纸浆洗涤过程稳态优化控制方法  

Stable-state Optimal Control for Pulp Washing Process

作　　者：单文娟 程晓云 汤伟[2] Shan Wenjuan;Cheng Xiaoyun;Tang Wei(Xi’an Aeronautical University,Xi’an 710077,China;Shaanxi University of Science&Technology,Xi’an 710021,China)

机构地区：[1]西安航空学院,西安710077 [2]陕西科技大学,西安710021

出　　处：《中华纸业》2022年第6期13-18,共6页China Pulp & Paper Industry

基　　金：国家自然科学基金项目(No.62073206);西安航空学院校级科研基金一般项目(No.2020KY0206)。

摘　　要：为了研究纸浆洗涤工段中工艺指标的矛盾情况,构建多目标优化,提高出浆量,降低清水使用量,达到高产、节能的目的。本文基于洗涤工艺过程的多段逆流特性,搭建了非线性数学模型,提出一种解决非线性问题的优化策略。引入自适应的惩罚因子,将带约束条件的模型转化为无约束问题,利用人工神经网络理论,对类似于灰箱模型的纸浆洗涤过程建立稳态模型,对于多目标模型,采用混沌优化算法进行目标优化。计算机仿真表明,稳态优化模型不仅保障了纸浆洗涤干净度,并且提高了洗后纸浆的出浆量。造纸厂实践表明,该方法能有效解决洗浆过程的工艺指标矛盾问题,整个过程得到了优化控制。该方案提高了出浆量,节约了洗涤用水,达到了节能减耗的目的。本文提出的稳态优化控制方法适应于大部分工业生产过程,对于非线性、多目标问题具有通用性,混沌优化算法可解决灰箱问题的求解精度、结果可靠性等要求,避免了局部最优的缺点。The purpose of this paper is to study the multiobjective optimization of paper washing process, in order to improve the pulp output, reduce the consumption of clean water, and achieve the purpose of high yield and energy saving. Based on the characteristics of pulp washing process, a new method was proposed for solving nonlinear constrained optimization problem,making the constraint problem into an unconstrained problem by introduction of adaptive penalty factor, then based on the stablestate neural network model of pulp washing, chaotic optimization algorithm was used for the multi-objective optimal control of pulp washing process. Results showed that the method is feasible in solving the optimal control of pulp washing process, it can bring greater economic benefits to pulp and paper enterprises. The steady-state optimal control method of pulp washing process can effectively solve the multi-objective problem of nonlinear and complex production process. It has good universality in solving accuracy, meeting constraints and reliability, and is easy to find the global optimal advantage.

关 键 词：纸浆洗涤过程 稳态优化 罚函数 混沌优化 神经网络 

分 类 号：TS736[轻工技术与工程—制浆造纸工程]