Iaeng Transactions on Engineering Technologies Volume 7

The following sections are included:

• PREFACE

• CONTENTS

In this paper, we focus on hierarchical multiobjective stochastic linear programming problems (HMOSLP) where multiple decision makers in a hierarchical organization have their own multiple objective linear functions together with common linear constraints. In order to deal with HMOSLP, a probability maximization model is applied. By considering the conflict between permissible objective levels and the values of the corresponding probability function in a probability maximization model, it is assumed that each of the decision maker specifies not his/her own permissible objective levels but the membership functions of both his/her own permissible objective levels and the corresponding probability functions. After eliciting membership functions, the interactive algorithm to obtain a satisfactory solution is proposed through the fuzzy decision, in which the hierarchical decision structure is reflected through the decision power and the proper balance between permissible objective levels and the corresponding probability function is attained. Interactive processes are demonstrated by means of an illustrative numerical example.

Staff scheduling has become increasingly important for both the public sector and private companies. Good rosters have many benefits for an organization, such as lower costs, more effective utilization of resources and fairer workloads and distribution of shifts. The process of constructing optimized work timetables for the personnel is an extremely demanding task. Driver rostering, preceded by vehicle scheduling and driver scheduling, is the last phase in the bus transit scheduling process. This paper presents a successful way to schedule days-off on a yearly basis and shifts on a monthly basis in one of the Finnish bus transportation companies. The days-off and shifts are scheduled using an algorithm that includes features from population-based methods, simulated annealing, tabu search and ejection-chains. The generated software has been integrated into a third-party vendor product.

This paper focuses on a new Stackelberg location problem on a network with demands whose weights are given uncertainly and vaguely. By representing them as fuzzy random variables, the optimal location problem can be formulated as a fuzzy random programming problem for finding Stackelberg equilibrium. By using both their α-level sets for fuzziness and their satisfaction level for a given probability for randomness, it can be reformulated as a version of conventional Stackelberg location problem on the network. Theorems for its complexity are shown based upon the characteristics of the facility location.

BEES algorithm, inspired by the natural foraging behaviour of honey bees, is one among new meta-heuristic optimisation techniques. However, the performance of the algorithm depends on its parameters levels and need to be determined and analysed before its implementation. This paper proposes the integrated algorithm based on steepest descent algorithm (SDA) and the Nelder—Mead modified simplex method (MSM) for investigating the proper BEES parameter levels for truckload trucking management. The dynamic multi-zone dispatching problem concentrates on the quantities of inbound and outbound in each area and it is modified from the multi-zone dispatching. The factors of the rearrangement penalty of the area, in each zone, including time periods are also included. In this research there are various levels of areas, zones, inbound, outbound and time periods. In practice both elements from this integrated algorithm are very easy to implement. An aim to integrate SDA and MSM is to produce faster and more accurate convergence when managing zones with minimal imbalance scenario via an application of the BEES algorithm. The main purpose of the paper is to demonstrate how the BEES can be improved by incorporating a proper parameter level selection strategy. Experimental results were analysed in terms of best solutions found so far, mean and standard deviation on both the imbalance and execution time to converge to the optimum. In a suite of tested problems taken from Thai leading transportation company, computational results of the investigation of parameter selection show that the integrated approach outperforms other two original techniques of the SDA and MSM in terms of solution quality and convergence rate. The comparison shows that the integrated algorithm provides largely favourable performance measures of accuracy, robustness and problem evaluation. Finally a recommendation of proper level settings of BEES parameters for some selected problem sizes that can be used as a guideline for future applications of the BEES. This is to promote ease of use of the BEES in real life problems. This study also found that number of zones affect iterations toward the optimum. Number of areas affects the imbalance. The parameters of zone and area are then the important variables for these multi-zone dispatching systems.

