A5058907026- Advanced Statistical Process Monitoring
- Scientific Measurement and Uncertainty Evaluation
- Fault Detection and Control Systems
- Advanced Statistical Methods and Models
- Industrial Vision Systems and Defect Detection
- Quality and Safety in Healthcare
- Electronic Packaging and Soldering Technologies
- Optimal Experimental Design Methods
- Integrated Circuits and Semiconductor Failure Analysis
- Neural Networks and Applications
- Facility Location and Emergency Management
- Advanced Neural Network Applications
- Risk and Safety Analysis
- Advanced Battery Technologies Research
- Spectroscopy and Chemometric Analyses
- Machine Learning in Materials Science
- Metal Forming Simulation Techniques
- Multi-Criteria Decision Making
- Supply Chain Resilience and Risk Management
- Reproductive Health and Contraception
- Reliability and Maintenance Optimization
- Advanced Control Systems Optimization
- Electrostatic Discharge in Electronics
- Outsourcing and Supply Chain Management
- Robot Manipulation and Learning
National Yunlin University of Science and Technology
2007-2023
Abstract In this study, a control chart is constructed to monitor multivariate Poisson count data, called the MP chart. The limits of are developed by an exact probability method based on sum defects or non conformities for each quality characteristic. Numerical examples used illustrate evaluated average run length (ARL) in simulation. result indicates that more appropriate than Shewhart-type when correlation between variables exists. Keywords: Attribute chartAverage lengthMultivariate...
The implementation of the National Health Insurance in Taiwan since 1995 has resulted Taiwanese residents having better medical security and care than before; however, it also increased demand for services. Thus, hospitals clinics have seen great growth recent decades. One problem is that program cannot fully support all clinics. This led to fierce competition between them. In such an environment, selecting optimal location can increase sectors' competitive advantage. serious aging...
Abstract The one-sided cumulative count of conforming (CCC) chart is a useful method to monitor nonconforming fraction in high-quality manufacturing processes. parameter assumed be known when implementing CCC chart. In this study, we investigated the impact estimated fraction, [pcirc] 0, run length distribution derived as well conditional probability false alarm rate (CFAR), average (CARL) and its standard deviation (CSDRL). Simulation results are conducted evaluate effect 0 show that values...
Many multivariate quality control techniques are used for variable processes, but few work attribute processes. To monitor attributes, controlling the false alarms (type I errors) and considering correlation between attributes two important issues. By taking into account these issues, a new chart is presented to bivariate binomial process. An example illustrated proposed method. evaluate performance of method, simulation study conducted compare results with those using both np skewness...
The wire bonding process is one of the most critical processes in semiconductor packaging. electrical performance and reliability IC chips must be evaluated by probe testing before bonding. In test, usually leaves traces on surface pad. Large marks tend to cause a decrease bond adhesion so increase possibility ball lifting from pads. Shear force an important measure between prediction shear helps understand quality advance. this study, six features were extracted automatic image recognition...
Due to the development of high-quality manufacturing processes, a cumulative count conforming (CCC) chart has been demonstrated be useful in processes for monitoring nonconforming fraction. When quality characteristic is asymmetric and one-directional shifting interest, one-sided control more appropriate than two-sided chart. We construct CCC based on previous runs improve its sensitivity. The results show that proposed conditional procedure method can detect shift quickly when process...
C-charts and U-charts are widely applied for Poisson distribution to monitor the number of defects in univariate process. More than one quality characteristic attribute data with is interest many industrial applications. A control chart constructed monitoring multivariate count data, called MP chart. The based on sum or nonconformities single In this study, limits process estimated by using multiple linear regression model since calculation complicated. Two examples provided illustrate...
ABSTRACT In this paper, a tabu search for solving the redundancy allocation problem (RAP) in multi-states (different performance levels) series-parallel systems (MSPS) is proposed. The modeled to find optimal system configuration that minimizes cost of component selection and level under given minimum required reliability constraints. proposed solution methodology offers several benefits compared with other approaches. computational results demonstrate advantages approach type problem.
In the medical, small shifts or changes that occur can to affect final results. Since each patient is different, they pose different risks will doctor's performance during surgery. This study uses Logistic model and Cox compare performance. From simulation Analysis method, we found quickly find patient's abnormal conditions. Thus, using this method doctors make earlier detection of patients with
ABSTRACT The most popular control charts in the multivariate statistical process are Hotelling T 2 for mean and | S chart variability. In this research, we will propose an alternative which combining information both chart. We construct a single to monitor variability simultaneously. average run length be used evaluate performance of new compared with
There are many cases showing that conventional pesticide residues in crops growing incidence, consumers gradually turn to accept the organic crops, and driven by booming healthy retail chain stores. This study is based on example of an store, analysis concept "trust" object supply between suppliers or partners; furthermore, explore impact core values - trust, Cooperation Appreciation innovation management model performance. Through literature review expert questionnaire identify key...
When processing multiple quality characteristics out of control, control charts are not effective tools for root cause analysis, but chart patterns can provide hints on assignable cause, and in this research, we use a fuzzy inference system based Hotelling's T <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> to accelerate analysis.
統計製程管制方法大多數是設計給常態製程資料，然而生產製程中，非常態製程也常出現，而統計學習方法(statistical learning techniques)中的支撐向量機為近幾年來新興的一項工具，可以把製程分類成失控或是控制中以監控製程，其中內核距離管制圖(k-chart)已被廣泛的運用於監控非常態製程。本研究探討k管制圖在監控非常態製程中的表現與特性，針對非常態資料中的對稱資料二元t分配與偏態的伽瑪分布資料，考慮不同的偏移型態及不同程度的相關係數，衡量k管制圖的績效，研究結果顯示k管制圖能有效監控中偏移至大偏移的情形。
Abstract Wire bonding is key to high-quality IC assembly in semiconductor packaging. An electrical test was conducted before wire bonding. During the test, needle probes touched surface of each bond pad resulting a probe trace left, called marks. The mark on affects adhesion between and copper ball. Generally, ball can be removed easily under shear testing if ratio area greater than 20%. A large causes defect yield. However, rejecting orders wafer batches with marks will result loss market...