Applications VI: Analysis of Random Data

  • John D. Robson
  • Colin J. Dodds
  • Donald B. Macvean
  • Vincent R. Paling
Part of the International Centre for Mechanical Sciences book series (CISM, volume 115)


Any attempt to enumerate the reasons for collecting data from random processes would necessarily be incomplete. Often conclusions can be drawn without any involved analysis; an experienced motorist can usually tell by “aural spectrum analysis” when the tappet clearance in his engine has increased just a few thousandth of an inch. Although a trained human operator has been a successful way of keeping complex sets of machinery under surveillance, we are now faced with the problems of fully automated engine rooms (for the understandable reason that trained operators are expensive items). Data on oil flows, steam pressures, bearing vibrations etc. have to be used to make decisions automatically. In some cases the penalty for a wrong decision is quite high, and then it is important to be sure as possible what valid conclusions can be drawn from the data. At this point the statistical nature of data analysis becomes evident. The reduced data based on an observation over a finite time T, i.e. an experiment, contains a component that could vary from sample to sample of length T, the remainder being what one really wants to know. For this reason it has been said that the major function of statistical procedures is to show us what we do not know from an experiment, not to tell us what we do know. This is far from being an admission of sterility, it serves as a warning about the misuse of random data.


Power Spectrum Amplitude Distribution Random Data Random Vibration Engine Room 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Bibliography for Chapter 17

  1. 17.1
    Harris B.(Editor): Spectral Analysis of Time Series. John Wiley & Sons Inc., 1967.Google Scholar
  2. 17.2
    Robinson E.A.: Multichannel Time Series Analysis with Digital Computer Programs. Holden-Day, San Francisco, 1967.MATHGoogle Scholar
  3. 17.3
    Cramér H.: Mathematical Methods of Statistics. Princeton University Press, Princeton 1961.Google Scholar
  4. 17.4
    Middleton D.: An Introduction to Statistical Communication Theory. McGraw-Hill, New York, 1960.Google Scholar
  5. 17.5
    McFadden J.A.: Probability Density of Output of a Filter when the Input is a Random Telegraphic Signal. IRE trans. Information Theory, 1959.Google Scholar
  6. 17.6
    Birnbaum Z.W.: Numerical Tabulation of the Distribution of Kolmogorov’s Statistic. J. Amer. Stat. Assocn., Vol. 47, p. 425, 1952.CrossRefMATHMathSciNetGoogle Scholar
  7. 17.7
    Baburin V.M.: Calculation of the Distribution Function for Random Processes from Experimental Data. Automation and Remote Control, Vol. 23, p. 519, 1962.MATHGoogle Scholar
  8. 17.8
    Fisher R.A.: Statistical Methods for Research Workers. Oliver and Boyd, Edinburgh, 1954.Google Scholar
  9. 17.9
    Bartlett M.S.: Periodogram Analysis of Continuous Spectra. Biometrica, Vol. 31, pp. 1–16, 1950.MathSciNetGoogle Scholar
  10. 17.10
    Bendat J.S. and Piersol A.G.: Analysis and Measurement Procedures. John Wiley and Sons, New York, 1971.MATHGoogle Scholar

Copyright information

© Springer-Verlag Wien 1971

Authors and Affiliations

  • John D. Robson
    • 1
    • 2
  • Colin J. Dodds
    • 1
  • Donald B. Macvean
    • 1
  • Vincent R. Paling
    • 1
  1. 1.University of GlasgowUK
  2. 2.UdineItaly

Personalised recommendations