Powerful methods for solving the noisy and complex multi-variable problems using finite sequential instructions can be categorised into optimisation and approximation algorithms. The latter might be defined as an iterative search process that efficiently performs the exploration and exploitation in the solution space aiming to efficiently find near optimal solutions. Various natural intelligences and inspirations have been adopted into meta-heuristics. In this work, bees, firefly and bat algorithms were adapted to find global optimums of non-linear continuous mathematical models with various levels of measurement noise. Considering the solution space in a specified region, some models contain global and multiple local optimums. Bees algorithm is a population-based search algorithm mimicked by the natural foraging behaviour of swarms of honey bees. Firefly algorithm is under a consideration of the flashing characteristics of fireflies. Finally, Bat algorithm is formulated based on the echolocation behaviour of bats. A series of computational experimental results were analysed in terms of best solutions found so far, mean and standard deviation on both the actual yields and execution time to converge to the optimum. Firefly algorithm seems to be better when the noise levels increase. Bees algorithm provides the better levels of computation time and the speed of convergence. In summary, the firefly algorithm is more suitable to exploit a search space by improving individuals’ experience and simultaneously obtaining a population of local optimal solutions. The bat algorithm with a balanced combination of the advantages of existing successful algorithms seems to gradually converge to the optimum.

The multi-dimensional nature of the ship evaluation and selection problem and the imprecise and subjective characteristics of the human decision making process make it difficult for the decision maker to consistently determine the relative importance of the criteria in a given situation. As a result, inconsistent criteria weightings are often produced, leading to unreliable decisions being made in selecting the most suitable ship for a specific shipping task. This chapter presents a decision support system approach for effectively assisting the decision maker in determining the criteria weightings in ship evaluation and selection. An example is presented for demonstrating the applicability of the approach for criteria weighting in the ship evaluation and selection process.

This paper proposes an exact and explicit solution algorithm to obtain the optimal solution of a fuzzy linear programming problem with the second-order cone and box constraints of decision variables. The main problem is not well-defined due to fuzziness, and so Yager’s ranking method is introduced. Then, the proposed model under Yager’s ranking method is transformed into a deterministic equivalent problem. The proposed solution algorithm is based on a parametric solution approach to determine the optimal strict region of parameters, and the main procedures are to perform deterministic equivalent transformations for the main problem and to solve the KKT condition of auxiliary problem satisfying the optimality.

In this paper, we develop a learning environment that combines “real lectures” and “e-Learning”. We study the feasibility of user-based homework assignments by the degree of comprehension and also quantitatively evaluate the degrees of achievements for each student. Not only active students but also inactive students can get assignments through the Web. Therefore, we expect that a familiar and easy study situation can be created by developing this e-Learning environment. Our system uses the accumulated learning history and information recommendation processing based on collaborative filtering for making assignments. Thus, giving assignments to each student based on his skill level can be combined with the tendency for covering every topic of study. The result of our research considers the feasibility of content-based filtering by text mining processing towards homework assignments in future systems.

In this paper a singly diagonally implicit Runge-Kutta-Nyström (RKN) method is developed for second-order ordinary differential equations with periodical solutions. The method has algebraic order four and phase-lag order eight at a cost of four function evaluations per step. This new method is more accurate when compared with current methods of similar type for the numerical integration of second-order differential equations with periodic solutions, using constant step size.

Turbulence transition in boundary layer flows arises from nonlinear wave generation, interaction, and amplification. These wave dynamics depend on the intricate phase synchronization of the propagating waves. Numerical simulation of the process encounters challenges in achieving sufficient computational resolution, dealing with truncated computational domains, and control of numerical errors. Application of highorder, optimized Combined Compact Difference numerical methods help to mitigate these issues and achieve realizations of nonlinear wave dynamics during turbulence transition. Validation of the numerical results attests to the accuracy of the model.

In this research, fuzzy multiple objective production and logistics planning model with interval demand and uncertain capacity is proposed. Two fuzzy goals are considered; profit goal and changes of workforce level goal. In conventional aggregate production planning (APP) models, logistics planning is not included. Although it is a critical criterion that creates extra cost. Moreover, demand is considered as a crisp demand using the forecast demand which always has an error. Actual demand is uncertain in nature and does not exactly equal to the forecast demand. So, APP with the possible interval demand that the best solution of the demand in each period can be selected is proposed in this research. This determination can extremely reduce change of workforce level, lost sales and costs of conventional plan. Uncertain capacity is also considered, which makes the model more realistic for use in the real application. The single objective model and the fuzzy multiple objective models with crisp demand are also compared and discussed with the proposed model. A case study of a real factory is illustrated to show the effectiveness of the proposed model.

