工程科学学报，第39卷.第10期：1546-1551,2017年10月 Chinese Journal of Engineering,Vol.39,No.10:1546-1551,October 2017 D0L:10.13374/j.issn2095-9389.2017.10.013;htp:/journals..usth.edu.cn BCOISOA-BP网络在磨矿粒度软测量中的应用 周 颖)四，杨京松)，付冬梅2)，岳彬) 1)河北工业大学控制科学与工程学院，天津3001302)北京科技大学自动化学院，北京100083 ☒通信作者，E-mail:zhouying.2007@163.com 摘要传统人群搜索(S0)算法通过计算搜索方向、搜索步长和搜寻更新个体位置三个步骤进行寻优.它的缺点在于计 算量大，种群之间信息交流少，导致寻优速度慢.针对人群搜索算法存在的缺点，本文提出二项交叉算子改进人群搜索算法 (BC0IS0A)对其改进.在计算搜索步长方面，本文采用随机数与最大函数值位置乘积判断子群位置，进而提高全局寻优计算 速率.在更新位置方面，本文提出二项交叉算子加强种群之间的联系，避免在更新搜索方向过程中，算法因局部最优而导致 过早收敛，进而达到快速、准确寻找最优解的目的.本文将以上二项交叉算子改进人群搜索-P神经网络算法应用在二段式 磨矿过程中，实现磨矿粒度在线软测量.仿真结果表明，与人群搜索算法和粒子群算法进行比较，二项交叉算子改进人群搜 索算法收敛速度更快，预测精度最高，满足对磨矿粒度实时检测的要求 关键词搜索步长；个体位置；二项交叉算子改进人群搜索算法；BP神经网络 分类号TG142.71 BCOISOA-BP network in grinding particle size soft sensor applications ZHOU Ying,YANG Jing-song,FU Dong-mei,YUE Bin' 1)School of Control Science and Engineering,Hebei University of Technology,Tianjin 300130,China 2)School of Automation,University of Science and Technology Beijing.Beijing 100083.China Corresponding author,E-mail:zhouying2007@163.com ABSTRACT The traditional seeker optimization algorithm (SOA)uses three steps for an optimal search:calculating the search di- rection,searching the step length,and updating the individual position.Its shortcomings are the large amount of calculation required and weak communication between populations,which results in low speed optimization.To address these disadvantages,this paper of- fers the binomial crossover operator improved seeker optimization algorithm(BCOISOA)as an improvement.In terms of computational search step length,this paper adopts a random number and maximum function product judgment subgroup location so that global opti- mization computation speed can be improved.In terms of update location,this paper puts forward two crossover operators to strengthen the connection between the populations.This avoids premature convergence of the algorithm during the process of updating the search direction,caused by the local optimum,and achieves a fast and accurate optimal solution.This article usesthe BCOISOA-BP neural network algorithm for a two-phase grinding process to achieve a grind size online soft sensor.Compared with the SOA and PSO algo- rithms,the simulation result shows that the BCOISOA algorithm has the fastest convergence speed and highest precision.It therefore satisfies the requirements of grind size real-time detection. KEY WORDS search steplength;individual position;BCOISOA;BP network 磨矿过程是将矿石研磨碾碎，将有用矿物与大量 的矿产精度指标，而且影响后续过程精矿的加工和金 的无用脉石分离，其中粒度大小的好坏不仅影响矿厂 属的回收.因此，磨矿粒度的检测尤其重要.然而磨 收稿日期：2016-12-01 基金项目：河北省高等学校科学技术研究资助项目(ZD2016071)工程科学学报,第 39 卷,第 10 期:1546鄄鄄1551,2017 年 10 月 Chinese Journal of Engineering, Vol. 39, No. 10: 1546鄄鄄1551, October 2017 DOI: 10. 13374 / j. issn2095鄄鄄9389. 2017. 10. 013; http: / / journals. ustb. edu. cn BCOISOA鄄鄄 BP 网络在磨矿粒度软测量中的应用 周 颖1) 苣 , 杨京松1) , 付冬梅2) , 岳 彬1) 1)河北工业大学控制科学与工程学院, 天津 300130 2)北京科技大学自动化学院, 北京 100083 苣通信作者, E鄄mail: zhouying2007@ 163. com 摘 要 传统人群搜索(SOA)算法通过计算搜索方向、搜索步长和搜寻更新个体位置三个步骤进行寻优. 它的缺点在于计 算量大,种群之间信息交流少,导致寻优速度慢. 针对人群搜索算法存在的缺点,本文提出二项交叉算子改进人群搜索算法 (BCOISOA)对其改进. 在计算搜索步长方面,本文采用随机数与最大函数值位置乘积判断子群位置,进而提高全局寻优计算 速率. 在更新位置方面,本文提出二项交叉算子加强种群之间的联系,避免在更新搜索方向过程中,算法因局部最优而导致 过早收敛,进而达到快速、准确寻找最优解的目的. 本文将以上二项交叉算子改进人群搜索鄄鄄BP 神经网络算法应用在二段式 磨矿过程中,实现磨矿粒度在线软测量. 仿真结果表明,与人群搜索算法和粒子群算法进行比较,二项交叉算子改进人群搜 索算法收敛速度更快,预测精度最高,满足对磨矿粒度实时检测的要求. 关键词 搜索步长; 个体位置; 二项交叉算子改进人群搜索算法; BP 神经网络 分类号 TG142郾 71 BCOISOA鄄鄄BP network in grinding particle size soft sensor applications ZHOU Ying 1) 苣 , YANG Jing鄄song 1) , FU Dong鄄mei 2) , YUE Bin 1) 1) School of Control Science and Engineering, Hebei University of Technology, Tianjin 300130, China 2) School of Automation, University of Science and Technology Beijing, Beijing 100083, China 苣Corresponding author, E鄄mail: zhouying2007@ 163. com ABSTRACT The traditional seeker optimization algorithm (SOA) uses three steps for an optimal search: calculating the search di鄄 rection, searching the step length, and updating the individual position. Its shortcomings are the large amount of calculation required and weak communication between populations, which results in low speed optimization. To address these disadvantages, this paper of鄄 fers the binomial crossover operator improved seeker optimization algorithm (BCOISOA) as an improvement. In terms of computational search step length, this paper adopts a random number and maximum function product judgment subgroup location so that global opti鄄 mization computation speed can be improved. In terms of update location, this paper puts forward two crossover operators to strengthen the connection between the populations. This avoids premature convergence of the algorithm during the process of updating the search direction, caused by the local optimum, and achieves a fast and accurate optimal solution. This article usesthe BCOISOA鄄鄄BP neural network algorithm for a two鄄phase grinding process to achieve a grind size online soft sensor. Compared with the SOA and PSO algo鄄 rithms, the simulation result shows that the BCOISOA algorithm has the fastest convergence speed and highest precision. It therefore satisfies the requirements of grind size real鄄time detection. KEY WORDS search steplength; individual position; BCOISOA; BP network 收稿日期: 2016鄄鄄12鄄鄄01 基金项目: 河北省高等学校科学技术研究资助项目(ZD2016071) 磨矿过程是将矿石研磨碾碎,将有用矿物与大量 的无用脉石分离,其中粒度大小的好坏不仅影响矿厂 的矿产精度指标,而且影响后续过程精矿的加工和金 属的回收. 因此,磨矿粒度的检测尤其重要. 然而磨