基于DBN-ELM的球磨机料位软测量方法研究  

Soft Sensor for Ball Mill Fill Level Based on DBN-ELM Model

作　　者：康岩[1] 卢慕超[1] 阎高伟[1] 

机构地区：[1]太原理工大学信息工程学院,山西太原030024

出　　处：《仪表技术与传感器》2015年第4期73-75,92,共4页Instrument Technique and Sensor

基　　金：山西省自然科学基金项目(2011011012-2);国家863项目(2013AA102306)

摘　　要：针对采用传统方法建立球磨机料位软测量模型存在测量精度不高和稳定性较低的缺点,提出一种结合深度信念网络和极限学习机的软测量方法。该方法以球磨机轴承振动信号为辅助变量,采用深度信念网络进行振动信号功率谱的特征提取,然后将提取的有效特征输入极限学习机进行模型训练,得到软测量模型。最后在小型实验室球磨机上进行试验和模型验证。结果表明,该方法与传统方法相比具有较高的测量精度和较好的稳定性。To solve the issue of low accuracy and weak stability of the traditional measurement method of ball mill fill level, a novel approach based on deep belief network and extreme learning machine was proposed.The vibration signal of ball mill bearing was selected as the instrumental variable. Deep belief network was employed to extract effective features from the power spectrum of vibration signal.Then effective features were put to the learning machine to proceed model training to obtain the soft sensor model. Lastly, the experiments were carried out on the lab-scale ball mill to validate the proposed method.The results show that the pro- posed method is more accuracy and stable than the traditional method.

关 键 词：球磨机料位 深度信念网络 特征提取 极限学习机 软测量 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]