This research aims to determine the factors affecting to nonconforming from surface mounting process and to optimize the condition of such factors in order to reduce the amount of nonconforming, permanently. The alternative improvement will be proposed from Taguchi and constrained response surface optimization methods. Surface mounting process consists of three sub-processes including solder printing, mounting, and reflowing processes. The most potential nonconforming were a component missed position and no component. From failure mode and effect analyses there were 10 potential factors affecting to these nonconforming. They also included one nonadjustable factor. By applying 2k factorial experiments, it was found that there were five potential factors which were the design aperture, the reflowing profile, the vision parameter, the number of fiducial marks and the number of push up pins. The interaction effect between the number of push up pins and the number of fiducial marks was also found. On the constrained response surface optimization a multiple linear regression model in a concept of the path of steepest descent, then, was applied to determine the suitable level of each factor. With the suitable level found, it was found that the variation of process was reduced providing the reduction of the defect rate from 17,234 to 5,450 PPM. Finally, Taguchi method was applied to determine the proper level of the potential factor to achieve the most reduction of defect. It was found that the defect rate was further reduced to 2423.5 PPM when the process condition was set as the proper level found.

A new fuzzy multi-objective Cell Formation (CF) model for Cellular Manufacturing System (CMS) is proposed in this research. Conventionally a facility planner solves a problem of CF by classifying parts into families and group machines into cells, respectively. However, existing methods for solving the CF problem are difficult and complicate. Moreover, efficient solutions of some of those methods are not guarantee. So, the efficient method based on two important performance measures, called Exceptional Elements (EE) and the Void Elements (VE) of a perfect grouping, are developed. Preemptive Fuzzy Goal Programming (P-FGP) is applied to these two performance measures for finding the efficient solution. The problems of grouping part families and machine cells can be simultaneously solved. Moreover, machines and parts grouping can also be adjustable to find the prefer solution. Numerical examples of existing literatures are demonstrated to compare and show the effectiveness of the proposed model over the conventional methods. Moreover, parameter setting of the proposed model is also suggested.

The current trend to shortening product lifecycles and the uncertainty in the dynamics of global markets force companies to increase flexibility in supply chain management. In spite of the importance of the interdependencies between product design and supply chains, the contribution of product design principles to supply chain management flexibility has not been, so far, sufficiently investigated in the scientific literature. This research will show the positive benefits of combining postponement and component commonality in the automotive industry. Insights from literature and a case study are combined to evaluated the impacts on the supply chain variables coordination, collaboration and configuration and its flexibility. The notion of this work is to convince companies in making combined component commonality and postponement decisions rather than separate ones.

Aerospace manufacturing system is very intricate due to the extreme material and information flow. Management of these variables needs Manufacturing Execution System (MES). A manufacturing execution system for the composite component manufacturing workshop of an aerospace enterprise has been developed. The material delivery system is the core module of the MES that described in detail. An active material delivery model is proposed to instead of the traditional passive material supply model. An improved work flow of material delivery is introduced in detail for optimizing the traditional material supply. Barcode technology is used in the material delivery system. By scanning the barcode, material information and working personnel information can be collected and inputted into the system in real time. The function module of material delivery system is introduced. The integration relationship between the material delivery system and other systems is also described. Web technology and browser/server architecture is adopted to design and develop the system. By implementing the material delivery system, the material supply time and cost can be reduced.

Small bias and high robustness at optimal variable settings are desirable properties to all the responses involved in a multiresponse optimization problem. An approach that considers those properties and can be easily used by practitioners is presented. Its feasibility is illustrated using two examples from the literature and the results compared with those of other effective methods.

Related information of optimal cutting parameters for machining and spring force operations is required for process planning. Numerous nonlinear constrained machining models have been developed with the objective of determining optimal operating conditions. The purpose of this article includes studying three meta-heuristic algorithms to test their efficiency in solving five benchmark machining models. Three promising meta-heuristics for the numerical process improvement consist of particle swarm optimisation (PSO), firefly (FFA) and hunting search (HuS) algorithms. A brief description of each algorithm is presented along with its pseudocode to facilitate the implementation and use of such algorithms by researchers and practitioners. Benchmark comparisons among the algorithms are presented in terms of processing time, convergence speed, and quality of the results. The experimental results show that FFA is clearly and consistently superior compared to the PSO and HuS with respect to precision as well as robustness of the results including design points to achieve the final solution. Only for simple data sets, all algorithms can obtain the same quality of performance measures. Apart from higher levels of performance measures, FFA is easy to implement and requires hardly any parameter tuning compared to substantial tuning for PSO and HuS algorithms.

