From GridLAB-D Wiki
Jump to: navigation, search

Mean-time to restoration

Please add an MTTR parameter as well. In addition, I recommend that we document how to convert MTBR and MTTR to the appropriate distribution parameters, e.g., exponential, pareto, etc. --Dchassin 21:48, 13 May 2010 (UTC).

  • Another way of probably stating the above, be sure to specify which distribution will be used where and how their sufficient statistics are determined, or if this will be another user specifiable option. -Ftuffner 20:46, 27 May 2010 (UTC)

Equations for metrics

  • Please provide the required formulas for the metrics to be calculated and how the input parameters are to be gathered and accrued. --Dchassin 21:48, 13 May 2010 (UTC).
  • There was offline discussion about possibly defining the interval to calculate the metrics over (e.g. 1 month, 1 year, etc.). Is this functionality still required and/or necessary? -Ftuffner 17:34, 1 June 2010 (UTC)

Distributions

I would express the full equations for the PDF on the MTBF and MTTR distributions. As part of this process, I'd adjust the notation as well. P(x) implies the PDF value at x. The way it is written, they look like variations of a uniform distribution, not the means as it indicates. Regarding the distributions, I am confused by the mean value for the exponential. Is there a particular reason it is calculated in this "non-standard" manner? In regards to the Pareto distribution, this distribution is typically defined by two parameters. How does the definition of the mean translate into those two parameters? -Ftuffner 17:34, 1 June 2010 (UTC)