基于改进灰色-多元回归组合预测模型的燃煤电厂智慧水务研究

doi:10.16085/j.issn.1000-6613.2020-0681

摘要/Abstract

摘要：

以某燃煤电厂水务系统为研究对象，对机组运行参数和水量历史数据进行筛选和关联性分析，根据前期水平衡测试结果，结合响应面分析验证，发现机组负荷、蒸发损失系数、浓缩倍率和循环水温升这四个因素能够对全厂供水量产生关键性影响。基于灰色理论和多元非线性回归分析，分别建立各因素的灰色预测模型GM(1, 1)，再将灰色模型预测值作为自变量输入到多元非线性回归方程中，得到了改进灰色-多元非线性回归组合的供水量预测模型，其模型拟合度R²为0.913且与真实值的平均相对误差为6.9%左右，实现了灰色模型和回归模型优势互补，有效地预测该电厂供水量未来变化；而供水量预测是智慧水务建设的关键所在，是水务管理和智能调度的主要依据，也是实现供水管网漏损和仪表故障报警的重要途径。

关键词: 灰色模型, 多元非线性回归, 供水量预测, 管网漏损, 智慧水务

Abstract:

The water affairs of a power plant has been taken as a research subject, its operation parameters associated with water history datas is analyzed and screened out, according to the previous water balance test results and validation with response surface analysis method, it is found that the generation load, coefficient of water evaporation loss, concentration ratio and rise of circulating water inlet and outlet temperature can make important influence on water supply demand. Based on the classical grey theory and multivariate nonlinear regression analysis, grey models GM(1, 1) are firstly used to forecast four impact factors, its predictive values are used as independent variable into equation, then the improved combination prediction model of grey system and multivariate nonlinear regression of water supply is obtained. The regression model of fitting accuracy(R²) come up to 0.913, the mean relative error between the predicted value and the truth is about 6.9%. That model can achieve complementary advantages of grey model and regression model, and effectively forecast the future tendency of water supplies. Water supply prediction is the key part of the construction of smart water affair of coal-fired power plant, which is the major basis for water management and intelligent distribution, also is an important way to achieve the detection of pipeline leakage and warning of monitoring instrument failure.

Key words: grey model, multivariate nonlinear regression, water supply prediction, pipeline leakage, smart water affair

中图分类号:

TM621

熊远南. 基于改进灰色-多元回归组合预测模型的燃煤电厂智慧水务研究[J]. 化工进展, 2020, 39(S2): 393-400.

Yuannan XIONG. Smart water affair of coal-fired power plant based on improved combination prediction model of grey system and regression analysis[J]. Chemical Industry and Engineering Progress, 2020, 39(S2): 393-400.

图/表 7

参考文献 18

1	韩买良. 火力发电行业用水分析及对策[J]. 工业水处理，2010, 30(2): 4-7.
	HAN Mailiang. Analysis and countermeasures on water consumption in thermal power plants[J]. Industrial Water Treatment, 2010, 30(2): 4-7.
2	潘荔, 刘志强, 张博. 中国火电节水现状分析及措施建议[J]. 中国电力，2017, 50(11): 158-163.
	PAN Li, LIU Zhiqiang, ZHANG Bo. Comprehensive analysis and related measures on current situation of water saving of thermal power generation in China[J]. Electric Power, 2017, 50(11): 158-163.
3	刘锁祥, 赵顺萍, 曹楠, 等. 供水管网漏损控制研究和实践[J]. 中国给水排水，2015, 31(10): 22-25.
	LIU Suoxiang, ZHAO Shunping, CAO Nan, et al. Research and practice of water loss control of water distribution network[J]. China Water & Wastewater, 2015, 31(10): 22-25.
4	杨宝红. 新形势下火电厂节水减排工作特点及关键[J]. 热力发电，2016, 45(9): 95-99.
	YANG Baohong. Features and key process of water saving and wastewater discharge reduction in thermal power plants at current situation [J]. Thermal Power Generation, 2016, 45(9): 95-99.
5	朱晓庆, 殷峻暹, 张丽丽, 等. 深圳市智慧水务应用体系研究[J]. 水利水电技术，2019, 50(S2): 176-180.
	ZHU Xiaoqing, YIN Junxian, ZHANG Lili, et al. Study on smart water application system in Shenzhen[J]. Water Resources and Hydropower Engineering, 2019, 50(S2): 176-180.
6	张晔明, 罗贤伟, 蒋怀德. 基于函数模型的水务智慧建模与应用[J]. 净水技术, 2019, 38(S2): 144-149.
	ZHANG Yeming, LUO Xianwei, JIANG Huaide. Smart modeling and application of water affairs based on function model[J]. Water Purification Technology，2019, 38(S2): 144-149.
7	金亚飚. 工业智慧水务的研究和探索[J]. 工业水处理，2020, 40(2): 11-13.
	JIN Yabiao. Research and exploration of industrial intelligent water works[J].Industrial Water Treatment, 2020, 40(2): 11-13.
8	谢丽芳, 邵煜, 马琦, 等. 国内外智慧水务信息化建设与发展[J]. 给水排水，2018, 54(11): 135-139.
	XIE Lifang, SHAO Yu, MA Qi, et al. Development and analysis of smart water system in China and abroad[J]. Water & Wastewater Engineering, 2018, 54(11): 135-139.
9	MILLS D A. Energy demands on water resources:analysis of competitiveconcerns[M]. Hauppauge, New York: Nova Science Publishers, 2011: 334-339.
10	MALCOLM A, BRUCE C, WEI C W. The performance of the urban water and wastewater sectors in Australia [J]. Utilities Policy, 2012, 20(1): 52-63.
11	TIWARI M K, ADAMOWSKI J. Urban water demand forecasting and uncertainty assessment using ensemble wavelet-bootstrap-neural network models[J]. Water Resources Research, 2013, 49(10): 6486-6507.
12	STEFANO A, MARCO F. Fuzzy neural networks for water level and discharge forecasting with uncertainty[J]. Environmental Modelling & Software, 2011, 26(4): 523-537.
13	BAI Y, WANG P, LI C, et al. A multi-scale relevance vector regression approach for daily urban water demand forecasting[J]. Journal of Hydrology, 2014, 517(2): 236-245.
14	邓聚龙. 灰理论基础[M]. 武汉: 华中科技大学出版社，2003.
	DENG Julong. Basic of grey theory[M]. Wuhan: Huazhong University of Science & Technology Press, 2003.
15	刘思峰, 杨英杰. 灰色系统研究进展(2004—2014)[J]. 南京航空航天大学学报，2015, 47(1): 1-18.
	LIU Sifeng, YANG Yingjie. Advances in grey system research(2004—2014)[J]. Journal of Nanjing University of Aeronautics & Astronautics, 2015, 47(1): 1-18.
16	陈民利, 高金良, 李坤, 等. 灰色模型在城市月用水量预测中的应用[J].哈尔滨商业大学学报(自然科学版)，2005, 21(6): 721-723.
	CHEN Minli, GAO Jinliang, LI Kun, et al. Application of GM(1, 1) in predicting city monthly water demands[J]. Journal of Harbin University of Commerce (Natural Sciences Edition), 2005, 21(6): 721-723.
17	YASAR A, BILGILI M, SIMSEK E. Water demand forecasting based on stepwise multiple nonlinear regression analysis[J]. Arabian Journal for Science & Engineering, 2012, 37(8): 2333-2341.
18	金菊良, 崔毅, 张礼兵, 等. 基于多智能体的城镇家庭用水量模拟预测分析[J]. 水利学报，2015, 46(12): 1387-1397.
	JIN Juliang, CUI Yi, ZHANG Libing, et al. Simulation and prediction analysis of urban household water demand based on multi-agent[J]. Journal of Hydraulic Engineering, 2015, 46(12): 1387-1397.

