基于免疫遗传和顺序支持向量分类机的DTG健康状态评估

An Health Condition Assessment Model Based on Immune Genetic Algorithm and Ordinal Support Vector Class Machine for DTG

针对无监督情况下动调陀螺仪健康状态评估问题,提出一种两阶段健康状态评估模型.首先,在利用HHT数据预处理的基础上,为克服FCM算法初值敏感及遗传算法过早收敛等问题,提出一种加权免疫遗传模糊C均值聚类模型;其次,针对聚类结果数据的顺序性,建立基于顺序支持向量分类机的健康状态评估模型.实例结果表明提出的聚类模型具有较高的收敛精度和收敛速度,评估模型具有较高的准确率和效率.

Aiming at the problem of health condition assessment for dynamically tuned gyroscope under unsupervised case,a two-stage health assessment model was proposed.First,by utilizing HHT method to preprocess data to overcome problems of sensitivity to center selection of FCM and prematurity of genetic algorithm,a weighted immune genetic fuzzy C-means clustering model was introduced.Then,according to the order of the clustering result data,an assessing health condition model based on ordinal support vector class machine was proposed.The experimental results show that the clustering model has higher convergence precision and speed;and the assessment model has higher accuracy and efficiency.