Study On Evaluation Of Comprehensive Performance Of The Coal Enterprises Based On AHP And BP Neural Network

The coal industry is the basic industry that supports the development of the national economy,and it is related to the development of the downstream industry and social stability.In recent years,coal prices remain low for a long time and the coal market has been depressed.Supply-side reforms in the coal industry are under pressure and the coal enterprises have been in trouble.At the critical moment of both opportunities and challenges,many enterprises go through the test of national policy,such as resource integration,mergers and reorganization,the upgrading of the industrial structure.The coal enterprise integrated performance evaluation becomes particularly important.On the basis of the traditional financial performance indicators,this article refer to a large number of related literature at home and abroad and the social responsibility evaluation index listed in the corporate social responsibility report.Combined with current situation and the particularity of the coal industry,coal enterprise integrated performance evaluation index system is constructed from eight dimensions includes financial performance(profit ability,debt paying ability,operating ability,growth ability)and non-financial performance(safety production,technology performance,environmental protection,social contribution).The performance evaluation model is established by using Analytic Hierarchy Process and BP neural network.Import the data from 5 listed coal companies in 2011-2015 as training samples and the 20 tertiary indicators of evaluation index system as the input vector.BP neural network model is constructed with comprehensive performance score for the output vector.Finally using the data of five companies in 2016 as the test sample,the model training and simulation verification of the comprehensive performance of the sample enterprises are carried out,and the comprehensive performance of the sample enterprises is analyzed according to the evaluation results.From the output of the model,the model can evaluate the comprehensive performance of coal enterprises(the relative error mean is 5.23%)within a reasonable error range.By using the trained network model with fixed weight threshold,the comprehensive performance evaluation of any coal enterprise can be realized,which has good practicability and operability.The evaluation results show that the comprehensive performance evaluation value of the five coal enterprises has been decreasing year by year since 2011,and the decline is large and the overall performance is at a low level.In 2016,there are four samples of the company's comprehensive performance score rebounded.This is because the supply-side reform of the coal industry has begun to bear fruit,and the supply and demand of the coal market has been balanced,and the operation of coal enterprises has significantly improved.Among the five enterprises,the comprehensive performance of Shenhua has been far ahead of other enterprises,which is worth deeply analyzing its reasons and providing ideas for the sustainable development of coal enterprises.Model constructed in this paper can be used for coal enterprise integrated performance evaluation and forecasting its future development potential,besides to provide certain reference for public policy of government departments.The model has good application value.