【摘要】 基于人工免疫的识别原理提出了一种转矩流变仪故障诊断方法,给出了相应的模型;通过对转矩流变仪的正常状态下获得的自体集否定选择,随机产生初始检测器集;在否定选择算法中引入故障类型判定疫苗,利用转矩流变仪正常工作样本和故障样本获得了合理的抗体集;该方法具有准确检测、定期更新和动态调整的功能,可以快速、准确的故障诊断和判定故障类型;最后通过转矩流变仪的双螺杆挤出机本体的故障诊断对算法进行了验证,结果表明,不同故障类型的识别准确率稍有不同。0类和2类的准确率达到93%以上,具有一定的可信度。更多还原

【Abstract】 The article puts forward a kind of torque rheometer fault diagnosis method based on the recognition principle of artificial immune;and gives the corresponding model.Through negative selection of the normal state self set of the torque rheometer,an initial detector set is randomly generated.At the same time,by introducing the fault type vaccine in the negative selection algorithm,the reasonable antibody sets can be obtained by the torque rheometer normal work sample and fault samples.The method owns accurate detection,regularly updated and dynamic adjustment function,so it can obtain quick and accurate fault diagnosis and fault type decision.Finally,through the example of torque rheometer double screw extruder ontology,the fault diagnosis algorithm has been validated,and the results showed that the accuracy of different fault identification is slightly different.Class 0and 2accuracy above 93%,it has certain credibility.更多还原