Higher Moments of Weighted Integrals of Non-Gaussian Fields

Gunnar Mohr

    Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

    Abstract

    In general , the exact probability distribution of a definite non-Gaussian random field is not known. Some information about this unknown distribution can be obtained from the 3rd and 4th moment of the integral. Approximations to these moments are calculated by a numerical technique based on recursive application of Winterstein approximations) moment fitted linear combinations of Hermite Polynomials of standard Gaussian variables). By use of computerized symbol manipulations it is practicable to obtain exact moments (up to order 4) of partial weighted sums of mutually dependent variables with known moments (including mixed moments) as fx log-normal variables or polynomials of standard Gaussian variables. This sceme is used when calculating the moments of the integral (and eventually approximating the integral by a Winterstein approximation).
    Original languageEnglish
    Title of host publicationProbabilistic mechanics & structural reliability : Proceedings of the Seventh Specialty Conference
    Place of Publicationnew york
    PublisherAmerican Society of Civil Engineers
    Publication date1996
    Pages254-257
    ISBN (Print)0-7844-0184-5
    Publication statusPublished - 1996
    Event7th Specialty Conference on Probabilistic Mechanics and Structural Reliability 1996 - Worchester, United States
    Duration: 7 Aug 19969 Aug 1996

    Conference

    Conference7th Specialty Conference on Probabilistic Mechanics and Structural Reliability 1996
    Country/TerritoryUnited States
    CityWorchester
    Period07/08/199609/08/1996

    Cite this