由于柴油机气缸缸盖振动信号具有复杂的时频特性,通过此类信号实现其气门故障的诊断较为困难。尤其在多种故障并发的条件下,故障确诊更为不易。为此,借助遗传算法提出一种基于统计规则的智能数据挖掘技术,对在不同气门状态下采集的大量柴油机气门缸盖振动信号进行知识挖掘,得到了进行多种气门故障确诊的有效诊断特征。试验表明,这一技术智能高效,结果准确无误,具有重要实践意义。 Due to the complex time-frequency characteristics of diesel engine cylinder head vibration signals, it is difficult to diagnose the valve faults by such signals. Especially in a variety of fault conditions, the fault diagnosis is more difficult. Therefore, based on the genetic algorithm, an intelligent data mining technology based on statistical rules is proposed to excavate the vibration signal of a large number of diesel cylinder valve head vibration signals collected under different valve states, and the effective diagnostic features for diagnosing various valve faults are obtained. Experiments show that this technology is intelligent and efficient, the result is accurate and has important practical significance.