A Digital-Twin Assisted Performance Prediction Model for Industrial Gas Turbines

Performance degradation of gas path components in industrial gas turbines increases the risk of the gas turbine failure. Accurate prediction of gas turbine performance parameters is important to ensure the safe and stable operation of the gas turbine. A digital-twin assisted performance prediction model for industrial gas turbines is established in this paper to predict the performance precisely, guide the operation control of the gas turbines and optimize the operation and maintenance strategy. An architecture of the digital-twin assisted performance prediction model is proposed including the physical entity, the virtual model, the twin-data, the virtual and real interaction, and the services. A general model of gas turbines is established considering the influence of the environment and working conditions. A model calibration method based on genetic algorithm is proposed by introducing the scaling factor of the performance maps. A performance prediction model with higher accuracy is obtained by the interaction and update between the monitoring data of physical entities and the virtual model. The proposed model is applied to the analysis of the monitoring data in variable working conditions of an in-service industrial gas turbine. The digital-twin assisted performance prediction model can predict the outlet temperature of the compressor outlet temperature, the exhaust temperature, the speed, and the power with an average prediction error of 1.39%. The proposed model contributes to the optimization of the operation and maintenance process of industrial gas turbines.