Proactive struck-by risk detection with movement patterns and randomness

摘要

This study proposes a proactive struck-by risk detection method for workers proximal to the laydown area. This approach regards the laydown area as a radiation source, and proactive struck-by risk detection is formulated as a location-based estimate of radiation quantity received by the worker. The position probability grid is adopted to track the worker and model their movement. The movement pattern model, which is learned from the trajectory records of the workers in the same trade, reflects the movement preferences of different trades, and the randomness model represents the randomness of worker movement. The worker's next location can be predicted by using the tracked location and movement model, and the struck-by risk can then be estimated. The methods potential for integration into currently available proximity warning systems is also demonstrated.