基于案例的推理在舰船维修费预测中的应用

针对市场经济条件下舰船维修费预测不确定性越来越大的问题,提出了一种基于案例推理的舰船维修费预测模型。首先,利用修船厂的历史数据,采用特征向量表示法构建舰船维修费案例库;其次,为提高案例检索效率,采用组合模糊聚类算法建立舰船维修费案例的分类索引结构,并提出基于最近邻法的两步检索策略检索相似案例;最后,采用加权平均法对相似案例进行修改以获取预测值,同时采用主动学习策略保存当前案例并更新案例库。实例仿真结果表明:与传统舰船维修费预测模型相比,基于案例的推理模型具有更高的预测精度,所提模型是有效的。

Application of case-based reasoning to ship maintenance cost forecasting

The case-based reasoning model was proposed to solve the problem of high uncertainty in forecasting ship maintenance cost under the conditions of the market economy.Firstly,the historical data of ship maintenance cost gathered at shipyards were used to build a case bank based on the eigenvector representation method.Secondly,in order to improve the efficiency of case retrieval,the indexical structure of ship maintenance cost cases was established by use of the combined fuzzy clustering method,and the two-step retrieval strategy based on the nearest neighbor algorithm was proposed.Finally,similar cases were modified to acquire the predictive values according to the weighted average method,and meanwhile an active learning strategy was adopted to reserve and refresh cases.The practical example indicates that the proposed model is very accurate in cost forecast compared with the conventional ones,so the proposed model is effective.