Abstract: According to the "Sandver theory model" and to resolving the overcapacity in the labor conflict risk status and the main problems in the investigation and analysis of Beijing Tianjin Hebei,the labor conflict risk early warning index system theory is designed based on the primary early warning index system,the rough set theory is used to analyze the relationship between the sample of enterprise historical data and the results of labor clash,and the attribute reduction iscarried to objectively determine the weight of each index,then combining subjective weight,to get the weight of each index,and reduce the first 22 three level index theory to 15 operating index system.After collecting the six industries with excess capacity represented by 48 enterprises and 246 copies of questionnaires and data in Beijing Tianjin Hebei,including iron and steel,coal,cement,electrolytic aluminum,flat glass,and boat,and on the basis of the subjective and objective weights and the standardization of data processing,this paper uses the multi attribute decision making method to evaluate comprehensively and have quantitative warning on the labor conflict risk status of the sample enterprises.The combination weighting method is used to enhance the objectivity and the scientific nature of the quantitative warning,and early warning results show that: there is a half of the sample enterprise labor relations having a certain risk,in which the enterprises are in the serious warning state,that is in the dangerous area accounting for 17%,and in the very severe state,that is in the very dangerous area accounting for 2%.The enterprises which are in the serious and the most dangerous areas are mainly state owned enterprises,accounting for 50%;private enterprises account for 30%;joint stock enterprises account for 16%.According to industry distribution,the enterprises mainly concentrate in the steel industry.

Key words: resolving excess capacity for BeijingTianjinHebei, rough set, conflict between labor and capital, risk earlywarning, multiattribute decision