Effect of a non-uniform basic temperature gradient and magnetic field on the onset of Béenard-Marangoni convection in a horizontal micropolar fluid layer bounded below by a rigid plate and above by non-deformable free surface subjected to a constant heat flux is studied. The lower rigid surface and the upper non-deformable surface are assumed to be perfectly insulating. Six different non-uniform basic state temperature profiles are considered. The resulting eigenvalue problem is solved using the Rayleigh-Ritz technique, and the influence of various parameters on the onset of convection is discussed.

This paper reports the implementation of a Driving Aptitude Test (DAT) for screening in potential candidates for a transportation company. The DAT involves both the aptitude test and personality tests. The aptitude test consists of various common perceptual and motor skills such as mechanical knowledge, spatial analysis, emergency preparedness and handling ability considered to be necessary for driving. The personality test examines drivers’ service and safety attitudes. It was shown that the use of the DAT can provide a reliable and systematic method to select potential and capable candidates for personnel decisions.

This study was undertaken to investigate the effects of demographic factors of age, gender, education level, major discipline, work nature, and years of work experience on the usability assessment of safety signs. The relationship between SUS (System Usability Scale) score and comprehension accuracy and the color associations for sign design were also assessed. Three hundred and ninety-eight participants were first asked to complete a self-administered questionnaire on safety sign comprehension and then a modified SUS questionnaire. The results showed that education level was the only demographic factor that had a major effect on sign usability. Participants with the higher diploma education perceived the sign usability significantly better than the diploma students. Besides, the perceived sign usability was found to be positively and significantly related to the comprehension accuracy, indicating the usability of safety signs can be judged by how effectively the signs could communicate with the readers. For the color associations in sign design, the criterion for determination of population stereotype was only achieved in red-danger association. The findings of this study could serve as a valuable reference on assessing the usability of safety signs and other graphic symbols with the SUS instrument.

Spatial Stimulus-Response (S-R) compatibility is an important concept for humanmachine interface design. Many studies have been conducted to investigate its impact and effects under different circumstances. However, studies concerning effects of spatial S-R compatibility on fingers are rare. The purpose of this paper is to review what have been done on finger S-R compatibility in order to facilitate the formation of research plans and methodologies for future finger reaction time studies.

This chapter describes the investigation of a converter based flexible ac transmission system (FACTS) device for the dynamic control of power flow in power networks. As one of the series-type FACTS devices, the static synchronous series compensator (SSSC) is the most versatile and powerful FACTS device that can provide effective means for controlling real and reactive power flow on the transmission line and improving the overall stability of power systems. However, converter based FACTS devices are still very costly and the internal converter circuitries have complex and coupled system dynamics which require advanced controllers to achieve satisfactory performances. This chapter aims to investigate a hardware simplified (2-leg, 4-switch) SSSC in performing various real and reactive power flow control functions during steady-state and dynamic operations of power systems. To achieve better dynamic performances, the space vector pulse width modulation (SVPWM) techniques are used. For comparative purposes, two topologies, i.e. 2-leg 4-switch and 3-leg 6-switch based SSSC are investigated in this study. Detailed mathematical models and simulation studies carried out in the Matlab/Simulink environment are firstly addressed. Typical results are then presented to verify the feasibility of the 2L-4S SSSC and the effectiveness of the proposed controllers.

This paper presents an analysis on the selection of an appropriate activation used in neural networks for fault classifications on underground cable. A decision algorithm based on a combination of discrete wavelet transforms (DWT) and back-propagation neural network (BPNN) is developed. Simulations and the training process for the backpropagation neural network are performed using ATP/EMTP and MATLAB. The variations of first scale high frequency component that detect fault are used as an input for the training pattern. Various cases studies based on Thailand electricity distribution underground systems have been investigated so that the algorithm can be implemented. Various activation functions in hidden layers and output layers are compared in order to find and to select the best activation function. It is found that the use of Hyperbolic tangent-function for the hidden layers, and Linear activation function for the output layer gives the most satisfactory accuracy in these particular case studies.

The semiconductor industry has given renewed interest to the asynchronous technology since a number of limiting factors exist in modern synchronous digital systems. NULL Conventional Logic (NCL) is a Delay-Insensitive (DI) clockless paradigm convenient for implementing asynchronous circuits but lacks efficient analysis methods and tools for specification and verification. Based on Delay Insensitive Sequential Process (DISP) specification, this chapter exemplifies application of formal methods by applying Process Analysis Toolkit (PAT) to model and verify behavior of NCL circuits. Some useful constructs (Boolean AND gate, toggle element), are successfully modeled and verified using PAT. The flexibility and simplicity of modeling, simulation and verification show the usefulness and applicability of PAT for NCL circuit design and verification.

