The warm-up period is studied because it affects the accuracy of the results in simulation of communication systems. During the warm-up period-when the simulation system has not reached the steady-state situation-, the system results (eg blocking probability) vary very rapidly from zero to 0.037 for the parameters used. In the batch-means method with 10,000 calls per batch and for 50 batches (500,000 calls), the CDMA blocking probability is 0.0192 with 99% confidence interval.
http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=5730377&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel5%2F5729958%2F5730309%2F05730377.pdf%3Farnumber%3D5730377
This paper reviews the performance of various methods used to detect the warm up length in steady state discrete event simulation. An evaluation procedure is used to compare the methods. The methods are applied to the output generated by a simple job shop model. The performance of the methods is tested at different levels of utilizations. Various measures of goodness are used to assess the effectiveness of the methods.
EVALUATION OF METHODS USED TO DETECT WARM-UP PERIOD IN STEADY STATE SIMULATION
One means for dealing with initialization bias in simulation experiments is to implement a warm-up period. This requires the correct estimation of the initial transient. A new method for determining the warm-up period, based upon the principles of statistical process control (SPC), is described. The method is tested on empirical data from a simulation model that has been used in a real-life study. In comparing the results to those from two commonly used warm-up methods, it appears that the SPC method performs well. The strengths and weaknesses of the approach are discussed.
A STATISTICAL PROCESS CONTROL APPROACH FOR ESTIMATING THE WARM-UP PERIOD
The problem of the initial transient has been discussed in the simulation literature over the past 40 years. Various approaches have been suggested for dealing with it. Deletion of the initial transient data by specifying a warm-up period seems to be the most common and favoured approach. But how can the length of the warm-up period be estimated? A review of methods for determining the warm-up period is provided. The methods are classified under five headings: graphical methods, heuristic approaches, statistical methods, initialisation bias tests and hybrid methods. A critique of the methods is provided with respect to their simplicity, ease of implementation, accuracy, dependence on assumptions and need to estimate parameters. It is concluded that no single method can be recommended and that further research is required particularly in testing and comparing the proposed methods on empirical data.
The Problem of the Initial Transient: Techniques for Estimating the Warm-up Period for Discrete-Event Simulation Models
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 paper is concerned with the first issue, and more specifically warmup estimation. A continuing research project is described that aims 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.
AUTOMATING WARM-UP LENGTH ESTIMATION
During the designated warm-up period, the meet management committee shall be responsible for ensuring that all Risk Management/Warm-up procedures are followed. It is incumbent on all participants, i.e. meet management, coaches, swimmers, and officials to comply with these procedures. Non-compliance with these procedures will result in a warning or removal of the swimmer from the warm-up pool, and possible further reporting to meet management and the Referee for action under rule SW 2.1.1. Meet management shall designate Safety Marshals to actively monitor the warm-up period in the competition pool and ensure compliance with these procedures. Safety Marshals should be notified of any disabled swimmers who are participating in warm-ups. Additional “practice” pools, if any, are to be supervised by lifeguards and coaches. “Safety Marshals shall be responsible to the Referee for all aspects of safety related to conduct during the warm-up period and shall ensure that all appropriate warm-up procedures are followed.” (CSW 2.13.1.5)
SWIMMING/NATATION CANADA RISK MANAGEMENT / WARM-UP PROCEDURES 2009
