基于ISABO-IBiLSTM模型的刀具磨损预测方法

针对现有的刀具磨损预测方法因为缺少优化算法及网络结构不完善而导致预测精度不高的问题,提出了一种将改进的减法优化器(SABO)算法和改进的双向长短时记忆(BiLSTM)网络相结合的刀具磨损状态预测模型(ISABO-IBiLSTM模型)。首先,采用截断法、Hampel滤波法、改进的完全自适应噪声集合经验模态分解(ICEEMDAN)-改进的小波阈值降噪法对加速度振动信号与力信号数据进行预处理。然后,提取预处理后的信号数据的时域、频域、时频域特征,并通过斯皮尔曼和最大互信息相关系数筛选特征,构建模型的输入。最后,利用改进的SABO算法对改进后的BiLSTM网络进行参数寻优,基于所得到的优化参数训练网络实现磨损预测。实验数据分析结果表明,所提出的ISABO-IBiLSTM模型对刀具磨损量的预测精度为98.49%~98.83%,较BiLSTM模型、改进的BiLSTM模型、改进的卷积神经网络(ICNN)-BiLSTM模型有了较大的提高。

Application of SABO-VMD-KELM in Fault Diagnosis of Wind Turbines

In order to improve the accuracy of wind turbine fault diagnosis,a wind turbine fault diagnosis method based on Subtraction-Average-Based Optimizer(SABO)optimized Variational Mode Decomposition(VMD)and Kernel Extreme Learning Machine(KELM)is proposed.Firstly,the SABO algorithm was used to optimize the VMD parameters and decompose the original signal to obtain the best modal components,and then the nine features were calculated to obtain the feature vectors.Secondly,the SABO algorithm was used to optimize the KELM parameters,and the training set and the test set were divided according to different proportions.The results were compared with the optimized model without SABO algorithm.The experimental results show that the fault diagnosis method of wind turbine based on SABO-VMD-KELM model can achieve fault diagnosis quickly and effectively,and has higher accuracy.

基于优化变分模态分解和包络峭度的轴承故障诊断

针对变分模态分解(VMD)的分解层数K和惩罚因子α难以选择问题,提出了用减法平均优化器(SABO)对参数寻优的方法。首先,采用SABO对K和α进行寻优,输出最优参数组合并代入到VMD中,将原始振动信号分解得到K个模态分量;然后,用最大包络峭度为指标提取K个模态分量中峭度最大的分量作为最优分量,并计算其相关时域和熵理论特征参数构造特征向量样本集;最后,将特征向量样本集输入到经网格搜索和五折交叉验证调参的支持向量机(SVM)中进行故障诊断。为了验证该方法的有效性,利用凯斯西储大学轴承数据集进行实验,实验结果表明:该方法分类效果更好,准确率达到99.44%;基于江南大学3种不同工况的轴承数据实验,最终故障诊断准确率都达到了95%以上。