Mine ventilation system planning using genetic algorithms

The most common problem in contemporary mining practice related to planing and analysis of ventilation systems is the optimization of partially regulated air distribution in mine Ventilation networks. This paper presents a two step optimization procedure for partly regulated air distribution in mine ventilation networks. The first step is the determination of air distribution in all branches using the well known Hardy-Cross method. In the second step the distribution and parameters of air flow regulators with minimum engaged power in the system are determined by a genetic algorithm. The optimization procedure in genetic algorithms is a loop through a "reproductive" cycle: selection, crossover and mutation, until a given condition is fulfilled. The proposed methodology which combines the Hardy-Cross method with a genetic algorithm gives ventilation engineers the opportunity to select the best solution out of a number of alternatives, under the limitations imposed by contemporary engineering practice. The main features of the implemented method are discussed and illustrated by a case study.