改进的布谷鸟算法优化极限学习机的石化轴承故障分类

针对通用的智能故障诊断方法在石化滚动轴承中准确率不理想的问题,提出一种通过改进的布谷鸟算法( CS )优化极限学习机( ELM )使诊断准确率提高的模型。将实测轴承振动信号降噪处理,计算不同嵌入维度下的关联维数作为 ELM 的输入信号;通过改进的布谷鸟算法获取极限学习机最优的隐含层偏置、输入权重,最后输出诊断结果。经过实验证明,该方法可以有效地克服测量信号时的干扰,可以对不同故障下的滚动轴承准确识别,并与多种模型对比,该方法的故障诊断准确率为 97.5% 。

Improved Cuckoo Algorithm for Optimizing Extreme Learning Machine for Petrochemical Bearing Fault Classification

Aiming at the problem of unsatisfactory accuracy of the general intelligent fault diagnosis method in petrochemical rolling bearings , a model is proposed to optimize the limit learning machine ( ELM ) by the improved cuckoo intelligent optimization algorithm ( CS ) to make the diagnosis accuracy improved.The measured bearing vibration signal is processed by noise reduction , and the correlation dimensions under different embedding dimensions are calculated as features as the input signal of ELM ; the optimal implied layer bias and input weight output diagnosis results of the limit learning machine are obtained by the improved cuckoo algorithm.The experimental results prove that the method proposed in this paper can effectively overcome the interference when measuring the signal , and can accurately identify the rolling bearing under different faults , and compared with a variety of models , the fault diagnosis accuracy of the method proposed is 97.5%.