Superheated Steam Temperature Prediction Based on Long Short-Term Memory Neural Network Optimized by Novel Whale Optimization Algorithm

doi:10.12096/j.2096-4528.pgt.24127

Abstract

Abstract:

Objectives In the process of participating in flexible peak shaving, thermal power units frequently increase and decrease load, leading to significant fluctuations in superheated steam temperature, which affects the operational safety and economic performance of the units. Accurately predicting the change trend of superheated steam temperature is crucial. Therefore, a superheated steam temperature prediction model based on long short-term memory (LSTM) neural network optimized by the novel whale optimization algorithm (NWOA) is proposed. Methods Principal component analysis (PCA) is used for variable selection to eliminate redundant variables in the initial data. The whale optimization algorithm (WOA) is improved by nonlinear convergence factor adjustment strategy, adaptive weighting strategy, and dynamic spiral update strategy to enhance the accuracy and global search ability of the algorithm. An LSTM model for superheated steam temperature prediction is established for a direct current boiler in a power plant in Xinjiang. The improved whale algorithm is used to optimize the parameters of this model, addressing the challenge of selecting the optimal hyperparameter combination. The prediction simulation is carried out using the power plant data. Results The improved algorithm demonstrates higher prediction accuracy. It can better predict the change trends of superheated steam temperature during periods of load increase and decrease. The mean absolute error and root mean square error of the model are reduced by 7.02% and 6.22%, respectively, compared to the model before improvement. Conclusions The improved algorithm effectively enhances the accuracy of the superheated steam temperature prediction model, providing a basis for the optimal design of the superheated steam temperature control system.

Key words: carbon peak, carbon neutrality, peak shaving, superheated steam temperature prediction, long short-term memory neural network, novel whale optimization algorithm, principal component analysis

CLC Number:

TK 32

Leihua FENG, Chaopeng NIE, Qifeng GUO, Feng YANG, Jinqi HE, Yucheng MAO. Superheated Steam Temperature Prediction Based on Long Short-Term Memory Neural Network Optimized by Novel Whale Optimization Algorithm[J]. Power Generation Technology, 2025, 46(6): 1212-1222.

Figures/Tables 15

Fig. 1 LSTM neural network unit structure

Tab. 1 Baseline functions

函数表达式	维数	搜索范围	最优值
$f 1 (x) = ∑ i = 1 n x i 2$	30	［-100，100］	0
$f 2 (x) = ∑ i = 1 n x i + ∏ i = 1 n x i$	30	［-10，10］	0
$f 3 (x) = ∑ i = 1 n - x i s i n (x i)$	30	［-500，500］	-1 256 9.5
$f 4 (x) = 1 4 000 ∑ i = 1 n x i 2 - ∏ i = 1 n c o s (x i i) + 1$	30	［-600，600］	0
$f 5 (x) = [1500 + ∑ j = 1 25 1 j + ∑ i = 1 2 (x i - a i j) 6] - 1$	2	［-65.5，65.5］	0.998
$f 6 (x) = [a i - x 1 (b i 2 + b i x 2) b i 2 + b i x 3 + x 4] 2$	4	［-5，5］	0.000 30

Tab. 1 Baseline functions

函数表达式	维数	搜索范围	最优值
$f 1 (x) = ∑ i = 1 n x i 2$	30	［-100，100］	0
$f 2 (x) = ∑ i = 1 n x i + ∏ i = 1 n x i$	30	［-10，10］	0
$f 3 (x) = ∑ i = 1 n - x i s i n (x i)$	30	［-500，500］	-1 256 9.5
$f 4 (x) = 1 4 000 ∑ i = 1 n x i 2 - ∏ i = 1 n c o s (x i i) + 1$	30	［-600，600］	0
$f 5 (x) = [1500 + ∑ j = 1 25 1 j + ∑ i = 1 2 (x i - a i j) 6] - 1$	2	［-65.5，65.5］	0.998
$f 6 (x) = [a i - x 1 (b i 2 + b i x 2) b i 2 + b i x 3 + x 4] 2$	4	［-5，5］	0.000 30

Fig. 2 Test function convergence curves

Tab. 2 Statistical information related to each principal component

主元	特征值	贡献率/%	累加贡献率/%
第1主元	8.308 6	69.238 7	69.238 7
第2主元	1.387 6	11.563 4	80.802 1
第3主元	0.988 0	8.233 4	89.035 5
第4主元	0.504 5	4.204 3	93.239 8
第5主元	0.259 2	2.160 2	95.400 0
第6主元	0.222 2	1.851 6	97.251 6
第7主元	0.123 2	1.027 4	98.279 0
第8主元	0.089 0	0.741 5	99.020 5
第9主元	0.052 6	0.438 1	99.458 6
第10主元	0.035 5	0.296 3	99.754 9
第11主元	0.025 0	0.209 2	99.964 1
第12主元	0.004 3	0.035 9	100

