Browsing by Author "Hsu, HM"

Now showing 1 - 15 of 15

A fuzzy reasoning based diagnosis system for (X)over-bar control charts
(2001-03-01)
Hsu, HM
;
Chen, YK
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
This paper describes a new diagnosis system, which is based on fuzzy reasoning to monitor the performance of a discrete manufacturing process and to justify the possible causes. The diagnosis system consists chiefly of a knowledge bank and a reasoning mechanism. The knowledge bank provides knowledge of the membership functions of unnatural symptoms that are described by Nelson's rules on (X) over bar control charts and knowledge of cause-symptom relations. We develop an approach called maximal similarity method (MSM) for knowledge acquisition to construct the fuzzy cause-symptom relation matrix. Through the knowledge bank, the diagnosis system can first determine the degrees of an observation fitting each unnatural symptom. Then, using the fuzzy cause-symptom relation matrix, we can diagnose the causes of process instability. In conclusion we provide a numerical example to illustrate the system.
A genetic algorithm for the optimal design of microwave filters
(1999-12-01)
Hsu, HM
;
Tsai, SP
;
Wu, MC
;
Tzuang, CK
;
電信工程研究所
This paper presents a multiple objective genetic algorithm for determining the component parameters in the design of a microwave filter. In this approach, we first propose a set of fuzzy-based membership functions for modeling the five design specifications (objective functions) of the microwave filter design problem. The five objective functions are then aggregated into a single-valued fitness function. Finally, a set of near-optimal component parameters for the microwave filter can be found by applying the proposed genetic algorithm. A large number of experiments were performed and demonstrated satisfactory results. This approach has three principal merits. First, it counteracts entrapment in a local optimal solution. Second, the processing of multiple objectives is relatively easy. Third, the generated solution is independent of the initial estimate. Significance: This paper presents a multiple objective genetic algorithm for determining the component parameters in the design of a microwave filter.
A tool planning approach considering cycle time constraints and demand uncertainty
(2005-09-01)
Wu, MC
;
Hsiung, Y
;
Hsu, HM
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
The tool planning problem is to determine how many tools should be allocated to each tool group to meet some objectives. Recent studies aim to solve the problem for the cases of uncertain demand. Yet, most of them do not involve cycle time constraints. Cycle time, a key performance index in particular in semiconductor foundry, should not be ignored. The uncertain demand is modeled as a collection of scenarios. Each scenario, with an occurrence probability, represents the aggregate demand volume under a given product mix ratio. A genetic algorithm embedded with a queuing analysis is developed to solve the problem. Experiments indicate that the proposed solution outperforms that obtained by considering only a particular scenario.
Aggregation of fuzzy opinions under group decision making
(1996-05-13)
Hsu, HM
;
Chen, CT
;
交大名義發表
;
National Chiao Tung University
In this paper, a method is proposed for aggregating individual fuzzy opinions into a group fuzzy consensus opinion. This paper presents a procedure for aggregating the expert opinions. First, we define the index of consensus of each expert to the other experts using a similarity measure. Then, we aggregate the experts using the index of consensus and the importance of each expert. Finally, a numerical example is given to apply our model.
Dynamic programming for delayed product differentiation
(2004-07-01)
Hsu, HM
;
Wang, WP
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
Product proliferation and demand uncertainty result in material requirement planning difficulties, which give rise to high inventories and low customer service. A design concept, delayed product differentiation, is advocated for reducing the impact due to inaccurate forecasts and shortening the order response time. In this paper, a dynamic programming model using an AND/OR graph is constructed to determine the product differentiation points. Taking into account the costs and benefits associated with delayed product differentiation points, we proposed an approach to suggest whether the designer ought to differentiate specific products from the common part set at each design stage. Finally, we illustrate an example to characterize the optimal product differentiation points. (C) 2003 Elsevier B.V. All rights reserved.
