Model formulas are the standard approach for specifying the variables in statistical models in the S language. Although being eminently useful in an extremely wide class of applications, they have certain limitations including being confined to single responses and not providing convenient support for processing formulas with multiple parts. The latter is relevant for models with two or more sets of variable, e.g., regressors/instruments in instrumental variable regressions, two-part models such as hurdle models, or alternative-specific and individual-specific variables in choice models among many others. The R package Formula addresses these two problems by providing a new class "Formula" (inheriting from "formula") that accepts an additional formula operator | separating multiple parts and by allowing all formula operators (including the new |) on the left-hand side to support multiple responses. / Series: Research Report Series / Department of Statistics and Mathematics

It is the extreme essential that the functions of the modernizing pharmacy stores integrate into the entire community and medical care system. Due to the environment alternation and the constant epoch evolution, however, pharmacy stores have not only been impacted by the national health insurance and the medicine profession categorization, but also affected by the logistic development and the business modernization. Under the circumstances, pharmacy stores have to confront transformation, and pursue the much better solutions in assuring the persistent business management in future. This research focuses on pharmacy stores around the middle of Taiwan, and proceeds with the questionnaire investigations without signature, aiming at four hundreds of their consumers in Taichung, Nantao, Chang Hua and adjacent districts. The contents of questionnaire include four sections: 1) The comparison of both modernizing and traditional pharmacies; 2) Personal daily custom of healthy life style; 3) Perceiving of import infant Milk powder serving by pharmacy stores; 4) Individual base information. Based on the analysis of the questionnaire investigation findings, the purport is to realize whether or not the purchasing custom of ordinary consumers would be changed in the existence of import infant milk powder serving by the modern pharmacy chain stores. This research discovered that: Since infant milk powder is the sole principal food of baby, parents are very much concerned about it. The manufacturers have the respective features for milk powder of their own brands, and babies have quite different physiques. The certain brand of infant milk powder might be suitable for some babies, but against another ones on the other hands. For such instances, more than 80% of consumers are in favor of milk powder serving by pharmacy stores, according to the findings of the research, and agree that they would go there for consults if they need to make a decision about milk powder. Besides, the more often milk powder they purchased from pharmacy stores, the more frequency that they interacted with medicine chemists.

The current bridge formula, Federal Bridge Formula B (BFB), established in 1974 to protect bridges against excessive overstress, is very restrictive on long combination vehicles due to an 80,000 lb gross vehicle weight limit. Without this limit the formula will not be able to protect bridges in the cases of longer trucks. A formula developed by the Texas Transportation Institute (T.T.I.) called the TTI-HS20 Formula addresses these issues. This formula, developed especially for bridges designed for the HS-20 truck, eliminates the need for the 80,000 lb limit. A generic formula developed to protect H15 and HS-20 bridges (James et al., 1986) was evaluated in a previous study (James and Zhang, 1991). The approach to evaluating the TTI-HS20 Formula follows the approach outlined in James and Zhang, 1991. Information was collected on two important elements: a set of test bridges representative of the lightest continuous bridges, and a set of test truck configurations representative of real truck traffic with a focus on long combination vehicles. Critical weights of the selected trucks for the representative bridges are calculated and plotted against the TTI-HS 20 formula and other proposed formulas. A final recommendation as to whether this formula should be adopted nationwide is made.

