摘要
针对采用振动法测量球磨机料位时,振动信号和料位之间存在非线性和时变性,采用传统方法存在测量精度低、稳定性差的问题,提出基于T-S模糊模型的球磨机料位表示和测量的方法。首先利用减法聚类对振动信号的功率谱特征值进行模糊前件辨识,确定模糊概念和规则数;再用最小二乘估计辨识后件参数;最后,利用模糊推理方法实现球磨机料位的软测量。在小型球磨机上的试验结果验证了T-S模糊模型对球磨机料位测量的有效性,与传统方法相比,T-S模糊模型方法具有测量精度高、稳定性好的特点。
When the ball mill level is measured by vibration, the vibration signals are nonlinear and time-varying. In order to deal with the problem of low precision and poor stability in traditional methods, the T-S fuzzy model is applied to represent and measure the level of ball mill. Firstly, the subtraction clustering is used to make the fuzzy antecedent recognition of vibration signal＇s power spectral eigenvalues, and the fuzzy concepts and rules are identified. Then the minimum variance estimation is used to identify the seccedent parameters. Finally, the fuzzy reasoning method is applied to achieve the soft measurement of ball mill. Compared the experimental measurement with the traditional methods, the experimental results demonstrate the high-precious and high-stable characteristics of the T-S fuzzy model method.
出处
《计算机仿真》
CSCD
北大核心
2014年第11期328-331,344,共5页
Computer Simulation
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)
关键词
球磨机
料位
模糊模型
减法聚类
Ball mill
Fill level
Fuzzy model
Subtractive clustering
