Evaluation of weigh-in-motion accuracy by simulation
Journal of the South African Institution of Civil Engineering
BKS, South Africa; BKS (Pty) Ltd., PO Box 3173, Pretoria, 0001, South Africa
A method for the evaluation of accuracy of weigh-in-motion (WIM) scales based on Monte Carlo simulation Is proposed. The objective of the method is to assist with decision on acceptance of the tested WIM scale as sufficiently accurate, or its rejection. Conceptually, the method considers a 'nebula' of possible populations that could supply the observed sample of n weighing errors. Excessive WIM weighing error is defined as an error falling outside a specified acceptance region. The percentage of excessive errors, Q, is then determined in each of the above populations. The distribution of Q Is constructed and the probability of Q exceeding a tolerable limit Qcrit is determined from this distribution. Verdict on acceptance or rejection is formulated in terms of the above probability and displayed graphically. Also graphically displayed are risks of wrong acceptance and wrong rejection. This allows the user to make decisions with a full view of consequences. The method is simple to use and its applications Indicate that, when using conventional WIM-accuracy testing procedures, one may under-estimate the risk of wrong decision. The objective of the paper is to Introduce and explain the principle of a method for correct evaluation of accuracy of weigh-in-motion scales. The purpose of the method is to assist with decisions on acceptance of the tested scale as sufficiently accurate, or its rejection.
Weighing error; WIM accuracy; Computer simulation; Decision making; Monte Carlo methods; Probability; Risk analysis; Building codes