Real-time Monitoring and Optimization Modification for Turbine Performance Based on Data Driven Model

Abstract

The performance of the unit is in a state of realtime degradation, and the traditional proprietary test method cannot accurately monitor this performance degradation in time. Aiming at the above problems, a real-time monitoring system and optimization scheme for steam turbine performance based on data driven model is proposed. Firstly, a steam turbine performance prediction model based on pattern recognition and prediction function is established, which can realize medium and long-term prediction of turbine operating economy and provide early warning for performance degradation. Then, performance analysis is performed for units performance degradation, respectively in thermal system and communication. The corresponding optimization scheme is proposed in the flow design. Finally, the 600MW supercritical unit is taken as the research case. The results show that the above method is effective and feasible in practice.