摘要
转炉炼钢生产过程数据特征选择是实现终点碳温预报的关键,针对生产过程高维数据不利于快速精确预测终点碳温的问题,提出一种改进遗传算法的转炉炼钢生产过程数据特征选择方法。首先采用皮尔逊相关系数衡量不同特征的重要贡献度,进而构造反映过程数据特征与终点碳温相关性的目标函数;然后通过目标函数定义了种群的最大、最小、平均适应度和随机个体适应度值4个变量,建立了一种自适应调节交叉变异概率机制,使得迭代寻优时种群分布更加合理的同时又提高了算法后期收敛速度,防止陷入局部最优。最后进行实际钢厂生产过程数据特征选择验证和对比实验,结果表明,特征选择平均用时为0.25 s,用于终点预报中温度误差在±5℃的精度为85.67%,碳含量预测误差在±0.01%的精度为80.67%。
Data feature selection of converter steelmaking process is the key step to realize the end point carbon content and temperature prediction.The high-dimensional data of production process are not conducive to the rapid and accurate prediction of the end point carbon temperature.To address this problem,an improved genetic algorithm is proposed to select the data feature of converter steelmaking process.Firstly,Pearson correlation coefficient is used to measure the important contribution of different features.Then,the objective function is formulated to reflect the correlation between process data feature and terminal carbon temperature.The maximum,minimum,average fitness and random individual fitness of the population are defined by the objective function.In this way,an adaptive crossover mutation probability mechanism is established.This method not only makes the population distribution more reasonable during the iteration optimization,but also improves the late convergence speed to prevent the algorithm from falling into local optimization.Through verification and comparison experiments of data feature selection in actual steel mills,results show that the average time of feature selection is 0.25 s,the accuracy of temperature error within±5℃in terminal prediction is 85.67%,and the accuracy of carbon content prediction error within±0.01%is 80.67%.
作者
刘辉
曾鹏飞
巫乔顺
陈甫刚
Liu Hui;Zeng Pengfei;Wu Qiaoshun;Chen Fugang(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650500,China;Yunnan Key Laboratory of Artificial Intelligence,Kunming University of Science and Technology,Kunming 650500,China;Yunnan Kungang Electronic Information Technology Co.Ltd.,Kunming 650500,China)
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2019年第12期185-195,共11页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金(61863018)项目资助.
关键词
转炉炼钢
特征选择
碳温预测
改进遗传算法
converter steel-making
feature selection
carbon temperature prediction
improved genetic algorithm
作者简介
刘辉,2012年于昆明理工大学获得博士学位,现为昆明理工大学副教授。主要研究方向为数据分析、图像处理、模式识别。E-mail:liuhui621@126.com;通信作者:曾鹏飞,2017年于华中科技大学武昌分校获得学士学位,现为昆明理工大学硕士研究生,主要研究方向为数据分析、图像处理、模式识别。E-mail:1431919050@qq.com
