A fast-track ship simulation model for port studies

Port studies are systematic investigations of the safety, or difficulty, of manoeuvring ships in a port in various environmental conditions. They are often used to enable decision-makers to analyse alternate port layouts for a proposed new port, or to assess the safety of manoeuvring a particular ship in an existing port. Port studies are generally carried out in ship manoeuvring simulators, in real time, using real pilots. These studies can be extremely time consuming and costly, as a port study requires a large number of simulations covering a range of weather conditions and ships. Although a computer can perform these simulations much faster than real time, a human pilot cannot. In order to reduce the numbers of scenarios which must be investigated in a full mission bridge simulator, it is proposed to develop a 'fast-track' method of rating the difficulty of simulation scenarios, using a simple computer search algorithm in place of a human pilot. As the first step in developing this search method, a simple model, capable of manipulating the rudder and engine only, has been created. Using the Australian Maritime Engineering Co-operative Research Centre's existing PC-based ship handling simulator, this computer search algorithm replaces the supervisory role of a human pilot in a simulation, and uses rudder and engine commands to search all possible routes from a start point to a given target point. The computer search method logs the numbers of successful and unsuccessful paths tested in this process. Two simple computer search models were created and used to evaluate the difficulty of a number of port exercises, and each of these port exercises was also evaluated in the Australian Maritime College's full-bridge simulator, using experienced master mariners. These mariners rated the difficulty of each port exercise on a preset scale, for comparison with the results of the computer assessments of difficulty. This thesis presents the models for the computer search algorithms, and the results of two series of tests using a 215m bulk carrier. The final set of tests shows good agreement between the assessments of difficulty given by the master mariners, and the assessments produced by the one of computer search methods. The other computer search method proved to be impractical to implement.