内燃机运行工况的两种识别方法

Two Fault Recognizing Measurements for Operation Condition of IC Engine

在模式集难以预先设定，但监测时只提供运行工况是否正常信息的前提下，尝试运用分形理论中的关联维（ＧＰ法关联指数）计算方法和模糊聚类原理，对内燃机故障样本进行识别．将以上两法应用于由１９０Ａ内燃机机身处测得的不同活塞裙部－缸套间隙的振动加速度样本群，和由４１３５Ｇ内燃机机身处测得的不同气环状况的振动加速度样本群，表明了所述方法是实用和有效的．同时也表明由活塞－活塞环－缸套－机身所构成的系统是非线性动力学系统．

While the standard pattern is difficult to establish in advance and the information about the operation condition (good or bad) of machine is only needed to be known,the correlation dimension of fractal and fuzzy collection are offered to recognize the samples in this paper.The measures are testified to be practical and effective while to be applied to the samples gotten from the block of 190A diesel with different pistion skirt cylinder gaps,and to the vibration acceleration samples collection gotten from the block of 4135G diesel's No.4 cylinder with the different operation condition of pressure rings.Meanwhile the system,which is formed by piston,cylinder,rings and block,is testified to be a dynamic one with the property of non linear and time variety.