Tab. 3 Weights of each variable in principal components

参数	第1主元	第2主元	第3主元
机组功率	0.322 4	-0.253 8	0.141 2
主汽压力	0.309 0	-0.134 3	0.244 6
中间点过热度	0.130 5	0.677 0	0.290 3
给水量	0.322 3	-0.245 5	0.105 4
给煤量	0.324 5	-0.056 3	-0.223 7
一次风温度	0.262 7	0.361 1	-0.234 5
一次风总量	0.284 4	0.232 6	-0.432 4
一次风压力	0.320 9	0.075 0	-0.311 3
一级减温水流量	0.219 9	0.317 0	0.519 4
二级减温水流量	0.277 4	0.314 1	-0.125 3
烟气含氧量	-0.320 2	0.031 7	-0.166 3
给水温度	0.305 9	-0.074 1	0.350 4

Fig. 3 Prediction process of NWOA-LSTM for superheated steam temperature

Tab. 4 Parameter optimization results

优化参数名称	约束范围	NWOA-LSTM	WOA-LSTM
隐藏层神经元数	［10，200］	112	78
学习率	［0.001，0.02］	0.003 1	0.002
正则化参数	［0.000 1，0.002］	0.000 292	0.001 8

Fig. 4 Total dataset prediction curves

Tab. 5 Prediction performance of WOA-LSTM and NWOA-LSTM models

模型	MAE/℃		RMSE/℃		MAPE/%
模型	训练集	测试集	训练集	测试集	训练集	测试集
WOA-LSTM	0.954 3	2.559 8	2.111 8	3.254 3	0.17	0.44
NWOA-LSTM	0.898 3	2.380 2	1.171 5	3.052 0	0.16	0.41

Fig. 5 Prediction curves of superheated steam temperature under steady-state load

