The real-time maneuvering simulations commanded by marine pilots are a fundamental tool to assess the nautical design of new terminals and verify larger vessels' feasibility in existing ports. It is unfeasible to make a large number of real-time simulations due to time constraints. Each simulation run takes approximately 1,5h. So, alternatives are employed to pre-evaluate a large number of environmental conditions, effectively grouping similar conditions into bins that can be represented by a single environmental condition, which can then be used for the simulation with real pilots.This work proposes the utilization of an automatic pilot that will control a vessel in a simulated scenario under thousands of different environmental conditions of a port location, using a fast-time simulator. &nbsp;The drifting angle time-series of the vessel is recorded, and then a cluster analysis algorithm is applied to these time-series to group conditions accordingly to how they affect the vessel's drifting. Since many mild environmental conditions produce weak effects in the vessel, the user can select a percentage of the cases which caused the strongest impact to exclude those milder environmental conditions.The auto-pilot uses a line-of-sight algorithm to keep the vessel on the desired path. The cluster analysis is made by the Partitioning Around Medoids algorithm using distances obtained from Dynamic Time Warping applied to the drifting angle time-series. To find the best number of groups in the data, the Silhouette method is used.The method was validated by using the selected environmental conditions for simulations carried with officers at the TPN-USP Ship Maneuvering Simulation Center, in 3 different port locations. For each case, a large number of environmental conditions (between 6000 and 8000 combinations of wind, current, local-sea and swell) were pre-run in the fast-time simulator and the clustering method was applied. Using the silhouette method we obtained two qualitatively different groups, that were further divided into three levels of intensity (1%, 5%, and 100% of the strongest drifting angles in the group).After each maneuver, the officer was inquired about their perception of the simulated maneuvers. We found an acceptable level of coherence between our assumption that the drifting angle defines both the maneuver difficulty and characteristics, in the majority of cases the officer grouped the maneuvers by the vessel’s drifting direction in a similar way to our algorithm, and the average perceived difficulty of a maneuver increases as the mean drifting angle increases.

•	Clustering Applied to Large Sets of Environmental Conditions for Selecting Typical Scenarios for Ship Maneuvering Real-Time Simulations

Paper Proceedings: https://asmedigitalcollection.asme.org/OMAE/OMAE2021/volume/85116

OMAE 2021

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.