关键词: 舰船维修成本, 案例检索, 模糊聚类

Abstract: Aiming at the greater uncertainty problem of ship maintenance cost forecast in the market economy, the case based reasoning method was used to solve the problem of ship maintenance cost forecast, and the case retrieval technology in forecast was studied particularly. The navy ship was a kind of type selection armament and attribute weights of ship maintenance cost cases were unequal. For these features, an improved similarity function named weighted cosine distance function which was used to measure the similarities of the ship maintenance cost cases accurately was designed. To improve the efficiency of case retrieval, based on the classification of ship maintenance cost case base, a new case retrieval method for ship maintenance cost cases based on classification strategy and combinatorial fuzzy clustering was proposed by integrating the fuzziness characteristic of maintenance cost case attributes. The effectiveness of proposed method was verified by an actual example.

Key words: ship maintenance cost, case retrieval, fuzzy clustering, similarity