Fig. 6 Prediction curves of superheated steam temperature during load decrease

Fig. 7 Prediction curves of superheated steam temperature during load increase

Tab. 6 Model prediction performance under steady-state load

模型	MAE/℃	RMSE/℃	MAPE/%
WOA-LSTM	1.092 9	1.426 7	0.19
NWOA-LSTM	0.770 4	0.994 9	0.13

Tab. 7 Model prediction performance during load decrease

模型	MAE/℃	RMSE/℃	MAPE/%
WOA-LSTM	1.155 1	1.352 2	0.20
NWOA-LSTM	0.957 1	1.195 2	0.17

Tab. 8 Model prediction performance during load increase

模型	MAE/℃	RMSE/℃	MAPE/%
WOA-LSTM	2.106 5	2.547 4	0.36
NWOA-LSTM	1.737 9	2.261 7	0.30

References 32

[1]	国家能源局．国家能源局发布2024年1-8月份全国电力工业统计数据[J]．电力勘测设计，2024(9)：39．
	National Energy Administration ．National Energy Administration released the statistical data of the national power industry from January to August 2024[J]．Electric Power Survey & Design，2024(9)：39．
[2]	姚玉璧，郑绍忠，杨扬，等．中国太阳能资源评估及其利用效率研究进展与展望[J]．太阳能学报，2022，43(10)：524-535．
	YAO Y B， ZHENG S Z， YANG Y，et al ．Progress and prospects on solar energy resource evaluation and utilization efficiency in China[J]．Acta Energiae Solaris Sinica，2022，43(10)：524-535．
[3]	陈文超，杜雯，王帅杰．风力发电对电力系统的影响研究[J]．中国设备工程，2023(11)：135-137．
	CHEN W C， DU W， WANG S J ．Study on the influence of wind power generation on power system[J]．China Plant Engineering，2023(11)：135-137．
[4]	杨明杰，胡扬宇，千海霞，等．计及风电、光伏、需求侧响应的藤Copula不确定性的综合能源配网分布鲁棒优化[J]．电测与仪表，2025，62(10)：147-155．
	YANG M J， HU Y Y， QIAN H X，et al ．Distributionally robust optimization for integrated energy distribution network considering vine Copula uncertainty of wind power，photovoltaic and demand side response[J]．Electrical Measurement & Instrumentation，2025，62(10)：147-155．
[5]	王长江，尹浩帆，姜涛，等．计及切换补偿的新能源电力系统暂态电压稳定评估[J]．电工技术学报，2025，40(17)：5487-5500．
	WANG C J， YIN H F， JIANG T，et al ．Transient voltage stability assessment approach for renewable energy power system considering switching compensation[J]．Transactions of China Electrotechnical Society，2025，40(17)：5487-5500．
[6]	王放放，杨鹏威，赵光金，等．新型电力系统下火电机组灵活性运行技术发展及挑战[J]．发电技术，2024，45(2)：189-198．
	WANG F F， YANG P W， ZHAO G J，et al ．Development and challenge of flexible operation technology of thermal power units under new power system[J]．Power Generation Technology，2024，45(2)：189-198．
[7]	严新荣，胡志勇，张鹏威，等．煤电机组运行灵活性提升技术研究与应用[J]．发电技术，2024，45(6)：1074-1086．
	YAN X R， HU Z Y， ZHANG P W，et al ．Research and application of operation flexibility improvement technology for coal-fired power unit[J]．Power Generation Technology，2024，45(6)：1074-1086．
[8]	刘云．我国能源电力发展及火电机组灵活性改造综述[J]．洁净煤技术，2023，29(S2)：319-327．
	LIU Y ．Overview of energy and power system development and the flexibility retrofit of thermal power units in China[J]．Clean Coal Technology，2023，29(S2)：319-327．
[9]	董信光，孙健，孔庆雨，等．超临界350 MW机组直流锅炉深度调峰能力试验[J]．热力发电，2018，47(7)：105-112．
	DONG X G， SUN J， KONG Q Y，et al ．Experimental study on depth peak-load regulation capacity of once-through boiler for a supercritical 350 MW unit[J]．Thermal Power Generation，2018，47(7)：105-112．
[10]	张驰，郭媛，黎明．人工神经网络模型发展及应用综述[J]．计算机工程与应用，2021，57(11)：57-69．
	ZHANG C， GUO Y， LI M ．Review of development and application of artificial neural network models[J]．Computer Engineering and Applications，2021，57(11)：57-69．
[11]	杨春来，殷喆，袁晓磊，等．基于神经网络模型的锅炉主蒸汽温度预测控制[J]．锅炉技术，2023，54(5)：30-36．
	YANG C L， YIN Z， YUAN X L，et al ．Predictive control of boiler main steam temperature based on long short-term memory neural network model[J]．Boiler Technology，2023，54(5)：30-36．
[12]	金志远，李胜男，谭鹏，等．基于长短时记忆神经网络的锅炉多参数协同预测模型[J]．热力发电，2021，50(5)：120-126．
	JIN Z Y， LI S N， TAN P，et al ．Multi-parameter collaborative prediction model of boilers based on long-short-term memory neural network[J]．Thermal Power Generation，2021，50(5)：120-126．
[13]	张国斌，郭瑞君，杜荣华，等．基于LSTM预估补偿的火电机组主蒸汽温度控制系统[J]．发电设备，2023，37(1)：51-58．
	ZHANG G B， GUO R J， DU R H，et al ．Main steam temperature control system based on prediction and compensation with LSTM in a thermal power Unit[J]．Power Equipment，2023，37(1)：51-58．
[14]	张文涛，马永光，董子健，等．基于相关向量机的热工参数软测量[J]．热力发电，2019，48(2)：90-95．
