There are two key issues in assuring the accuracy of estimates of performance obtained from a simulation model.  The first is the removal of any initialisation bias, the second is ensuring that enough output data is produced to obtain an accurate estimate of performance.  This part of the AutoSimOA project is concerned with the first issue, and more specifically warm-up estimation.  Our aim is to produce an automated procedure, for inclusion into commercial simulation software, for estimating the length of warm-up and hence removing initialisation bias from simulation output data.

The two papers below describe the work that has been carried out to date with the aim of producing an automated procedure to estimate the warm-up period.  Part 1 describes the extensive literature review that was carried out to find the various warm-up methods in existence.  It explains how we short listed candidate methods for further testing and details the testing procedure, including the creation of artificial data sets and performance criteria.  It contains the results of the pre-liminary testing of shortlisted methods and concludes that only one method, MSER-5, was found to be a robust and viable candidate for automation.  Part 2 describes the MSER-5 method and sets out the results of further testing of this method. 

Please click on the following links to read these two accounts of our research:   

PART 1: Review of methods and shortlisting.   

PART 2: MSER-5 - A full analysis.

Fuller details and a review of the 44 warm-up methods found in the literature can be viewed by clicking here.