S. F. Wong

Machine condition monitoring and fault diagnosis based on support vector machine

Due to the importance of rotating machinery as one of the most widely used industrial element, development a proper monitoring and fault diagnosis technique to prevent malfunction and failure of machine during operation is necessary. This paper presents a method for gearbox fault diagnosis based on feature extraction technique, distance evaluation technique and the support vector machines (SVMs) ensemble. The method consists of three stages. Firstly, the features of raw data are extracted through the wavelet packet transform (WPT) and time-domain statistical features. Secondly, the compensation distance evaluation technique is applied to select optimal feature via sensitivities ranking. Finally, the optimal features are input into the SVMs to identify different faults. The diagnosis result shows that the SVMs ensemble is able to reliable recognize not only different faults styles and severities but also the compound faults in high accurate rate.

Machine learning method with compensation distance technique for gear fault detection

In this paper, a condition monitoring and fault diagnosis method for rotating machineries using machine learning technologies including artificial neural network (ANNs) and support vector machine (SVMs) is described. The vibration signal is acquired from gearbox used in local power generation industry for analysis of potential defects. Wavelet packet transforms (WPT) and time domains statistical are used to extraction features, and the compensation distance evaluation technique (CDET) is applied to select optimal feature via sensitivities ranking. A comparative experiment study of the efficiency of ANN and SVM in predication of failure is carried out. The results reveal that the proposed feature selection and machine learning algorithms could be effectively used automatic diagnosis of gear faults.

Signal propagation model calibration under metal noise factor for indoor localization by using RFID

Radio frequency identification (RFID) technologies as an effective indoor localization solution are acquiring increasing attention for its low cost and compactness and it plays a significant role in industrial engineering. However, the location accuracy strongly relies on environment factor that will affect the signal propagation property (using by position system). To alleviate limitation imposed by this reason unlike others, this paper proposes a new calibrated Received Signal Strength Indication (RSSI) propagation curve for high accuracy under metal noise factor. Therefore this curve is obtained by taking the metallic effect characteristic into consideration. The proposed results are obtained from practical experiments by RF Code M250 and R150 tags. Aimed at incorporating these features into localization algorithm appropriately and then the accuracy of RFDD-based positioning system under metal noise factor can be improved substantially.

Applied RFID and Virtual Reality Technology in professional training system for manufacturing

Enhancing the professional knowledge in different levels of operator is a critical success factor to advance the performance of manufacturing industry. However, the traditional training system is lack of scientific method to transfer the professional knowledge (tacit knowledge) to the operator. Applied RFID and Virtual Reality Technology in Knowledge-based Training System can convert the tacit knowledge to the explicit knowledge to different levels of operator. The trainee can capture the basic operation skill through the web-enabled and knowledge-based training system. Moreover, the system can provide the working experience and operation history about the production and tool application to the trainee through RFID technology. They can quickly and conveniently search their target tools that will apply the real manufacturing processes without any human-being supervising. The cost of training resource can be saved, because the workload of supervisor or senior operation is reduced. The senior operation can report the updating and real-time situation of production and tools through RFID technology with knowledge-based training system. The closed loop knowledge can enhance the precision level of analysis results of production system. The industry can be sustainable development through this system.

Development of Web‐based Virtual Training Environment for Machining

With the booming in the manufacturing sector of shoe, garments and toy, etc. in pearl region, training the usage of various facilities and design the facility layout become crucial for the success of industry companies. There is evidence that the use of virtual training may provide benefits in improving the effect of learning and reducing risk in the physical work environment. This paper proposed an advanced web‐based training environment that could demonstrate the usage of a CNC machine in terms of working condition and parameters selection. The developed virtual environment could provide training at junior level and advanced level. Junior level training is to explain machining knowledge including safety factors, machine parameters (ex. material, speed, feed rate). Advanced level training enables interactive programming of NG coding and effect simulation. Operation sequence was used to assist the user to choose the appropriate machining condition. Several case studies were also carried out with animation of milling and turning operations.

Intelligent car seat design with ingress, egress and sit-to-stand services

With the population ageing, it is a critical factor in the public transportation to erect a physical setting that fosters independent living. However, none of research studies designed a barrier free car seat device for supporting elderly and disabilities ingress or egress the vehicle, as well as sit to stand movement. In this present work, it applied ergonomics concept to design intelligent barrier free car seat device. The device is controlled by mobile application to facilitate them. Moreover, 22 healthy subjects volunteered the effect of standup assistant function experiment. The result indicates this function helps sit-to-stand sequences compared with normal movement.

Signal loss of RFID technology with short distance and high frequency

To obtain an automatic and intelligent monitoring system, RFID is used to aid with CCTV system which is an innovative method. Therefore, some problems exist as integrating these two technologies, and signal loss is one of the problems as RFID signal is vital to the final result. Since the targets which are necessary to be monitored through the CCTV system are dynamic, the experiments in this research are set to obtain dynamic data with specific environments, which traditional signal loss equations are not available. Empirical loss equation is accurate but computational complexity, a simply empirical loss equation is found in this research and the results are compared with the other existing loss equations.

A study of shoulder rest's feet height variation on upper trapezius muscle activity in violinists

Violin is an important role in music history. Comfort of holding violin also becomes a concern since this affects the ability and quality of performance. Literature review pointed out that upper trapezius is the related muscle when playing violin and proved that the performance of playing with a shoulder rest is greater than without it. However, the height of shoulder rests two feet was only stated that all ‘high’ is better then all ‘low’, none of any researches study the different combination of height variation of the two feet. In this study, eight subjects were invited to participate the experiment and twelve shoulder rests were provided to test by applying human factors. Surface electromyography was used to collect the left and right upper trapezius muscle activities. The results showed there were significant differences and it indicated that the variation height of shoulder rests feet could affect the upper trapezius.

Posture analysis in ergonomic study for bus station design

Bus station is one of the important facilities in mass transportation service. Nevertheless, when people are waiting for bus, they will choose different standing posture such as upright standing and leaning against wall, referring to different bus station construction. Hence the study of different standing posture provides ergonomic perspective to design bus station. The postural control in leaning-against-wall posture during prolonged standing is investigated. In present paper 20 healthy volunteers were instructed to upright stand or lean against wall for 15 minutes. The pressure distribution of foot was collected, and the standard deviation of center of pressure (CoP) in AP and ML direction and asymmetrical standing condition were analyzed. The result indicated that in leaning-against-wall posture, the percentage time and frequency of asymmetrical standing and standard deviation of CoP in AP direction are significantly different between two postures.

Study on the production forecasting based on grey neural network model in automotive industry

The automotive industry has been dramatically developed these years. However, the whole process of automotive production chain is directly affected by the accuracy of its production forecasting model, such as safety inventory quantity, out of stock losses, and timely performance. Therefore, to improve the accuracy of production forecasting, this paper uses the Chinese automotive industry as a case study, which has been the largest in the world measured by automobile unit production since 2008. It presents three kinds of combined models based on grey neural network, which are parallel grey neural network, series grey neural network, and inlaid grey neural network, compared the single model GM(1,1) and neural network. The experimental results demonstrate that the combined models are higher forecasting accuracy than the single model.