摘要
极限学习机(extreme learning machine,ELM)具有学习速度快、测试精度高的优点。近年来被广泛用于模式识别和故障诊断等领域,但是ELM固有的随机性对其泛化性能和精度有很大的影响。蝙蝠算法(bat algorithm,BA)是一种新型的智能优化算法,具有良好的全局搜索能力。将蝙蝠算法引入到极限学习机输入权值和阈值的优化中,有机结合2种算法的优点,建立了基于蝙蝠算法优化极限学习机的故障模型,以带通滤波器作为测试电路,并和ELM、DE-ELM、SAE-ELM进行对比,仿真和实验结果表明蝙蝠算法有效地改善了ELM网络的诊断精度和泛化能力。
In recent years,Extreme learning machine is widely used in the fields of pattern recognition and fault diagnosis etc.because of quick learning speed and high testing accuracy,but randomization of extreme learning machine may lead to non-optimal performance.Therefore,bat algorithm,which has the features of global convergence,is introduced in the parameter optimization of extreme learning machine.A parameter optimization model based bat algorithm is established in order to combine the advantages of the algorithms,applied in bandpass filter analog circuit fault diagnosis,and compared with the ELM,DE-ELM,SAE-ELM.Simulation show that ELM improved by BA can achieve good generalization performance and robustness.
出处
《电子测量技术》
2015年第2期138-141,共4页
Electronic Measurement Technology
关键词
极限学习机
蝙蝠算法
故障诊断
模拟电路
extreme learning machine
bat algorithm
fault diagnosis
analog circuit