基于灰色神经网络的船舶柴油机活塞环退化可靠性研究

Study on reliability of piston ring degradation of ship diesel engine based on grey neural network

柴油机能否可靠运行直接影响着船舶任务的执行。传统的柴油机可靠性评估方法依赖大量的统计数据、忽略了个体差异,因而进行评估的能力有限。引入灰色神经网络的方法对退化数据进行处理。通过两台某型船舶柴油机样机耐久性试验得到的活塞环磨损退化数据,验证了此方法的有效性。此方法理论上可以在少量历史数据的基础上,减少耐久性试验的时间,对失效物理过程未知的小样本退化数据的可靠性分析上,有一定应用前景。

Whether the diesel engine can run in a reliable way is related to the security of the whole ship.Using traditional methods to assess the reliability is not accuracy.Because it relies on a lot of statistical data,and the individual differences is ignored.This article introduces the grey neural network method into analyzing degradation data.The effectiveness of this method is verified by piston rings’ wear degradation data.This method based on a small amount of historical data.It can reduce the time of the durability test,and has the certain application to prospect the failure which has unknown physical process.