The Evaluation Methods And Applied Research Of BP Neural Network On Mine Ecological Safety

Mineral resources is an indispensable of important material foundation for human existence social development and economic construction, but in the past for a long period of time,extensive mining model, backward production technology,imperfect management system caused a more serious impact or even destroy on the integrity of the system and the health status of mine ecological environment, resulting in the gradual deterioration of the trend of ecological safety in mines, and seriously hindered the sustainable development and utilization of mining resources. Therefore, quantitative research on mine ecological security is very necessary.In this paper, on the basis of the results of previous studies, In view of the problem of quantitative technology evaluation mine ecological security,from the integrity and health of mine ecosystem point of view, in combination with BP artificial neural network basic theory, the mine ecological security evaluation model first was proposed and established based on BP artificial neural network,it formed technical methods of quantitative analysis and evaluation of mine ecological security system,which provided scientific guidance for the protection and restoration of mining ecological environment,and had certain theoretical and practical significance.The main research results as follows:(1) According to the characteristics of the Shanxi Coal Mine Environment, this article combined with the theory of ecological security assessment and BP neural network method, from four aspects of Mine ecological security assessment on division level, feature description, establishment of index system, formulation of indicators standards of evaluation and establishment of evaluation model,the technical methodology of quantitative analysis and research mine ecological security was established.(2) Combined with the mine ecological security features, the mine ecological security is divided into ideal, good, early warning, poor, poor five levels to use the thought of aliquots and normalized data processing method.By the interaction between the research and analysis of mining natural resources, environment and human behavior, PSR conceptual model extension framework was proposed, and follow the principle of index selection, it selected25the evaluation index factor, and established the index system of mine ecological security evaluation; combined with the actual situation of the study area and with reference to the relevant standards, the index standard system of mine ecological security assessment model was first formulated based on BP artificial neural network.(3) From the perspective of the demand of model value and the cohesive force of technical methods, the randomly interpolation method expanded training samples,which solved the problem of insufficient training samples; the hidden layer nodes of BP neural network are optimized to use gold segmentation algorithm, and determined the best topology of the network model, mine ecological safety evaluation model was first established based on BP neural network.(4)The study takes Tangshangou coal mine in Datong as an example,carried out empirical study on mine ecological security evaluation model,the evaluation results basically accords with real situation of mine area,and have verified the BP neural network evaluation model of mine ecological safety is practicality and operability.