Ho Y.-L.,Hsing Wu Institute of Technology
Journal of Convergence Information Technology | Year: 2012
This study provides a hybrid method of two-plant production decision. The Back-Propagation Network (BPN) is firstly applied to predict the future market demand of the target product for the next two years. A system dynamics (SD) model is then constructed to simulate the cost-orientated production decision of two plants. BPN is applied in this study because of its high-speed computing power, fast recall speed, high learning accuracy, and fault tolerance. In addition, the proposed system model can be divided into 4 subsystems, demand, judgement, production, and human resource. The proposed model considers that time delay of material demand and order demand would be affected by the amount of market demand, cycle time and the total human resource input, cost-orientated judgement of production decision, production capacity, product shortage, and human resource policy to satisfy the practical applications. The BPN results can offer an accurancy prediction of market demand, and SD simulation results can provide the references of cost-orientated production decision of two plants under the constrained production capacity, employee strategy, material inventory and product shortage situations. Besides, some unpredictable situation inserted, SD model can quickly simulate the changes of system behavior, and then make enterprise have enough time to prepare those unpredictable situations. This work shows that any enterprise can apply the proposed hybid method to solve its case easily by changing input data only.
Hsieh H.-C.,Hsing Wu Institute of Technology |
Chiang M.-L.,Chaoyang University of Technology
Journal of Parallel and Distributed Computing | Year: 2011
Reliability is an important research topic in distributed computing systems consisting of a large number of processors. To achieve reliability, the fault-tolerance scheme of the distributed computing system must be revised. This kind of problem is known as a Byzantine agreement (BA) problem. It requires all fault-free processors to agree on a common value, even if some components are corrupt. Consequently, there have been significant studies of this agreement problem in distributed systems. However, the traditional BA protocols focus on running ⌊(n-1)3⌋+1 rounds of message exchange continuously to make each fault-free processor reach an agreement. In other words, since having a large number of messages results in a large protocol overhead, those protocols are inefficient and unreasonable, especially for some network environments which have large number of processors. In this study, we propose a novel and efficient protocol to reduce the number of messages. Our protocol can collect, compare and replace the received values to find the reliable processors and replace the values sent by the unreliable processors. Subsequently, each processor can agree on a common value through three rounds of message exchange. Furthermore, the proposed protocol can use the minimum number of messages to tolerate the maximum number of faulty components in a distributed system. © 2011 Elsevier Inc. All rights reserved.
Lan K.-T.,National Taiwan University of Science and Technology |
Lan C.-H.,Hsing Wu Institute of Technology |
Kuo C.-H.,Nanhua University
Journal of Convergence Information Technology | Year: 2012
It is known that the global marine resources are exhausted little by little and the fish amounts are decreasing year by year. To restore and nurture the marine resources, more and more artificial reefs which have been proved to be effective to attract and gather marine lives are widely deployed by the countries adjacent to the sea or ocean. However, few artificial reefs are deployed on a fixed base or with a finite quantity. Besides, the maximum fractal dimension is rarely considered to design and deploy a three dimension artificial reef ecology (3D-ARE). Therefore, the study is working on a better solution of how to deploy a 3D-ARE which has the largest fractal dimension on the basis of the evolutionary technology. In this research, the evolutionary algorithm (EA) is employed to design a 3D-ARE which has the maximal fractal dimension. The larger the fractal dimension is, the better the deployment is. The searching space is very large and complex with many local optimal values to trap the individuals of the population. Therefore, Cauchy mutation which is better at exploring the rugged landscape and expanding searching space is well-applied in this research. By examining and analyzing different sizes of bases and numbers of reefs, the EA on the 3D-ARE design and deployment is confirmed to be feasible. With the conclusion of this study, to construct a 3D-ARE with the limited base-space, resource, and budget is able to achieve. In the future, the possibilities of further research about the related issues will be expectable and practicable.
Chen C.-Y.,Hsing Wu Institute of Technology |
Kuo J.-T.,China University of Science and Technology |
Chung Y.-C.,China University of Science and Technology
Environmental Technology (United Kingdom) | Year: 2013
The use of matured compost as an inoculation agent to improve the composting of vegetable and fruit wastes in a laboratory-scale composter was evaluated, and the commercial feasibility of this approach in a pilot-scale (1.8×104 L) composter was subsequently confirmed. The effect of aeration rate on the physico-chemical and biological properties of compost was also studied. Aeration rate affected the fermentation temperature, moisture content, pH, O2 consumption rate, CO2 production rate and the formation of odour. The optimal aeration rate was 2.5 L air/kg dry solid/min. The CO2 production rate approached the theoretical value during composting and was linearly dependent on temperature, indicating that the compost system had good operating characteristics. The inoculation of cellulolytic bacteria and deodorizing bacteria to compost in the pilot-scale composter led to an 18.2% volatile solids loss and a 64.3% volume reduction ratio in 52 h; only 1.5ppmv odour was detected. This is the first study to focus on both operating performance and odour removal in a pilot-scale composter. © 2013 Copyright Taylor and Francis GRouép, LLC.
Ho Y.-L.,Hsing Wu Institute of Technology |
Peng W.-N.,Ardentec Corporation
Advances in Information Sciences and Service Sciences | Year: 2012
According the fast changes of customer desire, the core subject of business administration for an enterprise is to create its competitive advantage. Based on the research motivation of efficient and fast decision, the main objective is to construct a mathematical model with the considerations of production yield, multi-production intervals, multi-order, multi-plant, multi-material purchase, diverse order deadline, inventory cost, quantity discounts for each material, material holding cost, and the production setup cost to provide a referenced decision-making platform under the current complicated competitive environment. This work tries to combine the synchronized, real-time and the high computing capability of computer to handle the integrated problem of production planning, material purchasing and inventory planning for diverse products with different deadlines under the considerations of the least cost with finite production capacity. In addition, this model can not only be categorized as an Integer Nonlinear Programming (INLP) problem, but function as a decision-making tool for real-time analyses of the planning for multi-order by multi-plant production. Moreover, the above-mentioned mathematical model is constructed by using the syntax of Lingo 11.0 extended version, and the built-in Global Solver is selected as the solving method to search the global optimum. A exemplified case is then followed to describe the solving results. Through the proposed model, a factory can use the proposed model to solve other cases easily by changing its input parameters only.