基于改进型阴性选择算法的车辆故障检测方法研究

根据现有的阴性选择算法提出适于车辆故障检测的改进型阴性选择算法。在改进型算法中，对数据串编码、自体集合的生成、检测器数目的确定进行了研究。对某军车上使用的防滑差速器正常工作和故障状态下的样本数据进行故障检测，结果表明了改进型阴性选择算法的有效性，并与神经网络检测方法所得到的结果进行比较，表明了改进型阴性选择算法具有较高的检测准确度。

Research on Vehicle Fault Diagnosis Method Based on Improved Negative Selection Algorithm

According to the negative selection algorithm in existance, an improved negative selection al-gorithm which adapts to vehicle online testing was presented. The methods of coding data string, cre-ating self sets and determining number of detectors in the improved negative selection algorithm were analyzed. The detected results by the noise data of normal and fault non-slipping differential used in a certain military vehicle show that the improved algorithm is effective and more accurate compared with the neural network testing method.