Novel prediction method of process and system performance for organic Rankine cycle based on neural network

doi:10.11949/0438-1157.20181521

Abstract

Abstract:

Organic Rankine cycle (ORC) is one of the most promising technologies for medium and low temperature thermal energy-electric energy conversion, and has received more and more attention in recent years. Working fluid is the carrier for energy transport or conversion in the ORC. Because of the diversity of the heat source and working substances, the screening of working fluids and the optimization of the system are very important to improve the comprehensive performance of the ORC. Accurate prediction of working fluid properties is significant for the accurate prediction and optimization of the ORC performance. Based on the artificial neural network and group contribution method (ANN-GCM), a prediction method for the ORC performance is presented. A group table covering 11 groups established, 7958 of data are derived from REFPROP for ANN training, obtaining the correlation of the energy transform and entropy difference in DRC. The performance of the ORC is tested by using 21 common refrigerants in 1584 working conditions. The error of predicting the thermal efficiency, output power, and exergetic efficiency of the ORC system with the experimental data is 1.01%, 1.02% and 1.61%. Comparing with the traditional method, the prediction accuracy is significantly improved.

Key words: neural networks, thermophysical properties, prediction, group contribution method, organic Rankine cycle

摘要：

有机朗肯循环（ORC）是中低温热能-电能转换中最具前景的技术之一，近年来受到越来越多的关注。工质是ORC实现的载体，由于热源及可选工质的多样性，工质筛选及系统的优化对于提升ORC综合性能非常重要，而物性及过程特性的准确预测是关键。提出了基于神经网络-基团贡献法的ORC系统性能计算方法，建立了涵盖11个基团的基团表，从REFPROP中调用51种工质7958组数据进行神经网络训练，获得了ORC中各个热力过程能量转换和熵差的计算关联式。计算了21种常用工质在1584组工况下的ORC系统性能，并与基于传统方法计算的ORC系统性能参数进行了对比。结果显示预测得到的ORC系统热效率、净输出功和系统?效率与用REFPROP计算得出的结果相比误差分别为1.01%、1.02%和1.61%，相比传统方法，预测精度有显著提高。

关键词: 神经网络, 热力学性质, 预测, 基团贡献法, 有机朗肯循环

CLC Number:

TK 11⁺2

Yupeng WANG, Junwei LIANG, Xianglong LUO, Yifan LI, Jianyong CHEN, Ying CHEN. Novel prediction method of process and system performance for organic Rankine cycle based on neural network[J]. CIESC Journal, 2019, 70(9): 3256-3266.

王羽鹏, 梁俊伟, 罗向龙, 李逸帆, 陈健勇, 陈颖. 基于神经网络的有机朗肯循环过程及循环性能计算方法[J]. 化工学报, 2019, 70(9): 3256-3266.

Figures/Tables 13

Fig.1 Flow for evaluating performance of working fluids in ORC

Fig.2 Schematic diagram of typical ORC (a) and T-s diagram of ORC(b)

Table 1 Functional groups

基团分类
—CH₃	—CH₂—	>CH—	>C<	═CH₂	═CH—
—F	—Cl	—O—	—I	═C<

Fig.3 Flow of training data acquisition and ANN operation

Table 2 51 training working fluids for neuron network

工质
R1234yf	R134a	R11	R113	R114	R115	R123	R125	R143a
R152a	R218	R227ea	R236ea	R236fa	R245ca	R245fa	propane	R32
butane	isobutane	pentane	ipentane	R1233zd	R1234ze	R141b	R142b	Re143a
R161	R21	dimethylether	R124	R1216	Re245cb2	Re245fa2	propylene	C₄F₁₀
R12	Re347mcc	CF₃I	isobutene	C₅F₁₂	neopentne	hexane	isohxane	heptane
acetone	R365mfc	cis-butene	trans-butene		dimethyl carbonate		diethyl ether

