Study On Mine Ventilation Resistance Coefficient Inversion

In the early 70’s, the research on the theory and algorithm of mine ventilation network solution has been mature, but today the ventilation network solution is still not widely used in the actual application of mine. The ventilation resistance coefficient "ventilation resistance coefficient is of uncertainty and still in the overall change, that’s one of the 3 major bottleneck problems with ventilation network solution. The problem is not resolved yet, hindering the application of ventilation network solution in practice. The accurate determination of ventilation resistance factors of ventilation system simulation is more and more cannot be ignored. Both empirical formula calculation and field test have respective limitations. Empirical formulas are usually only approximate summarized for some special circumstances; also some constant parameters depend on human experience more, leading to larger subjective errors. And the workload of field test is very large, also time-consuming and laborious. Simulation calculation results and actual ventilation system do not match because wind resistance data obtained from empirical formula or field test have errors. These data cannot be applied to the ventilation system simulations.How to inverse mine ventilation resistance coefficient by a few representative data of air flows and node pressures, this is a topic worth studying. The researches on this aspect are seldom discussed. "Research for an intelligent diagnosis system of the mine ventilation system based on simulation" is based on the National Natural Foundation of China (60772159).The matrix equations were established for the ventilation resistance coefficient inversion on the basis of the three basic laws of fluid network, proved the ventilation resistance coefficient inverse problem is ill posed and still have multiple solutions with the condition of both multipoint one observation or less measuring points in many observations. The observational data of air flows and node pressures are limited, so the number of equations less than the number of unknown variables.The mathematical model of ventilation resistance coefficient inversion is established based on the least squares principle. The measured pressure and deviation calculation of pressure and flow measured and calculated flow deviation are considered as the objective function, containing node pressure, air flow and the ventilation resistance coefficient range constraints. The ventilation resistance coefficient inversion problem was converted to a nonlinear optimization the problem through the establishment of the model. Genetic algorithm and particle swarm algorithm were adopted to solve ventilation resistance coefficient inversion problem, which is established based on the least squares principle. GA and PSO were both improved to enhance the global search and the local search ability of the algorithm for the ventilation resistance coefficient inversion problem.On this basis, some measured values with high relative sensitivity can be used. A method of layouting measuring point was put forward that combined with the ventilation system sensitivity theory and the theory of clustering analysis, reflecting the influence on node presses and air flows of the resistance coefficient. The method tries to classify the nodes and edges, to look for a few measuring points of nodes and edges, to reflect the actual state of ventilation system in the maximum possible.Finally, particle swarm optimization method was applied to the ventilation system of Sihe 2# well, and the method is been validated. The research results of the thesis which can be used in future study are theoretical foundation of mine ventilation resistance coefficient inversion and real-world application, which has important significance to the mine ventilation system.