Development and Implementation of Mine Ventilation Network Calibration Using a Two-step Method

Calibrating an entire mine ventilation network model from survey data has long been considered a challenging and sometimes impractical task. This paper proposes a solution to this issue with the development of a quadratic programming optimization model. The optimization model takes the minimum modification of input resistances as the objective function, uses the mesh pressure equations as equality constraints, and incorporates the lower limits of airway resistances as inequality constraints. A two-step method is developed to solve the model, which includes finding an initial feasible solution and iteratively solving the final solution using the active set method. A computer program has been developed to implement the optimization model. To demonstrate its practicality, a study case using a 35-branch ventilation network is presented, which indicates that the solution process converges after only 57 iterations, a significant improvement over the times of trial and error, which could reach 3.4 x 1010 for the worst-case scenario. Additionally, it is shown that considering only the equality constraints results in negative resistances, while incorporating the lower limits of resistance ensures that the corrected resistances are within a reasonable range. Using the resistance corrected by the optimal model to simulate the ventilation system via traditional methods generates quantities and fan pressures that perfectly match the measured values, eliminating the tedious process of manually adjusting airway resistances repeatedly.