Operational Risk management is an important part of risk management. Facing to today's increasingly frequent operational risk of commercial banks, commercial banks need to develop practical operational risk control procedures and relevant measures, The establishment of Early-warning systems of operational risk would play a very important role in the future risk management. By means of the BP Neural Network analysis and nonlinear model methods, the paper would analyze and resolve the nonlinear relationship between several Key Risk Indicators (KRI) factors of commercial banks operational risk and risk results. According to establish the Early-warning system model of operational risk management for commercial banks, the paper expect that the research could provide some important references to prevent and control effectively operational risk for commercial banks.