Application of Genetic Algorithms for Reliability Assessment of Two Mine Hoisting Systems

This paper discusses the application of a computerised model based on genetic algorithms (GAs), called GenRel, for reliability assessment of underground mine hoisting systems. The purpose of this paper is to investigate whether GenRel can be applied to predict future failure data of a mine hoisting system based on historical records of failures. The incentive of selecting the GAs for prediction of failures is that GAs are a class of evolutionary algorithms which imitate the biological evolution procedures such as reproduction, selection, crossover and mutation. The reliability of mining equipment changes over time due to an array of factors (eg equipment age, the operating environment, number and quality of repair). These factors affect the equipment's failure patterns and have complex impacts on the equipment's reliability characteristics. The failure patterns are assumed to follow a biological evolution process, and thus GAs can be considered applicable in the modelling process.To conduct the reliability assessment, a study was carried out with historical failure data of hoists in the time period from January to December 2007. Two failure data sets of two mine hoisting systems were collected from two typical underground mines in Ontario, Canada, which are denoted as the NA mine and the SB mine. The failure data sets were prepared in the format of time between failures (TBF). Then, these data sets were entered in GenRel to generate predicted failure data sets for the period of January to March 2008. The paper discusses the statistical similarity of the actual failure data sets for the period of January to March 2008 with the predicted failure data set generated by GenRel in the same time period.