基于强化学习的重介质选煤过程优化控制

doi:10.11799/ce202201025

煤炭工程 ›› 2022, Vol. 54 ›› Issue (1): 137-141.doi: 10.11799/ce202201025

基于强化学习的重介质选煤过程优化控制

胡金良¹,李彤昀²,王光辉³

1. 国能准能集团科学技术研究院
2. 中国矿业大学
3. 中国矿业大学化工学院

收稿日期:2021-08-04 修回日期:2021-10-24 出版日期:2022-01-14 发布日期:2022-07-07
通讯作者: 胡金良 E-mail:67141171@qq.com

Optimal control of dense medium coal preparation process based on reinforcement learning

Received:2021-08-04 Revised:2021-10-24 Online:2022-01-14 Published:2022-07-07
Contact: Jinliang Hu E-mail:67141171@qq.com

摘要/Abstract

摘要： 决定重介质选煤产品质量的主要影响因素是重介质悬浮液密度。但是由于过程复杂，设备众多，呈强非线性特性，导致对实现重介质悬浮液密度的优化控制存在难点。为此，针对重介质选煤过程及其特性，提出了一种基于强化学习的优化控制方法，用于在线更新密度设定值。所提方法将策略提升和策略评价两步迭代采用不同的神经网络实现，建立了无模型的控制器。最后，在MATLAB仿真平台上，将该方法与传统PI控制方法相比较，验证了所提方法的有效性。

关键词: 重介质选煤, 悬浮液密度, 强化学习, 优化控制

Abstract: The main influence factor on the quality of dense medium coal preparation product is the density of dense medium suspension. However, due to the complex process, numerous equipment and strong nonlinear characteristics, it is difficult to realize the optimal control of dense medium suspension density. In this paper, an optimal control method based on reinforcement learning is proposed to update the density setting-value online. The two step iterations of strategy updating and strategy evaluation are implemented by different neural networks. And a model-free controller is established. Finally, on the MATLAB simulation platform, the proposed method is compared with the traditional PI control method to verify the effectiveness of the proposed method.

中图分类号:

TD94

胡金良, 李彤昀, 王光辉. 基于强化学习的重介质选煤过程优化控制[J]. 煤炭工程, 2022, 54(1): 137-141.

参考文献

[1] ZHANG Wenjun, OU Zeshen. Reliability design and methods of reliability distributing of coal preparation system. Journal of China Coal Society, 2000, 25(5): 542-545.
(张文军, 欧泽深. 选煤工艺系统的可靠性设计与可靠度分配. 煤炭学报, 2000, 25(5): 542-545.)
[2] NAPIER-MUNN T. The dense medium cyclone-past, present and future. Minerals Engineering, 2018, 116: 107 – 113.
[3] 陈志强,徐春江,齐正义,黄亚飞.刮板运行速度对浅槽重介分选机流场的影响[J].煤炭科学技术,2014,42(02):117-119.
[4] ZHANG L J, XIA X H. A model predictive control for coal beneficiation dense medium cyclones. IFAC Proceedings Volumes, 2014, 47(3): 9810 - 9815.
[5] ZHANG L J, XIA X H, ZHU B. A dual-loop control system for dense medium coal washing processes with sampled and delayed measurements. IEEE Transactions on Control Systems Technology, 2017, 25(6): 2211 – 2218.
[6] MEYER E J, CRAIG I K. The development of dynamic models for a dense medium separation circuit in coal beneficiation. Minerals Engineering, 2010, 23(10): 791 – 805.
[7] MEYER E J, CRAIG I K. Coal dense medium separation dynamic and steady-state modelling for process control. Minerals Engineering, 2014, 65(2): 98 – 108.
[8] WANG Z, KUANG Y, DENG J, et al. Research on the intelligent control of the dense medium separation process in coal preparation plant. International Journal of Mineral Processing, 2015, 142: 46 – 50.
[9] CAO Zhenguan. Study of prediction control on heavy medium density in the process of coal preparation. Xuzhou: China University of Mining and Technology, 2014.
(曹珍贯. 重介选煤过程中重介质的密度预测控制研究. 徐州: 中国矿业大学, 2014.)
[10] 邱佳楷, 王然风, 付翔. 重介质悬浮液密度宽域智能控制系统设计[J]. 工矿自动化, 2019, 45(7): 33-37.
[11] 代伟,张凌智,褚菲,马小平.重介质选煤过程模型与数据混合驱动的自适应运行反馈控制[J].控制理论与应用,2020,37(02):283-294.S.
[12] Wei Dai,Lingzhi Zhang,Jun Fu,Tianyou Chai,Xiaoping Ma. Model-data-based switching adaptive control for dense medium separation in coal beneficiation[J]. Control Engineering Practice,2020,98.
[13] 胡金成. 结合主动学习策略的在线自适应半监督软测量建模方法[D].中国矿业大学,2020.
[14] Dual-Rate Adaptive Optimal Tracking Control for Dense Medium Separation Process Using Neural Networks.[J]. IEEE transactions on neural networks and learning systems,2020,PP.
[15] Lewis F L, Vrabie D, Syrmos, V L. Optimal control[M]. John Wiley & Sons, 2012.
[16] Kiumarsi B, Lewis F L. Actor–Critic-Based Optimal Tracking for Partially Unknown Nonlinear Discrete-Time Systems[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 26(1):140-151.
[17] Bertsekas D P, Tsitsiklis J N. Neuro-dynamic programming: an overview[C]. IEEE Conference on Decision & Control. IEEE, 1995,1: 560-564.
[18] Igelnik B, Pao Y H. Stochastic Choice of Basis Functions in Adaptive Function Approximation and the Functional-Link Net[J]. IEEE Transactions on Neural Networks, 1995, 6(6): 1320-1329.
[19] Selmic R R, Lewis F L. Deadzone compensation in motion control systems using neural networks[J]. IEEE Transactions on Automatic Control, 2000, 45(4): 602-613.
[20] Sutton R S. Learning to predict by the methods of temporal differences[J]. Machine Learning, 1998, 3(1): 9-44.

基于强化学习的重介质选煤过程优化控制

Optimal control of dense medium coal preparation process based on reinforcement learning

PDF (Mobile)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 3

编辑推荐

Metrics

本文评价

[1]	汪进雅，张孝松，李强，等. 霍尔辛赫选煤厂重介质消耗控制研究与实践[J]. 煤炭工程, 2015, 47(8): 76-79.
[2]	乔继. 沙曲选煤厂工艺优化设计和布置探讨[J]. 煤炭工程, 2015, 47(12): 22-24.
[3]	牛国前. 巴关河选煤厂重介工艺系统改造与实践[J]. 煤炭工程, 2012, 44(9): 56-58.