The considered Pareto-optimal state observer allows isolation of faults appearing in the ship propulsion system (engine plus propeller). The discussed project is based on multiobjective optimization of vector criteria taking into account faults external disturbances, noises and modeling uncertainty. The procedure of multiobjective searching for a Pareto-optimal set of solutions is supported by evolutionary computing, which allows for a global search in multidimensional spaces of optimized parameters that is immune to possible multimodality of partial objective functions. At the same time the multiple solutions yielded by the Pareto-optimal approach, which are generally criticized for their ambivalency or non-uniqueness. are fully utilized applied in the process of ranking the individuals of parental pools. In order to make a final evaluation of Pareto-optimal solutions obtained each of them is characterized by a global-optimality level index which relates the solution to attainable values of all the partial quality indices.
Pareto-optimal observers for ship propulsion systems by evolutionary computation
Berechnung Pareto-optimaler Beobachter mittels genetischer Algorithmen für die Fehlererkennung bei Schiffsantrieben
2000
6 Seiten, 5 Bilder, 18 Quellen
Conference paper
English
Pareto Optimal Ship Course Keeping Controller
| British Library Conference Proceedings | 2001
Electronic computation of torsional vibrations in ship propulsion systems
| Engineering Index Backfile | 1960
Optimal propulsion systems for all ship types
| Tema Archive | 1999
Optimal propulsion systems for all ship types
Online Contents | 1999