PDF全文下载：
201905015.pdf
文章编号： 16726987（2019）05010106； DOI： 10.16351/j.16726987.2019.05.015
程换新, 黄震， 骆晓玲
(青岛科技大学 自动化与电子工程学院，山东 青岛 266061）
摘要： 在热电厂脱硫过程中，pH值直接影响脱硫的效率，若pH测量仪器受到环境的影响被破坏，会给生产造成巨大的损失。为了降低这种损失，采用改进的孪生支持向量机回归模型对pH值进行预测，首先将粒子群算法的权值和学习因子进行改进，然后用改进之后的粒子群算法对孪生支持向量机回归模型的惩罚参数和核函数的参数等进行寻优，再将最优的参数代入孪生支持向量机预测模型中，最后用MATLAB工具箱对pH值历史数据进行仿真，并与未改进的孪生支持向量机和BP神经网络预测技术进行比较。结果表明：该方法对脱硫系统中pH值的预测精度高，平均相对误差比未改进的孪生支持向量机和BP神经网络的预测结果小，能够显著改善脱硫装置的效率。
关键词： 粒子群算法； 孪生支持向量机； BP神经网络； pH值
中图分类号： TP 181文献标志码： A
引用格式： 程换新, 黄震， 骆晓玲. 基于改进孪生支持向量机的热电厂脱硫系统pH值预测\[J\]. 青岛科技大学学报（自然科学版）， 2019， 40（5）： 101106.
CHENG Huanxin，HUANG Zhen, LUO Xiaoling. pH prediction desulfurization system in thermal power plant based on improved twin support vector machine\[J\]. Journal of Qingdao University of Science and Technology（Natural Science Edition), 2019, 40(5)： 101106.
pH Prediction Desulfurization System in Thermal Power Plant Based on
Improved Twin Support Vector Machine
CHENG Huanxin， HUANG Zhen, LUO Xiaoling
(College of Automation and Electrical Engineering, Qingdao University of Science and Technology, Qingdao 266061, China)
Abstract: In the desulfurization process of thermal power plants, the pH value directly affects the efficiency of desulfurization. If the pH measuring instrument is damaged by the environment, it will cause huge losses to the production. In order to reduce this loss, the improved twin support vector machine regression model is used to predict the pH value. Firstly, the weight and learning factor of the particle swarm optimization algorithm are improved, and then the improved particle swarm optimization algorithm is used to support the support vector machine regression model. The penalty parameters and kernel function parameters are optimized, and then the optimal parameters are substituted into the twin support vector machine prediction model. Finally, the MATLAB toolbox is used to simulate the pH value historical data, and the unmodified twin support vector machine and BP neural network prediction techniques are compared. The results show that the method has high prediction accuracy for the pH value in the desulfurization system, and the average relative error is smaller than that of the unmodified twin support vector machine and BP neural network, which can significantly improve the efficiency of the desulfurization device.
Key words: particle swarm optimization； twin support vector machine； BP neural network； pH value
收稿日期： 20181123
基金项目： 国家海洋局重大专项项目（2016496）.
作者简介： 程换新(1966—), 男, 教授.
