Abstract: Port handling crane starts or brakes frequently, running in bad condition. In order to detect the dynamic lifted load of the port crane, self-adaptive filter is developed, based on the identification of dynamic model. The self- adaptive filter can eliminate efficiently the signal of acceleration interference when the crane starts or brakes, and the signals of momentum impact interference. The off-line pre-estimating algorithm for model parameters is founded, and the on-line detecting speed is raised. When the crane is in the zero-speed, zero-load, the self zero-tracking will act, and zero drift of the detecting system can be corrected. The applying results show that the new technique has raised the accuracy of detecting lifted load.

Key words: Lifted load, Detection, Dynamic model, Identification, Self-adaptive