Table 3 Total number of groups using for training neuron network

基团	—CH₃	—CH₂—	>CH—	>C<	═C<	═CH—	═CH₂	—F	—Cl	—I	—O—
出现次数	47	28	14	50	6	9	3	147	20	1	10

Table 4 Parameters of using ANN-GCM for predicting heat transfer in ORC process

ANN-GCM相关参数		W _p/kW	W _t/kW	Q _evap/kJ	s _p/(J/(kg·K))	s _evap/(J/(kg·K))	s _t/(J/(kg·K))
隐层神经元个数		24			20
隐层传递函数		tansig			logsig
训练算法		L-M			B-R
R ²	训练集	0.9987	0.9999	0.9999	0.9981	0.9998	0.9999
	验证集	0.9979	0.9994	0.9999	—	—	—
	测试集	0.9981	0.9999	0.9999	0.9950	0.9641	0.9999
	总集	0.9983	0.9999	0.9999	0.9976	0.9941	0.9999
AAD/%	训练集	5.8428	1.6693	0.4365	4.8560	0.3800	0.8130
	验证集	1.3102	0.6451	0.3238	—	—	—
	测试集	2.0479	0.8767	0.2770	5.9408	1.8656	0.9509
	总集	4.5937	1.3968	0.3957	5.0187	0.6029	0.8336
Bias/%	训练集	-2.1722	-0.0856	-0.0081	-0.0102	-0.0003	-0.0003
	验证集	0.5396	0.2184	0.0690	—	—	—
	测试集	2.0479	0.8767	0.2770	-0.0106	0.0136	-0.0007
	总集	-1.5178	-0.0448	-0.0006	-0.0103	0.0018	-0.0004
RMS	训练集	0.0023	0.4738	2.6511	8.0677	66.8208	8.0677
	验证集	0.0004	0.1299	2.1518	—	—	—
	测试集	0.0019	0.4949	1.8710	38.1364	114.3525	84.2851
	总集	0.0445	0.6522	1.5682	25.9051	44.9142	69.7206

Fig.4 Distribution of ARD of ORC process parameters

Table 5 21 working fluids used to validate model

工质	CAS	分子式	相对分子质量	工质	CAS	分子式	相对分子质量
R123	306-83-2	CF₃CHCl₂	152.93	R290	74-98-6	CH₃CH₂CH₃	44.096
R125	354-33-6	CHF₂CF₃	120.02	R600	106-97-8	CH₃CH₂CH₂CH₃	58.122
R134a	811-97-2	CF₃CH₂F	102.03	R600a	75-28-5	CH(CH₃)₂CH₃	58.122
R143a	420-46-2	CH₃CF₃	84.041	R601	709-66-0	CH₃CH₂CH₂CH₂CH₃	72.149
R152a	75-37-6	CH₃CHF₂	66.051	R601a	78-78-4	(CH₃)₂CHCH₂CH₃	72.149
R218	76-19-7	CF₃CF₂CF₃	188.02	n-C₆H₁₄	110-54-3	CH₃(CH₂)₄CH₃	86.175
R227ea	431-89-0	CF₃CHFCF₃	170.03	C₅F₁₂	678-26-2	CF₃CF₂CF₂CF₂CF₃	288.03
R236ea	431-63-0	CF₃CHFCHF₂	152.04	R1233zd	102687-65-0	CHCl═CHCF₃	130.5
R236fa	690-39-1	CF₃CH₂CF₃	152.04	R1234yf	754-12-1	CH₂═CFCF₃	114.04
R245ca	679-86-7	CH₂FCF₂CHF₂	134.05	R1234ze	29118-24-9	CF₃CH═CHF	114.04
R245fa	460-73-1	CF₃CH₂CHF₂	134.05

Fig.5 Bias using different methods

Table 6 AADs of process parameters using different methods

方法	AAD/%
方法	Q _evap	Q _con	W _P	W _t	W _net	η _th
本文方法	0.26	0.31	1.72	0.97	1.01	1.02
Su等^[24]	5.05	5.08	10.7	7.25	8.28	4.89
Joback等^[17]	10.95	10.98	52.69	10.26	13	5.42

