Multivariate input modeling with Johnson distributions

Stanfield, Paul; Wilson, James R.; Mirka, Gary A.; Glasscock, Naomi F.; Psihogios, Jennie P.; Davis, Joseph R.

doi:10.1145/256562.256990

Cited by 25 publications

(14 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Estimation of marginal distributions can be based on their moments [33] and the dependency expressed by the correlation matrix of the underlying MVN distribution. Monte Carlo simulation can be performed by simulating from the underlying MVN distribution and applying the inverse translation, which attains the correct marginals [34].…”

Section: Joint Distributionsmentioning

confidence: 99%

Handling Input Correlations in Pharmacoeconomic Models

Naveršnik¹,

Rojnik²

2012

Value in Health

View full text Add to dashboard Cite

show abstract

Section: Joint Distributionsmentioning

confidence: 99%

Handling Input Correlations in Pharmacoeconomic Models

Naveršnik¹,

Rojnik²

2012

Value in Health

View full text Add to dashboard Cite

show abstract

“…This family of distribution was found best to fit the marginal distribution of multivariate as well as the correlation between them. [5] The procedures that are outlined in the same reference have also been followed here.…”

Section: Fig 7: Correlation Matrix Of the Measured User Loadsmentioning

confidence: 99%

Generalized method for characterizing mechanical loads in mobile devices

Lumbantobing

Tiku

2012

2012 IEEE 62nd Electronic Components and Technology Conference

View full text Add to dashboard Cite

This paper presents the concept and design of a test vehicle for characterizing applied mechanical loads on small form-factor devices. These loads can include forces, moments, and torques. A case-study utilizing a version of this test vehicle is discussed. The test vehicle weighs 100 grams and has a form factor similar to commonly available media players like Zune HD. Specific usage scenarios were selected to represent extreme loading conditions. The resulting multivariate, stochastic data were analyzed using the Bounded Johnson System distribution. This statistical treatment eases the use of the empirical data in subsequent analysis, such as determination of the appropriate in-lab loading conditions to quantify the mechanical robustness of a product. IntroductionRecent proliferation of mobile consumer electronic devices has led to a need for more efficient methods to better predict the reliability of these devices in various use scenarios.[1] These scenarios range from transporting these devices in pockets, back-packs, and purses; to leaving them in parked vehicles or other non-climate-controlled spaces. Some of these use cases were not seen by consumer products in the past.Miniaturization, growing device complexity, and reduction in development time due to time-to-market pressures exacerbate this need.The first step in predicting reliability is to accurately characterize the expected mechanical and other stress profiles for the product during its use life. The ability to quickly and accurately quantify the design margins for a product based on this knowledge minimizes the risk of costly product returns and the bill-of-material costs due to unnecessary over-design. This also allows the opportunity for concurrent evaluation of multiple device architectures while evaluating trade-off scenarios for reliability, manufacturability, industrial design, etc. This consequently can reduce the number of unnecessary design compromises to be made during development.To evaluate reliability at both component and board levels, mechanical stresses are an essential input. [2-4] These stresses may be obtained from crude estimations, historical data from similar products, or evaluating the actual component or board placed inside a product prototype. The issue with the first two approaches is the accuracy of the estimates; and the third approach constraints the estimates around a specific design and requires a higher level of design maturity. This paper presents a generalized methodology for characterizing mechanical loads that a handheld device may experience while in use. It can be used very early in the development phase to support high quality decisions on competing design options.

show abstract

“…They focus on the important, but relatively simple, task of fitting classical univariate distributions to data and incorporating the fitted distributions into simulation software. The last few years have produced some easily attainable noncommercial software and ideas for creating more-complex input models, including Avramidis and Wilson (1994), Cario (1996), Cario andNelson (1997a, 1997b), Chen (1999), Wilson (1997, 1998), Kuhl, Wilson and Johnson (1997), Song and Hsiao (1993), Song, Hsiao and Chen (1995), Stanfield et al (1996), and Wagner and Wilson (1995, 1996a, 1996b. These (and some other ideas) are discussed in Nelson and Yamnitsky (1998).…”

Section: The Logical and Input Modelsmentioning

confidence: 99%

“…The early approach to creating flexible multivariate models was to assume a particular marginal distribution, as done, for example, by Lal (1980, 1982) and Lewis and colleagues (e.g., Lewis and Orav 1989). Wagner and Wilson (1995) discuss a generalization having Bézier marginal distributions and Stanfield et al (1996) discuss a generalization having Johnson marginals.…”

Section: Modeling Dependencymentioning

confidence: 99%