In this article, we proposed a noise detector with high linearity in CMOS 0.18um process. By removing the precharging capacitor and exchanging the connection of control signals, we can provide a stable charging voltage source to enhance power supply noise detection linearity from 0.9291 to 0.9986. By using separated supply voltage sources, we set the detection path with a higher supply voltage of 2.5V to turn on the detection circuitry immediately when supply voltage drops. In this way, the noise detection range can be enlarged and the noise detection accuracy can be enhanced. Finally, the power supply noise detection error can be lowered to 9.62%

The growing interests of Business-To-Business e-Commerce (B2B e-Commerce) have led to a rapid progress towards “e” transformation especially for Small and Medium Sized Enterprises (SMEs). Literature is indicative of the growth of the B2B sectors in all industries, and B2B e-Marketplace is one of the sectors that have witnessed a rapid increase. To achieve this broad perspective, there is a need to explore the e-Marketing abilities and services derived from B2B e-Marketplace. This study provides a new systematic framework for Malaysian SMEs to compete in the dynamic “e” environment derived by empirical results obtained from various industrial sectors.

We consider such a mobile ad hoc network system that consists of two kinds of nodes: the servers that provide services and the clients that request services provided by servers. An SCDS (Server-centered CDS)-based approach and a maintenance algorithm of SCDS are proposed for improving performance of the system that is suffering from high resource contention. The essence of the proposed approach is to reduce the possibility of bottleneck occurrences caused by the servers, and thus to improve the packet delivery fraction. Results of simulation experiments demonstrated the effectiveness of the proposed approach.

It is noted that Elliptic curve cryptosystem (ECC) has been playing one of major roles in wireless sensor networks (WSNs), which was first proposed by Miller [10] and Koblitz [9] that relies on the difficulty of elliptic curve discrete logarithmic problem (ECDLP). It is gaining wide acceptance as an alternative to the conventional public key cryptosystem such as RSA [24], DSA [25], DH [26]. Also it is known that the wireless sensor networks based on the rapid progress of wireless communications and embedded micro electro mechanical systems technologies are becoming important part in our daily life. The security of the WSN becomes one of the major concerns in its applications. Even ECC prominently offers great potential benefits for WSN security there is still a lot of work needs to be done due to WSN has very restraint running conditions such as limited energy source, capability of computing, etc. It is well known that a scalar multiplication operation in ECC accounts for about 80% of key calculation time on wireless sensor network motes. In this paper we present an optimized dynamic window based on our previous research works. The whole quality of service (QoS) has been improved under this algorism in particularly the power consuming is more efficiently. The simulation results showed that the average calculation time, due to fuzzy conditions decreased from previous 26 to current 9 as a whole the calculation time, decreased by approximately 18% in comparison to our previous algorithms in an ECC wireless sensor network [23]

One of the main challenges for Wireless Sensor Networks (WSN) is their limited lifetime due to finite node energy. Although we can overcome this problem by optimizing the power consumption of nodes through clustering, optimum clustering of WSN is an NPHard problem. Therefore this paper presents a hybrid algorithm based on Genetic Algorithms and Particle Swarm Optimization (or artificial bee colony) that overcomes clustering problems by finding the best number of cluster heads, the cluster heads themselves and the cluster members. Simulation results reveal that the proposed scheme outperforms the simple Genetic Algorithm based clustering scheme, PEDAP, PEGASIS and LEACH.

Multimode fiber (MF) at high frequency region has recently been considered as a high data rate medium for signal transmission. This high data rate transmission is made possible since there are many available passbands, which can be utilized as channels for sending many subcarrier signals. These passbands are located at different frequencies and vary from fiber to fiber. These different frequencies can be determined from the delays of guided modes in the fiber. The passbands at high frequency region of multimode fiber are studied in this paper. Rayleigh distribution is used to statistically model the delays of guided modes. The characteristics in terms of the average magnitude response and the average bandwidth of these passbands are considered by varying many parameters for different multimode fibers. From the study, it is found that the average magnitude response of the passbands depends on the number of guided modes in the fiber. Additionally, the average bandwidth of these passbands depends on the standard deviation of the delays.