Intelligent ship command task assignment considering facial expression and ability states of ship operators

Abstract

The traditional ship safeguard tasking scheme is formulated by the shipmaster based on his own experience. This experience-based method for safeguard task assignment is more applicable when the safeguard tasks are less and the ship operators meet the demand. However, when the safeguard task become very complex and difficult, this method can’t perform a scientific and reasonable task assignment decision to the operators that is assigned to engage to this task. Especially, this approach ignores the variability influence of the ship’s each operator’s personnel work state and the specialized capability, which resulting in unreasonable task assignment and low utilization of safeguard resources. To solve this problem, we design an intelligent ship safeguard task assignment by acting the facial expression and ability states of ship operators as the core. Firstly, the ship operator’s facial expression is recognized by the Convolutional Neural Networks(CNN). Secondly, a time series analysis is performed by using the recorded time from the ship operator to complete a certain type of task, and thus get the time prediction that he completes the same task in next. Finally, the intelligent ship safeguard tasking scheme is designed to complete the whole task by the genetic algorithm based on the operator’s emotion and personal ability. The experimental results show that the model can efficiently perform the ship safeguard task assignment by considering the accuracy of emotion recognition and ability prediction.