Fuzzy credibility relation method for multiple criteria decision-making problems
(1997-01-01)
Hsu, HM
;
Chen, CT
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
This paper dears with the problem of ranking alternatives under multiple criteria. A fuzzy credibility relation (FCR) method is proposed. Owing to vague concepts represented in decision data, in this study the rating of each alternative and the weight of each criterion are expressed in fuzzy numbers. Then we define the concordance, discordance, and support indices. By aggregating the concordance index and support index, a fuzzy credibility relation is calculated to represent the intensity of the preferences of one alternative over another. Finally, according to the fuzzy credibility relation, the ranking order of all alternatives can be determined. A numerical example is solved to highlight the procedure of the FCR method at the end of this paper. (C) Elsevier Science Inc. 1997.
Fuzzy credibility relation method for multiple criteria decision-making problems
(1997-01-01)
Hsu, HM
;
Chen, CT
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
This paper dears with the problem of ranking alternatives under multiple criteria. A fuzzy credibility relation (FCR) method is proposed. Owing to vague concepts represented in decision data, in this study the rating of each alternative and the weight of each criterion are expressed in fuzzy numbers. Then we define the concordance, discordance, and support indices. By aggregating the concordance index and support index, a fuzzy credibility relation is calculated to represent the intensity of the preferences of one alternative over another. Finally, according to the fuzzy credibility relation, the ranking order of all alternatives can be determined. A numerical example is solved to highlight the procedure of the FCR method at the end of this paper. (C) Elsevier Science Inc. 1997.
Fuzzy logic control of aggregate production planning using genetic algorithms
(1999-12-01)
Hsu, HM
;
Lin, YJ
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
In general, a successful application of fuzzy logic control depends on fuzzy rules, the structure of the rules, and membership functions. This research presents a fuzzy logic control (FLC) model for aggregation production planning by using a genetic algorithm (GA) to simultaneously search the membership functions and inference rule set. It is shown that the genetic algorithm enables us to generate an optimal combination of the membership functions and the rule set. The approach presented in this research is illustrated by using Holt's paint factory data. The comparison of our results with those of Rink and Turksen shows that the proposed approach is the more favorable one. Significance: This research presents a fuzzy logic control (FLC) model for aggregation production planning by using a genetic algorithm (GA) to simultaneously search the membership functions and inference rule set.
Improving the RF performance of 0.18 mu m CMOS with deep n-well implantation
(2001-10-01)
Su, JG
;
Hsu, HM
;
Wong, SC
;
Chang, CY
;
Huang, TY
The radio-frequency (RF) figures of merit of 0.18 mum complementary metal-oxide-semiconductor (CMOS) technology are investigated by evaluating the unity-current-gain cutoff frequency (F-t) and maximum oscillation frequency (F-max). The device fabricated with an added deep n-well structure is shown to greatly enhance both the cutoff frequency and the maximum oscillation frequency, with negligible dc disturbance. Specifically, 18% increase in F-t and 25% increase in F-max are achieved. Since the deep n-well implant can be easily adopted in a standard CMOS process, the approach appears to be very promising for future CMOS RF applications.
Investigations of bulk dynamic threshold-voltage MOSFET with 65GHz "normal-mode" Ft and 220GHz "over-drive mode" Ft for RF applications
(2001)
Chang, CY
;
Su, JG
;
Hsu, HM
;
Wong, SC
;
Huang, TY
The RF properties of bulk dynamic threshold-voltage MOSFET (B-DTMOS) with a deep n-well isolation was investigated both under the normal DTMOS mode and two newly-proposed DTMOS operation modes: moderate (0.6V < Vgs = Vbs < 0.85V) and over-drive (Vgs = Vbs > 0.85V) modes. While Ft can be improved to 65GHz at 12.5mA with 1.5V Vds bias under normal-mode DTMOS operation, a high Ft of 220GHz with good linearity and stability is achieved under over-drive mode of operation.
