of electricity consumption is significantly different from those of the elasticity when the per capita GDP, per capita consumption expenditure, and per capita investment do not exceed the corresponding thresh value. Not only does the local electricity consumption significantly promote economic growth of the local region, the electricity consumption spillover effects of the adjacent provinces can also promote the economic growth of the local region. If the spatial spillover effect of electricity consumption is neglected, the supportive effect of electricity consumption on economic growth will be underestimated.

Submitted on January 2, 2017; Revised on April 15, 2017; Accepted on June 13, 2017
References: 10

Although a large variety of outlier detection techniques have been developed, the algorithms pay less attention to the impact of structure factor on semantics for string data, and the threshold is difficult to be given automatically with unknown distribution law of string data, so the accuracy of string outlier detection is difficult to be ensured. This paper presents a similarity entropy-based self-adaptive string outlier detection method to address this issue. Firstly, semantic similarity is calculated by matrix computation based on word matching, and structure similarity is calculated by considering the structure factors. On this basis, string data is mapped into similarity cells, and they are detected to identify outlier data by using similarity distance. In order to reduce the sensitivity problem of threshold, the similarity entropy histogram is constructed to determine the dynamic threshold. The simulation experiments are conducted to prove the feasibility and rationality of this method, and the results show that this method can reduce sensitivity problem of threshold and ensure accuracy.

Submitted on February 27, 2017; Revised on April 2, 2017; Accepted on May 17, 2017
References: 13

Adapting the frequently changed topology is the main task of data delivery in vehicular ad hoc networks. Making advantage of the characteristics that the mobility patterns and positions of vehicles are predicable, this paper presents an improved dynamic hop choosing mechanism for data delivery. It exploits predicting the positions and mobility patterns of vehicles, and also takes into account the vehicular traffic density. Accordingly, the next hop chosen by the proposed strategy is seldom out of reach even if the topology is quickly changed. Simulation results show that our approach improves the packet delivery ratio and reduces the network latency when compared with state of the art protocols.

Submitted on February 13, 2017; Revised on May 25, 2017; Accepted on June 17, 2017
References: 11

Cervical cancer, with an annually increasing incidence rate, is becoming the leading cause of death among women in China. However, studies have shown that the early detection and accurate diagnosis of cervical cancer contribute to the long survival of cervical cancer patients. The machine learning method is a good substitute for manual diagnosis in the analysis of Pap smear cervical cell images, reflecting its effective and accurate classification. In the present study, a framework for cervical cancer diagnosis is presented based on a random forest (RF) classifier with ReliefF feature selection. Using preprocessing, segmentation, and feature extraction, 20 features were extracted. In the feature selection phase, 20 features were ranked according to weight using ReliefF. In the classification phase, the RF method was used as a classifier, and different dimensions of features were selected to train the classifier. To examine the efficacy of the proposed method, the Herlev data set collected at Herlev University Hospital was used, in which 917 Pap smear images were categorized into two classes: normal and abnormal. After a 10-fold cross validation, the experimental results showed that the best classification performance was obtained with the top 13 features based on the RF classifier, which were better than Naive Bayes, C4.5, and Logistic Regression. The accuracy was 94.44%, and the AUC value was 0.9804. The results also confirmed the effectiveness of cytoplasm features in the classification.

Submitted on January 29, 2017; Revised on April 12, 2017; Accepted on June 23, 2017
References: 47

TLD (Tracking-Learning-Detection) algorithm can be good for a long time to track the target in rotation, occlusion, illumination and other circumstances. However, in the case of uneven illumination, occlusion, tracking target fuzzy and so on, the problem of false tracking or tracking failure often occurs. Aiming at the shortcomings of TLD tracking algorithm, this paper will take TLD as the basic framework of target tracking and improve the detection module. When the tracking target has better texture feature, the SLBP (Semantic Local Binary Pattern) classifier is used to replace the nearest neighbor classifier in the detection module, which converts the image into SLBP texture feature vector to classify the samples using. In this paper, TLD-SLBP, MEEM, SCM, Struck and TLD are compared by using CVPR2013Benchmark test platform. The experiment results show that the TLD-SLBP algorithm has a higher success rate than other algorithms.

Submitted on February 3, 2017; Revised on April 22, 2017; Accepted on June 16, 2017
References: 14

It is very important to classify and identify the quality of potato by computer vision. In order to realize the accurate and fast classification of potato, a classification and recognition method based on support vector machine and PCA is proposed. Study uses normal potato, green potato, germinated potato and damaged potato as the experiment sample. A total of 600 images were collected where 150 images of each sample was collected. The SVM multi classifier is designed to train the classifier based on the PCA principal component vector, and the key parameters of the classifier are optimized to improve the overall recognition rate of 96.6%. Separately, the normal potato recognition rate is 97.5%, the greened potatoes is 96.3%, the damaged potato is 95.0% and the germinated potato is 97.5%. The research results provide technical support for the intelligent grading of fruit and vegetable quality.

Submitted on February 3, 2017; Revised on May 2, 2017; Accepted on June 8, 2017
References: 17

Applying cognitive radio to railway communication systems is a cutting-edge research area. This paper aims to solve the optimization problem of the global channels opportunistic accessibility in railway cognitive radio environments. In particular, we propose an efficient cooperative model for multiple wayside base stations. This model consists of Bayesian inference to calculate the probability of successful transmission on a single station along with team collaboration to maximize network performance within a group of base stations. Instead of only performing the traditional sensing and assigning, the base stations have an ability to learn from the interactions among others and the environment to gain prior knowledge. The base station agents further analyze prior knowledge and perform optimal channel assignment for global network performance. Using our cooperative model of channels opportunistic accessibility, we have shown that the model can also reduce the computational complexity in high-mobility communication environments.