The Texas A&M University Formula SAE program currently has no rigorous method for analyzing or predicting the overall dynamic behavior of the student-designed racecars. The objective of this study is to fulfill this need by creating a full vehicle ADAMS/Car model incorporating an empirical tire-road force model and validating the longitudinal performance of the model by using vehicle responses recorded at the track. Creating the model requires measuring mass and inertia properties for each part, measuring the locations of all the kinematic joints, testing the Risse Racing Jupiter-5 shocks to characterize damping and stiffness, measuring engine torque, and modeling the tire behavior. Measuring the vehicle performance requires installation of the Pi Research DataBuddy data acquisition system and appropriate sensors. The 2002 Texas A&M University Formula SAE racecar, the subject vehicle, was selected because it already included some accommodations for sensors and is almost identical in layout to the available ADAMS/Car model Formula SAE templates. The tire-road interface is described by the Pacejka ??94 handling force model within ADAMS/Car that is based on a set of Goodyear coefficients. The majority of the error in the model originated from the Goodyear tire model and the 2004 engine torque map. The testing used Hoosier tires and the 2002 engine intake and exhaust configuration. The deliverable is a full vehicle model of the 2002 racecar with a 2004 engine torque map and a tire model correlated to longitudinal performance recorded at the track using the installed data acquisition system. The results of the correlation process, confirmed by driver impressions and performance of the 2004 racecar, show that the 2004 engine torque map predicts higher performance than the measured response with the 2002 engine. The Hoosier tire on the Texas A&M University Riverside Campus track surface produces 75??3% of peak longitudinal tire performance predicted by the Goodyear tire model combined with a road surface friction coefficient of 1.0. The ADAMS/Car model can now support the design process as an analysis tool for full vehicle dynamics and with continued refinement, will be able to accurately predict behavior throughout a complete autocross course.

Thesis (Ph. D.)--University of Queensland, 2002. / Includes bibliographical references.

農林水産研究情報センターで作成したPDFファイルを使用している。

Thesis (doctoral)--University of Stockholm, 1984. / Abstract (1 leaf) inserted. Bibliography: p. 79-80.

Thesis (Ed.M.)--Boston University

This thesis studies how to eliminate syntactic complexity in Lμ, the modal μ calculus. Lμ is a verification logic in which a least fixpoint operator μ, and its dual v, add recursion to a simple modal logic. The number of alternations between μ and v is a measure of complexity called the formula’s index: the lower the index, the easier a formula is to model-check. The central question of this thesis is a long standing one, the Lμ index problem: given a formula, what is the least index of any equivalent formula, that is to say, its semantic index? I take a syntactic approach, focused on simplifying formulas. The core decidability results are (i) alternative, syntax-focused decidability proofs for ML and Pμ 1 , the low complexity classes of μ; and (ii) a proof that Ʃμ 2 , the fragment of Lμ with one alternation, is decidable for formulas in the dual class Pμ 2 . Beyond its algorithmic contributions, this thesis aims to deepen our understanding of the index problem and the tools at our disposal. I study disjunctive form and related syntactic restrictions, and how they affect the index problem. The main technical results are that the transformation into disjunctive form preserves Pμ 2 -indices but not μ 2 -indices, and that some properties of binary trees are expressible with a lower index using disjunctive formulas than non-deterministic automata. The latter is part of a thorough account of how the Lμ index problem and the Rabin–Mostowski index problem for parity automata are related. In the final part of the thesis, I revisit the relationship between the index problem and parity games. The syntactic index of a formula is an upper bound on the descriptive complexity of its model-checking parity games. I show that the semantic index of a formula Ψ is bounded above by the descriptive complexity of the model-checking games for Ψ. I then study whether this bound is strict: if a formula Ψ is equivalent to a formula in an alternation class C, does a formula of C suffice to describe the winning regions of the model-checking games of Ψ? I prove that this is the case for ML, Pμ 1 , Ʃμ 2 , and the disjunctive fragment of any alternation class. I discuss the practical implications of these results and propose a uniform approach to the index problem, which subsumes the previously described decision procedures for low alternation classes. In brief, this thesis can be read as a guide on how to approach a seemingly complex Lμ formula. Along the way it studies what makes this such a difficult problem and proposes novel approaches to both simplifying individual formulas and deciding further fragments of the alternation hierarchy.

The problem with designing and simulating a race car entirely through CAD and other computer simulations, is that the real world behavior of the car will differ from the results outputted from CFD and FEA analysis. One way to learn more about how the car actually handles, is through telemetry and data logging of many different sensors on the car while it is running at racing speeds. This data can help the engineering team build new components, and tune the many different systems on the car in order to get the fastest time around a track as possible.