Genetic programming for prediction of heat stress hazard in underground coal mine environment

Working against nature and an uncertain environment makes underground mining a hazardous profession. Every year hundreds of miners lose their valuable lives due to mine hazards. Increasing demand for coal necessitates the extraction of coal at a higher rate. As a result, easily minable shallow coal deposits are depleting speedily, and in near future, deep-seated deposits will be left for mining by underground methods. With rising depth and deployment of high-capacity machines, increasing heat stress becomes a major hazard in the underground mine environment posing threat to the miners’ health, productivity and safety. Ignoring the effect of heat stress may lead to dangerous circumstances, even result in death. To avoid such unwanted event, it has become imperative to predict the heat stress to reduce its adverse impact in underground coal mines. Therefore, in this study a detailed field survey is conducted to collect the environmental data of three underground coal mines. Genetic programming (GP) is done to develop a relation between the environmental parameters and heat stress, by taking the mine survey data as input. A good correlation coefficient (R = 0.9816) is obtained between the GP predicted heat stress and actually measured heat stress, which indicates that GP can be effectively used to predict the heat stress in underground mines. A sensitivity analysis (SA) is done to determine the effect of input parameters on heat stress. The SA results revealed that all six input parameters have a considerable effect on the heat stress; however, the dry-bulb temperature has the highest effect (0.98) on heat stress.