Submitted on December 27, 2016; Revised on March 2, 2017; Accepted on June 17, 2017
References: 28

In recent years, wireless sensor networks (WSNs) technology has been widely used in various fields because of its advantage of low construction cost. Based on the characteristics of sensor-limited power, there are presently many studies that explore how to extend the life cycle of WSNs under limited power conditions. Some research designed sensor wake-up scheduling mechanisms according to the quorum system that can not only achieve the effect of energy-saving, but also ensure the rendezvousing opportunities between the sensors. However, there is a problem of rendezvous idle when the sensor performs a sensing task. In order to improve the power-saving efficiency of the sensors, this paper designs an energy balance quorum system(EBQS). In addition to ensuring the opportunity to rendezvous between the sensors, this system can also balance the remaining capacity of the sensors and improve the power-saving efficiency of the traditional Quorum System depending on the remaining capacity of the sensors and the way they rendezvous. The experimental results show that the EBQS proposed in this paper can effectively reduce the number of slots compared to the traditional quorum system scheduling mechanism.

Submitted on January 24, 2017; Revised on March 16, 2017; Accepted on May 25, 2017
References: 10

Given a query point set and an object point set, a multi-source query of k nearest neighbors (MQ-kNN) returns the query set for its k closest objects. However, most existing nearest neighbor query algorithms are based on a single-source query point (SQ-kNN), and the query point is often the user's location. However, in some cases, a query can be a point set. For example, a user wants to choose a house from the existing idle houses (query points) and hopes that its surrounding facilities (object points) are best. For this kind of application, we study the problem of MQ-kNN on road networks and try to solve MQ-kNN query problems. A basic algorithm based on Dijkstra algorithm is proposed as an original algorithm by calculating SQ-kNN repeatedly. Then, two improved algorithms are proposed by taking all query points as a whole and adopting the effective pruning strategy. Comprehensive experiments on five road network datasets clearly demonstrate the efficiency of this method.

Submitted on February 13, 2017; Revised on May 5, 2017; Accepted on June 25, 2017
References: 15

An active learning method is put forward to filter Chinese spam. In terms of training the filtering model, labeling all of the emails seems to be costly and time-consuming, while unlabeled emails can be easily accessed. Misclassification and a low-certainty method is proposed to reduce the number of labeled emails. The ROSVM model is also utilized as the online filtering model. The experimental results show that the proposed method not only decreases the number of training emails and the computational cost of spam filter, but also improves the accuracy of the filter.

Submitted on February 20, 2017; Revised on May 11, 2017; Accepted on June 15, 2017
References: 21

Multiple-instance learning (MIL) has been successfully utilized in image retrieval. Existing approaches cannot select positive instances correctly from positive bags, which may result in low accuracy. Inspired by the characteristic that consistencies are always among instances and instances, and bags and instances, we propose a new algorithm called multiple instance learning based on instance-consistency (MILIC) to mitigate this issue. First, we select potential positive instances effectively in every positive bag through the minimum cost of instance-consistency. Second, we use the L1-LR to select irrelevant instances from potential positive instances to further improve the retrieval efficiency. Then, we design a novel feature representation scheme based on the irrelevant potential positive instances to convert a bag into a single instance. Band on the feature representations, we finally conduct object-based image retrieval and image categorization by adopting the standard single-instance learning (SIL) strategy, such as the support vector machine (SVM), to verify the effectiveness of our proposal.

Submitted on March 1, 2017; Revised on April 27, 2017; Accepted on June 15, 2017
References: 16

Researching on destination prediction has a particularly important influence on the location-based services' popularization. The traditional destination prediction algorithm is to retrieve the historical trajectory data to find the same trajectory sequences as the query trajectory and then derive the most likely location to be the predicted result. However, due to the limitation of the historical trajectory data, this method has low efficiency and accuracy. Thus, in this paper, we propose the Prediction algorithm based on time (PBT algorithm), which considers the influence of the factor of time on destination prediction. Experiments based on real data show that in terms of destination prediction, the PBT algorithm not only alleviates the limitation of the historical data in the traditional algorithm to make the results more realistic, but also is more effective.

Submitted on February 14, 2017; Revised on May 3, 2017; Accepted on June 15, 2017
References: 11

In wireless sensor networks, the LAP-(D, k) query problem can be seen as a special Top-k query problem with the constraint that the Euclidean distance between any two locations corresponding to the data items in the Top-k query results should not be smaller than the given value D. To solve this problem, a novel dynamic-programming-based heuristic algorithm named LAPDK is proposed. LAPDK firstly divides the sensing field into hexagonal cells using some geometry methods. Then, it finds the approximate solution of the LAP-(D, k) query problem based on parts of the data generated by the sensor nodes in some preferential cells using the dynamic programming technique. Finally, it further optimizes the solution based on the sensing data received by the Sink node. Simulation results show that LAPDK not only decreases the energy cost of WSNs but also obtains a better approximation ratio compared to the existing state-of-the-art scheme for the LAP-(D, k) query problem.

Submitted on February 24, 2017; Revised on April 26, 2017; Accepted on May 30, 2017
References: 16

Bearing fault diagnosis is a key research content of condition-based maintenance for machineries. Because of noise interference, incipient bearing fault is always difficult to be found. Traditionally, a large number of filtering algorithms used are limited to the cases of Gaussian noise or linear operation. In this paper, the adaptive Myriad filter and alpha stable model are elaborated. Myriad filter is a non-linear filter, which can be effectively applied in impulsive environment. The order of Myriad filter can be determined by alpha stable model. Finally, both laboratory fault data and real fault data from a public data set are used to verify the efficiency of the proposed method.

Submitted on December 7, 2014; Revised on November 18, 2016; Accepted on April 23, 2017
References: 10