论文部分内容阅读
针对故障诊断中故障类样本难于获取以及不均衡类问题,提出了基于粒子群和滑动窗口的支持向量数据描述(M-SVDD)故障诊断方法.该方法利用粒子群优化支持向量数据描述的核参数,同时引入滑动窗口技术,通过大窗口大小来控制故障诊断模型的训练样本数,根据小窗口的预测误差变化动态调整大窗口的大小.采用该方法对铜转炉吹炼过程进行故障诊断的实验结果表明,该方法能有效抑制过拟合现象,具有故障敏感性高、泛化能力强等特点.
In order to solve the problem of difficult to obtain fault samples and unbalanced classes in fault diagnosis, this paper proposes a fault diagnosis method based on particle swarm optimization and sliding window support vector data description (M-SVDD). This method uses particle swarm optimization to support the kernel parameters described by vector data , And the sliding window technology is introduced to control the number of training samples of the fault diagnosis model by the size of the large window and dynamically adjust the size of the large window according to the prediction error of the small window.This method is used to diagnose the failure of the copper converter blowing process The results show that this method can effectively suppress the over-fitting phenomenon and has the characteristics of high fault sensitivity and generalization ability.