Optimization decision support system for safe ship control

Abstract

This paper describes the application of selected methods of optimal control theory for a ship to determine a safe trajectory during the passing of other ships encountered in restricted visibility at sea. The methods of comparison of safe ship control in a collision situation – multi-stage positional non-cooperative and cooperative game, multi-step matrix non-cooperative game, dynamic optimization with neural constrains of the state control process and kinematics optimization – have been introduced here. The synthesis of computer navigator decision supporting algorithms using dynamic programming and dual linear programming methods has been presented. The considerations have been used to illustrate examples of a computer simulation of the algorithms to determine the safe and optimal own ship trajectories in a navigational situation in restricted visibility on the North Sea.