To quantitatively analyzing the degradation degree of performance of gas-turbine combined cycle(GTCC) units

the performance-time degradation research model for the unit at different load is proposed. On the basis of historical operation data of PG 9351FA GTCC unit

GA-BP neural network is adopted to establish the prediction models for gas turbine efficiency and GTCC efficiency in each year. Moreover

physical model for the GTCC overall characteristics is established in THERMOFLEX simulation software

to investigate the influence of compressor and gas turbine with different degradation degrees on the system performance. The results show that

the relative error of prediction of test set based on the data-driven efficiency model is within 3.5% and the correlation coefficient R exceeds 0.9

which demonstrates the good generalization ability of the neural network. Both the gas turbine efficiency and GTCC efficiency at the same unit load decline gradually year by year

while they are apparently recovered at high load after A-class overhaul. The influence of turbine performance degradation on system performance is approximately twice as large as that of the compressor degradation. The work provides theoretical guidance to the study of degradation mechanism

which has engineering application value.