有色冶金设计与研究
有色冶金設計與研究
유색야금설계여연구
NONFERROUS METALS ENGINEERING & RESEARCH
2014年
4期
8-11
,共4页
高炉料位%机械探尺%雷达探尺%模糊GK聚类%RBF网络
高爐料位%機械探呎%雷達探呎%模糊GK聚類%RBF網絡
고로료위%궤계탐척%뢰체탐척%모호GK취류%RBF망락
stock line of blast furnace%mechanical gauge rod%radar gauge rod%fuzzy GK clustering%RBF neural network
通过分析机械探尺和雷达探尺在高炉料位检测上的优缺点，采用模糊GK聚类算法，对雷达探尺测量数据进行聚类处理，并把聚类获得的参数用于构建一个RBF网络。利用机械探尺数据训练已建立的RBF网络，建立了基于机械探尺数据的修正模型，通过修正模型对雷达探尺测量数据的逐一修正，实现雷达探尺和机械探尺测量数据的有机融合。仿真结果和工业数据证明，基于机械探尺数据建立的修正模型具有较高的精度和较好的实用价值。
通過分析機械探呎和雷達探呎在高爐料位檢測上的優缺點，採用模糊GK聚類算法，對雷達探呎測量數據進行聚類處理，併把聚類穫得的參數用于構建一箇RBF網絡。利用機械探呎數據訓練已建立的RBF網絡，建立瞭基于機械探呎數據的脩正模型，通過脩正模型對雷達探呎測量數據的逐一脩正，實現雷達探呎和機械探呎測量數據的有機融閤。倣真結果和工業數據證明，基于機械探呎數據建立的脩正模型具有較高的精度和較好的實用價值。
통과분석궤계탐척화뢰체탐척재고로료위검측상적우결점，채용모호GK취류산법，대뢰체탐척측량수거진행취류처리，병파취류획득적삼수용우구건일개RBF망락。이용궤계탐척수거훈련이건립적RBF망락，건립료기우궤계탐척수거적수정모형，통과수정모형대뢰체탐척측량수거적축일수정，실현뢰체탐척화궤계탐척측량수거적유궤융합。방진결과화공업수거증명，기우궤계탐척수거건립적수정모형구유교고적정도화교호적실용개치。
By analyzing the characteristics of the mechanical gauge rod and the radar gauge rod on the stock line measurement of the blast furnace, the Fuzzy GK Clustering Method is firstly adopted to realize the measurement data clustering of the radar gauge rod. Then, the parameters obtained by the Fuzzy GK Clustering Method is used to establish a RBF neural network (RBFNN), and the RBFNN is trained by the measurement data of the mechanical gauge rod. Finally, a correction model based on mechanical gauge rod data is built to correct the radar gauge rod data, which can realize the effectively fusion of the mechanical gauge rod data and radar gauge rod data. The proposed method overcomes the discontinuous measurement of the stock line of the mechanical gauge rod, and weak anti-disturbance ability, high accuracy fluctuations and poor stability of the measurement of radar gauge rod. Both the simulation results and industrial validation show that the proposed correction model has high accuracy and good practical value for industrial production.
