工业技术与职业教育

为解决船舶柴油机故障无法提前预判处理的问题，提出了基于粗糙集和DS证据理论的方法。粗糙集对海量数据进行处理，属性约简提取重要特征参数，得到简化的信息系统。DS证据理论从信息系统中得到基于证据信任度的基本可信度分配，利用证据合成规则有效融合证据信息，为船舶柴油机状态提供预测参考。经过验证，表明该方法对船舶柴油机燃油系统的故障具有较高的识别率。

A method based on rough set and DS evidence theory is proposed to solve the problem of inability to predict and handle ship diesel engine faults in advance.Rough set is used to process massive data,extract important feature parameters through attribute reduction,and obtain a simplified information system.The DS evidence theory obtains the basic credibility allocation based on evidence credibility from information systems,effectively integrates evidence information using evidence synthesis rules,and provides predictive references for the status of ship diesel engines.After verification,it has been shown that this method has a high recognition rate for faults in the fuel system of marine diesel engines.