预测精度等级	小误差概率P	后验差比值C
优良(一级)	＞0.95	＜0.35
合格(二级)	＞0.8	＜0.5
勉强(三级)	＞0.7	＜0.65
不合格(四级)	≤0.7	≥0.65

预测精度等级	小误差概率P	后验差比值C
优良(一级)	＞0.95	＜0.35
合格(二级)	＞0.8	＜0.5
勉强(三级)	＞0.7	＜0.65
不合格(四级)	≤0.7	≥0.65

月份	进塔气温/℃	平均负荷/MW	蒸发损失系数K/℃^-1	循环水浓缩倍率	循环水温升/℃	厂区每月总取水量/t
2019.1	-5.5	653	0.000890	3.5	8.9	618961
2019.2	2.9	798	0.001058	4.4	9.0	745164
2019.3	5.2	724	0.001104	4.5	9.1	890625
2019.4	11.8	723	0.001196	3.8	9.6	933904
2019.5	15.7	674	0.001294	3.7	10.2	998627
2019.6	22.6	786	0.001391	3.4	10.3	1641500
2019.7	28.9	1027	0.001518	3.4	10.2	1802109
2019.8	35.2	1008	0.001644	3.5	10.5	2121371
2019.9	23.6	976	0.001472	3.1	10.6	2255383
2019.10	21.5	1035	0.001430	3.3	10.3	2323354
2019.11	15.3	964	0.001346	3.3	10.2	1957250
2019.12	4.5	1025	0.001190	3.1	10.0	1337604

月份	进塔气温/℃	平均负荷/MW	蒸发损失系数K/℃^-1	循环水浓缩倍率	循环水温升/℃	厂区每月总取水量/t
2019.1	-5.5	653	0.000890	3.5	8.9	618961
2019.2	2.9	798	0.001058	4.4	9.0	745164
2019.3	5.2	724	0.001104	4.5	9.1	890625
2019.4	11.8	723	0.001196	3.8	9.6	933904
2019.5	15.7	674	0.001294	3.7	10.2	998627
2019.6	22.6	786	0.001391	3.4	10.3	1641500
2019.7	28.9	1027	0.001518	3.4	10.2	1802109
2019.8	35.2	1008	0.001644	3.5	10.5	2121371
2019.9	23.6	976	0.001472	3.1	10.6	2255383
2019.10	21.5	1035	0.001430	3.3	10.3	2323354
2019.11	15.3	964	0.001346	3.3	10.2	1957250
2019.12	4.5	1025	0.001190	3.1	10.0	1337604

预测参数	月平均负荷	蒸发损失系数K	循环水浓缩倍率	循环水温升
P	1	0.83	0.87	0.81
C	0.28	0.36	0.19	0.38
评估结果	优良	合格	合格	合格
相对误差	0.078	0.079	0.031	0.013