	ZHANG W T， MA Y G， DONG Z J，et al ．Soft measurement for thermal parameters based on correlation vector machine[J]．Thermal Power Generation，2019，48(2)：90-95．
[15]	杨婷婷，罗海玉，李浩千，等．基于IWOA优化TCN模型的磨煤机故障预警研究[J/OL]．华北电力大学学报(自然科学版)，2024：1-9．(2024-01-18)．．
	YANG T T， LUO H Y， LI H Q，et al ．Research on fault early warning of coal mill based on improved whale algorithm optimizing time convolution network[J/OL]．Journal of North China Electric Power University (Natural Science Edition)，2024：1-9．(2024-01-18)．．
[16]	张志勇，陆金桂，张猛．基于WOA-BP神经网络的磨煤机出粉量估算[J]．电子测量技术，2022，45(22)：157-161．
	ZHANG Z Y， LU J G， ZHANG M ．Estimation of powder output of coal mill based on WOA-BP neural network[J]．Electronic Measurement Technology，2022，45(22)：157-161．
[17]	张宇晨，姜雪松，李春伟，等．基于Bootstrap误差修正的电力负荷短期预测深度学习模型[J]．热力发电，2023，52(3)：121-129．
	ZHANG Y C， JIANG X S， LI C W，et al ．Deep learning model for short-term power load prediction based on Bootstrap error correction[J]．Thermal Power Generation，2023，52(3)：121-129．
[18]	SHAO L， GUO Q J， LI C，et al ．Short-term load forecasting based on EEMD-WOA-LSTM combination model[J]．Applied Bionics and Biomechanics，2022，2022(1)：2166082． doi:10.1155/2022/2166082
[19]	BRODZICKI A， PIEKARSKI M， JAWOREK-KORJAKOWSKA J ．The whale optimization algorithm approach for deep neural networks[J]．Sensors，2021，21(23)：8003． doi:10.3390/s21238003
[20]	LIU L S， ZHANG R S ．Multistrategy improved whale optimization algorithm and its application[J]．Computational Intelligence and Neuroscience，2022，2022(1)：3418269． doi:10.1155/2022/3418269
[21]	王梓辰，窦震海，董军，等．多策略改进的自适应动态鲸鱼优化算法[J]．计算机工程与设计，2022，43(9)：2638-2645．
	WANG Z C， DOU Z H， DONG J，et al ．Adaptive dynamic whale optimization algorithm based on multi-strategy improvement[J]．Computer Engineering and Design，2022，43(9)：2638-2645．
[22]	李安东，刘升．混合策略改进鲸鱼优化算法[J]．计算机应用研究，2022，39(5)：1415-1421．
	LI A D， LIU S ．Multi-strategy improved whale optimization algorithm[J]．Application Research of Computers，2022，39(5)：1415-1421．
[23]	SUN Y J， CHEN Y ．Multi-population improved whale optimization algorithm for high dimensional optimization[J]．Applied Soft Computing，2021，112：107854． doi:10.1016/j.asoc.2021.107854
[24]	张志豪，李军，王林，等．基于内模控制和改进自抗扰控制的主汽温串级控制系统[J/OL]．中国电机工程学报，1-10[2024-04-09]．．
	ZHANG Z H， LI J， WANG L，et al ．Cascade control system of main steam temperature based on internal model control and improved active disturbance rejection control[J/OL]．Proceedings of the CSEE，1-10[2024-04-09]．．
[25]	张宇帆，艾芊，林琳，等．基于深度长短时记忆网络的区域级超短期负荷预测方法[J]．电网技术，2019，43(6)：1884-1892．
	ZHANG Y F， AI Q， LIN L，et al ．A very short-term load forecasting method based on deep LSTM RNN at zone level[J]．Power System Technology，2019，43(6)：1884-1892．
[26]	MIRJALILI S， LEWIS A ．The whale optimization algorithm[J]．Advances in Engineering Software，2016，95：51-67． doi:10.1016/j.advengsoft.2016.01.008
[27]	YUAN C， ZHAO D， HEIDARI A A，et al ．Artemisinin optimization based on malaria therapy：algorithm and applications to medical image segmentation[J]．Displays，2024，84：102740． doi:10.1016/j.displa.2024.102740
[28]	张永，陈锋．一种改进的鲸鱼优化算法[J]．计算机工程，2018，44(3)：208-213．
	ZHANG Y， CHEN F ．A modified whale optimization algorithm[J]．Computer Engineering，2018，44(3)：208-213．
[29]	武泽权，牟永敏．一种改进的鲸鱼优化算法[J]．计算机应用研究，2020，37(12)：3618-3621．
	WU Z Q， MU Y M ．Improved whale optimization algorithm[J]．Application Research of Computers，2020，37(12)：3618-3621．
[30]	何庆，魏康园，徐钦帅．基于混合策略改进的鲸鱼优化算法[J]．计算机应用研究，2019，36(12)：3647-3651．
	HE Q， WEI K Y， XU Q S ．Mixed strategy based improved whale optimization algorithm[J]．Application Research of Computers，2019，36(12)：3647-3651．
[31]	徐文光，麦鴚，杨开明，等．基于改进WOA参数优化的反应釜PID控制器[J]．现代电子技术，2024，47(7)：150-153．
	XU W G，MAI G， YANG K M，et al ．Reactor PID controller based on improved WOA parameter optimization[J]．Modern Electronics Technique，2024，47(7)：150-153．
[32]	李峰．基于主元分析法的火电厂锅炉故障检测分析[J]．现代制造技术与装备，2022，58(12)：49-51．
	LI F ．Fault detection and analysis of boiler in thermal power plant based on principal component analysis[J]．Modern Manufacturing Technology and Equipment，2022，58(12)：49-51．