| 陈浩洲,满奕.面向造纸生产的混合整数非线性求解框架的应用研究[J].中国造纸,2026,45(9):162-173 |
二维码（扫一下试试看！） |
| 面向造纸生产的混合整数非线性求解框架的应用研究 |
| Application Research of a Mixed Integer Nonlinear Solving Framework for Paper Production |
| 收稿日期：2026-03-02 修订日期：2026-04-30 |
| DOI：10.11980/j.issn.0254-508X.2026.09.019 |
| 关键词: 混合整数非线性规划 可行泵算法 外部逼近算法 造纸生产优化 |
| Key Words:mixed integer nonlinear programming feasible pump algorithm external approximation algorithm paper production optimization |
| 基金项目:国家自然科学基金项目22478141国家自然科学基金项目（22478141）。 |
|
| 摘要点击次数: 386 |
| 全文下载次数: 59 |
| 摘要:造纸生产系统转型升级过程涉及高维度、非线性以及离散和连续变量耦合的实时优化问题，具有约束复杂、搜索空间庞大等难点，导致在处理此类问题时，存在无效迭代多以及求解效率不足等情况，研究面向造纸生产数学模型的求解框架是解决这些问题的关键。本研究基于可行泵算法与外部逼近算法，设计了一种面向造纸生产的全局混合整数非线性求解框架，实现了对于复杂生产约束和动态实时优化问题的高效求解。结果表明，本求解框架在造纸企业30天的备浆错峰用电调度案例中，备浆加权电价平均降低约7.32%，减少了78.3%的迭代次数；在车间生产调度7组案例中，以100%的成功率找到最优解，生产周期和成本平均优化4.15%和4.98%，平均缩短了25.8%的计算时间。 |
| Abstract:The transformation and upgrading process of the paper production system involves real-time optimization problems with high dimensions， non-linearity， and coupling of discrete and continuous variables. It presents challenges such as complex constraints and a vast search space， leading to issues such as excessive ineffective iterations and insufficient solution efficiency when dealing with such problems. Therefore， developing a solution framework for the mathematical model of paper production is the key to solving these problems. This study， based on the feasible pump algorithm and the external approximation algorithm， designed a global mixed-integer nonlinear solution framework for paper production and achieved efficient solutions for complex production constraints and dynamic real-time optimization problems. The results showed that in a 30-day case study of pulp preparation peak load power scheduling in a paper manufacturing enterprise， the weighted average price of pulp preparation was reduced by approximately 7.32%， and the number of iterations was reduced by 78.3%. In 7 cases of workshop production scheduling， the optimal solution was found with a 100% success rate， with the production cycle and total production cost optimized by 4.15% and 4.98% on average， respectively， and the computing time was shortened by 25.8% on average. |
| 查看全文
HTML
查看/发表评论 下载PDF阅读器 |