By adopting the recursive least square method

a close-loop identification of the main steam temperature control system has been carried out.Based on features of the cascade control structure

and through double linearity transformation model

the identification model has been converted

and the identification links being divided

thereby

the identification parameter equation being deducted and obtained.Based on the integration criterion of time multiplying absolute error

the off-line optimization of PID parameters has been realized

and the effectiveness of above-mentioned method being verified on the emulation model.The said identification model has been used for optimization of PID parameters in the main steam temperature control system of one 600 MW unit

results show that the dynamic regulation performance of the main steam temperature control system has been remarkably enhanced

at the same time

the target of off-line optimization for the control system can be reached.