Study On Thoeries For Network Algorithm In Muti-level Fan Station Ventilation Fashion

Multi-stage fan station ventilation mode is used in some metal mines, when the fan power of a fan station is great than the balance point of power, the opposite direction airflow will be occur in both sides of ventilation network, and unidirectional circuits occurr in a ventilation network in the same time. With unidirectional circuit, all the existing ventilation network algorithms will fail, but many ventilation network algorithms, such as ventilation network equilibrium graph plotting algorithm, diagonal structure identification, ventilation network mapping, and automatic assignment of initial values during network solution, are based on it. Based on this, this paper mainly researches algorithms for ventilation networks with unidirectional circuits, and the paper are based on the National Science Foundation-funded project "Simulation-based Intelligent Diagnosis of Mine Ventilation System of Systems" (60772159).Based on the analysis of characteristics of ventilation networks with multi-stage fan stations, the paper derives the unidirectional circuit formula, which is formed from the critical point. To solve the failure problem of path method in drawing ventilation network equilibrium graph with unidirectional circuits, a source addition equivalence transform method is proposed. The rectangular block number of a ventilation network equilibrium graph depends on the ventilation network topological transformation; this paper describes the ventilation network topological transformation using the chromosomes in genetics, and solves the problem of binary and integer encoding. I firstly propose an optimization method that drawing the ventilation network equilibrium graph based on the genetic algorithm, namely the number of minimum cut rectangle problem. It is proved that this method has good convergence through drawing the ventilation network equilibrium graph of Jinchuan No.2 Mine by using the computer software.In allusion to the problem of branch crossing in plotting network graphs, an optimizing plotting algorithm for mine ventilation network graph is explored. A study of node hierarchy is done based the longest path method, and parallel genetic algorithm based on migration strategy is used to set the orders of nodes on every layer. Through the plotting network of Jinchuan NO.2 Mine, the results suggest that this algorithm was proved to be effective in reducing branch crossing number and making a better hierarchy layout.Diagonal structure algorithms for a ventilation network with multi-stage fan stations are also explored. The diagonal structure determination algorithm, path method, for directed graph is mainly studied. A new definition of generalized diagonal structure is put forward.