摘要
针对采用音频法测量球磨机料位时存在特征值随机性强等不确定性因素,引入云理论进行球磨机料位概念表示,并利用云模型实现球磨机料位测量。其过程是对数据进行预处理并提取特征值,然后采用逆向云算法对不同料位下振声信号的特征进行基本料位概念提取,经过概念提升成粗粒度的料位概念表示后,形成不确定性推理的前件云模型;同时依据料位值信息构造推理后件云模型,以此建立云不确定性推理规则集。最后,通过云模型规则推理实现球磨机料位的软测量。多种方法对比实验结果说明了模型的有效性和实用性。
Aiming at the existence of uncertainty factors like randomness in the process of measuring fill level （FL） of ball mill by analyzing acoustic signals, cloud model system was utilized to represent the FL concepts, meanwhile, the uncertainty reasoning model was built to measure FL of ball mill. The features were extracted from the preprocessed data. Then the backward cloud generator was employed to generate the basic FL concepts and build the antecedent clouds by synthesizing the basic FL concepts. Simultaneously, the consequent clouds were built by FL values. Thus the cloud reasoning rule base was established. At last, FL soft measurement of ball mill was realized by cloud model reasoning. The comparative experiments have proved that this model is feasible and practical.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2014年第14期2281-2288,共8页
Proceedings of the CSEE
基金
国家自然科学基金项目(60975032)
山西省自然科学基金项目(2011011012-2)
山西省青年学术带头人支持计划(TYAL)
山西省科技攻关项目(20110321036)~~
关键词
球磨机料位
测量
概念表示
云模型
音频法
ball mill fill level
measurement
conceptrepresentation
cloud model
acoustic method