Modeling demand driven with forecast regulation, safety stock setting and capacity analysis
(2000-09-01)
Hsu, HM
;
Wang, WP
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
High demand uncertainty has been a major challenge for production planning in assemble-to-order (ATO) environments. Regulating dealers' demand forecasts, determining appropriate safety stock levels and deciding required capacity levels are some of the typical problems solved by ATO firms. The main objective in this study is therefore to construct a mathematical model to deal with the aforementioned managerial, problems under the consideration of product life cycle. According to the various cost and demand characteristics in different phases of a product life cycle, production managers may apply the proposed model to find a near optimal solution set of regulation factors for dealers' demand forecasts, appropriate levels of safety stocks and the number of key machines. Finally, an example is given to illustrate our model. Significance: This paper presents a linear programming model that integrates the activities of forecasting adjustment, material preparation and key machine quantity determination in an ATO environment. The proposed model reflects the phenomena of a product life cycle passing through the different appearances of the demand and cost in distinct phases.
Possibilistic programming in production planning of assemble-to-order environments
(2001-04-01)
Hsu, HM
;
Wang, WP
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
For the sake of alleviating the influences of demand uncertainty in assemble-to-order (ATO) environments, the strategies of regulating dealers' forecast demands, determining appropriate safety stocks, and deciding the numbers of key machines are usually adopted by manufacturers. In this paper, we propose a possibility linear programming model to manage these production planning problems. The proposed model accomplishes forecasting adjustments, material management, and production activities. Because of price fluctuations, material obsolescence, and the time value of, capital. the ambiguity of cost is considered in the objective function of the model. We substituted the fuzzy objective function with three crisp objectives: minimizing the most possible cost, maximizing the possibility of obtaining lower cost, and minimizing the risk of obtaining higher cost. Zimmermann's fuzzy programming method is then applied for achieving an overall satisfactory compromise solution. Finally, an example is given to illustrate our model. (C) 2001 Elsevier Science B.V. All rights reserved.
Scrap rules for small lots in wafer fabrication
(2002)
Wu, MC
;
Chiou, CW
;
Hsu, HM
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
This paper presents a genetic algorithm for determining the-rules for scrapping small lots in a semiconductor fab. Small lots are lots which carry less than 25 wafers due to yield problem. Simulation results show that the proposed approach can increase contribution margin up to 24% in the case of low yield and high average, selling price.
Scrapping small lots in a low-yield and high-price scenario
(2004-02-01)
Wu, MC
;
Chiou, CW
;
Hsu, HM
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
Some wafers in a lot may become spoiled after they are processed at a workstation; such a lot is called a small lot. In a low yield and high price scenario, scrapping small lots may increase revenue and profit; yet, this notion has seldom been examined. This study presents a model for formulating the decision-making problem of scrapping small lots. A genetic algorithm is used to solve the problem when the solution space is large. An exhaustive search method is used when the solution space is small. Some numerical examples are used to evaluate the outcome of scrapping small lots. The profit obtained by the proposed scrapping method may be up to 23% higher than that obtained without scrapping.
Tool planning in the scenario of multiple existing semiconductor fabs
(2005-11-01)
Wu, MC
;
Hsiung, Y
;
Hsu, HM
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
The tool planning problem involves determining how many tools should be allocated to each tool group to achieve a set of objectives. Most previous studies assume that a demand mix has been given for a new factory to be planned. However, when a semiconductor company has several existing fabs, the demand mix for the new fab is not explicitly known, and needs to be allocated from the demand mix of the whole company. This paper presents an integrated approach to determine the optimal demand mix and associated tool plan for the new fab that can minimize the tool cost of the new fab while each fab (new or existing) is requested to meet a predefined target in its mean cycle time. Simulation experiments indicate that the proposed solution is better than that obtained by a heuristic method used in industry. The saving in tool cost for a typical tool planning problem can be over US$ 70 million.