The Algorithm Research Of Coal Mine Safety Evaluation Based On Hidden Danger Analysis

Coal as a basic industry of our country, it is main energy sources and has apivotal position in economic and social development of China. However in recentyears the situation of coal mine safety production is not optimistic, it causes not onlya serious threat to personnel security and the social harmony and stability, but also ahuge economic losses. As a result, to carry out the research of coal mine safetyevaluation positively, to evaluate the number of factors which impact coal minesafety production objectively and reasonably, to provide practicable solution for coalmining enterprises, that are important things of great social significance which relateto sustainable development of economic construction and social security and stabilityand are major issues in our country have long faces and sustained research and theurgent need to be solved.In this paper, several aspects as principles of safety, improved algorithm,simulation result and system design of achieving the related research of coal minesafety evaluation algorithm step by step. The main work is divided into the followingseveral aspects.⑴This paper put forwards the4M-C-I theory, basing on the hidden dangersmen、machine、medium、management、culture and legal and regulatory informationestablishes and gradually improves the safety evaluation index system which in linewith the mine production actual.⑵Combined with the genetic algorithm, this paper further improvesoptimization the initial parameters of Wavelet Neural Network, overcoming the othershortcomings as dependences on the settings of network parameters and easily to fallinto local minima existed in Wavelet Neural Network, enhancing the accuracy of thecoal mine safety evaluation results.⑶In order to achieve the optimization algorithm of Support Vector Machine,this paper proposes a method of adjusting steps to look for the optimal penaltyparameter C and loss parameter g. And through MATLAB simulation verifies theimproved algorithm has much higher operational efficiency.⑷Based on B/S network architecture this paper designs and develops the coalmine accidents and safety evaluation prototype system, develops and implements the key safety evaluation module with mixed programming ideas. In this module, severaldifferent evaluation algorithm results such as Wavelet Neural Network, GeneticAlgorithms-Wavelet Neural Network, Support Vector Machine, C g-Support VectorMachine are given.The research work in this paper not only provides effective model approach ofcoal mine safety evaluation but also effective ways for coal mining enterprises toimplement of decision management.