Study On Analysis And Decision-making Of Industrial Explosive Enterprise Security Investment

Security production is a problem must be carefully considered in the industrial explosive enterprise throughout the whole production process. To increase security investment can prevent or eliminate the occurrence of certain accidents, reduce casualties and economic losses and can bring benefits far outweigh the cost of inputs. In this paper, the theoretical analysis tools were applied to qualitative and quantitative analysis based on the statistical data of security investment of industrial explosive enterprise, which had guidance to practical work. The main work and conclusions of the study are as follows:Gray relational model was selected to calculate the correlation degree between security investment items and security benefit of a real explosive enterprise. The descending sort of correlation degree when the correlation coefficient values of0.1and0.5were:the cost of security equipment inspection, security protective articles, security education training, daily security facilities, dangerous source management and hidden dangers rectification, emergency rescue, security evaluation, occupational disease prevention, which verify the accuracy of the model. The enterprise can optimize security investment structure and achieve reasonable allocation of safety investment according to the sort.The sample data of a industrial explosive enterprise from the year2007to2011were used to train the BP neural network, selecting the optimum candidate with the least network errors and training times. This paper also made network approximation of different training functions to confirm the single hidden layer BP network structure which had10hidden layer neurons, the trainlm function and an error of0.0059. After training for69times with the network parameters were set, the network error fulfilled the requirements of the target. The error between the predicted value and the actual value of the investment in2012could be controlled in small range, and the total investment error was only0.028%. The example indicated the accuracy and feasibility of network prediction. The investment value in2013was predicted after that. Compared with the value in2012, the growth of the predicted value ranked as follows:the cost of security equipment inspection, security protective articles, security education training, dangerous source management and hidden dangers rectification, daily security facilities, emergency rescue, security evaluation, occupational disease prevention, which basically consistent with the results of grey relational model, can provide scientific basis for the enterprise security investment decision-making.Finally, the game analysis between enterprise and employees, enterprise and competitors concluded that the enterprise with a reasonable safety investment in the long-term would get better security status and profits. And recommendations about the improvement of safety investment were carried out both from two aspects of enterprises and government regulation.