In order to establish an accurate prediction model for heat consumption rate of steam turbines

an integrated modeling method was proposed by combination of oppositely adaptive whale optimization algorithm (AWOA) and fast learning network (FLN). Compared to basic whale algorithm

improved particle swarm optimization algorithm

and differential evolution algorithm

the improved whale algorithm had higher convergence accuracy and faster convergence speed. A prediction model for heat consumption rate of a 600 MW supercritical steam turbine generator set in a thermal power plant was established from the collected operation data

which was also compared to FLN model

improved particle swarm optimization

differential evolution algorithm

and whale optimization algorithm. The results show that the AWOA-FLN prediction model had higher prediction accuracy and stronger generalization ability

which therefore could predict heat consumption rate of steam turbine more accurately.