In the localization structure of moving long baselines(MLBL), although the measurement information, which can
be got by using the acoustic propagation time of flight(TOF) and Bayesian filters such as the extended Kalman filter(EKF),
is utilized to improve the localization accuracy of a low self-localization capability unmanned underwater vehicle(UUV),
the higher measurement errors will reduce the extent of this improvement. A correlation assumption is proposed and the
error correction algorithm(ECA) is constructed according to the characteristics of the underwater acoustic communication.
Under the setting conditions, the measurement errors are depressed by using of the correlation between the errors, and the
rough estimates of the measurements are achieved. The simulation results show that the localization accuracy of the UUV
equipped with low precise proprioceptive localization sensors can be improved significantly by combining the measurement
rough estimates with one of Bayesian filters.