Research On Mined-out Area Hazard Evolution Model Based On Fuzzy Petri Nets

While the mining of mineral resources service for the national economic construction and improve the people's material life, the mineral's mine left the plenty of mined-out area, and the number of it is increasing year by year. Because of the mined-out area's roof fall, rock burst could caused land slides, ground fissures, ground collapse, and other forms of disasters, these disasters lead to the tremendous loss of life and property.Firstly, On the analyzed the mined-out area's risk factors and evolution processing basis, this paper combined the mined-out area's risk evolution with fuzzy Petri net theory, and put forward the mined-out area's hazard evolution model-HFPN, which is based on the fuzzy Petri net. In this model, the risk factors and events in hazard evolution dividing used place and transition in Petri net express, the facility degree in risk factors equivalent with the credibility degree in Petri net, according to membership degree of fuzzy sets to fetch quality, and the interval is [0,1], and given the model's reasoning algorithm. Secondly, considering the processing of hazard evolution in mined-out area, which not only the risk factors could influenced the evolution process,but also the management action of mined-out area could influenced it, in order to make the established model can fit the actual situation of mined-out area hazard evolution, then draw the both-branch fuzzy set theory, and it can make the affected factors of risk evolution from one-sided impel evolution extend to positive and negative aspects which is promotion(risk factors) and control(management action), the fetch value area of influence factor's facility degree also extended to [-1,1]. Combined the both-branch fuzzy set theory with fuzzy Petri net theory, and established the mined-out area hazard evolution model-HBBFPN based on the both-branch fuzzy set theory. In HBBFPN, we considered the condition of the very low credibility degree of affirmative proposition, the negated proposition of it is more credit, then we put forward change weight algorithm, and use it transform the affirmative proposition which equivalent negative proposition's credibility degree. In addition, on the base of the analysis model and changed weight algorithm put forward the model's reasoning algorithm. Finally, the analysis results showed that in previous models, the weight's fetch are all based on the past experience, this method is very well to used assess the initial place, but when estimate the weights between the model still using this method may produced the model is not work very well because of human estimate value error, and may get the wrong evolution results. In order to solve this problem, combined the artificial neural network with fuzzy Petri net, and put forward the mined-out area hazard evolution model-HNNFPN, which is based on the artificial neural network and fuzzy Petri net, and gives the model improved BP learning algorithm and the model's reasoning algorithm.This paper put forward three kinds of models and reasoning algorithm, which is verified proving by real examples, the results showed that the model and its algorithm's the correctness and rationality, and also expressed this model can play a role in guiding the treatment of the mined-out area, and provide the new idea for the mined-out area's hazard research. This model can play a role in guiding the treatment and research for the mined-out area hazard new ideas.