Fig.6 Distribution of ARD of exergy loss of each process and exergetic efficiency of ORC

Table 7 Errors of predicting isomers

工质	EATII	T _evap/K	T _con/K	T _s/K	Error/%
工质	EATII	T _evap/K	T _con/K	T _s/K	W _p	W _t	Q _evap	Q _con	W _net	$η t h$
R236ea	35.248	360	298.15	5	1.18	0.27	0.04	0.01	0.25	0.21
R236fa	36.5556	360	298.15	5	0.59	0.56	0.12	0.04	0.60	0.48
R245ca	29.715	410	298.15	5	1.61	0.87	0.31	0.18	0.85	0.54
R245fa	28.2336	410	298.15	5	0.08	0.72	0.26	0.14	0.75	0.49

Table 7 Errors of predicting isomers

工质	EATII	T _evap/K	T _con/K	T _s/K	Error/%
工质	EATII	T _evap/K	T _con/K	T _s/K	W _p	W _t	Q _evap	Q _con	W _net	$η t h$
R236ea	35.248	360	298.15	5	1.18	0.27	0.04	0.01	0.25	0.21
R236fa	36.5556	360	298.15	5	0.59	0.56	0.12	0.04	0.60	0.48
R245ca	29.715	410	298.15	5	1.61	0.87	0.31	0.18	0.85	0.54
R245fa	28.2336	410	298.15	5	0.08	0.72	0.26	0.14	0.75	0.49

References 31

1	宋建忠, 张小松, 李舒宏, 等 . 太阳能有机朗肯循环系统的实验特性[J]. 化工学报, 2014, 65(12): 4958-4964.
	Song J Z , Zhang X S , Li S H , et al . Experimental characteristics of solar organic Rankine cycle system[J]. CIESC Journal, 2014, 65(12): 4958-4964.
2	刘继芬, 王景甫, 马重芳, 等 . 中低温地热发电循环参数的优化[J]. 化工学报, 2011, 62(S1): 190-196.
	Liu J F , Wang J F , Ma C F , et al . Optimization of cycle parameters in low-medium temperature geothermal power generation[J]. CIESC Journal, 2011, 62(S1): 190-196.
3	杨凯, 张红光, 宋松松, 等 . 变工况下车用柴油机排气余热有机朗肯循环回收系统[J]. 化工学报, 2015, 66(3): 1097-1103.
	Yang K , Zhang H G , Song S S , et al . Waste heat organic Rankine cycle of vehicle diesel engine under variable working conditions[J]. CIESC Journal, 2015, 66(3): 1097-1103.
4	黄仁龙, 罗向龙, 梁志辉, 等 . 基于分液冷凝的R245fa/pentane混合工质朗肯循环多目标优化[J]. 化工学报, 2018, 69(5): 2040-2048.
	Huang R L , Luo X L , Liang Z H , et al . Multi-objective optimization of Rankine cycle using R245fa/pentane based on liquid-vapor separation[J]. CIESC Journal, 2018, 69(5): 2040-2048.
5	许俊俊, 罗向龙, 王永真, 等 . ORC工质选择的多级非结构性模糊决策分析[J]. 化工学报, 2015, 66(3): 1051-1058.
	Xu J J , Luo X L , Wang Y Z , et al . Optimum selection of ORC working fluid using multi-level fuzzy optimization and non-structural fuzzy decision[J]. CIESC Journal, 2015, 66(3): 1051-1058.
6	Imran M , Usman M , Park B , et al . Multi-objective optimization of evaporator of organic Rankine cycle (ORC) for low temperature geothermal heat source[J]. Applied Thermal Engineering, 2015, 80: 1-9.
7	Györke G , Deiters U K , Groniewsky A , et al . Novel classification of pure working fluids for organic Rankine cycle[J]. Energy, 2018, 145: 288-300.
8	Bao J , Zhao L . A review of working fluid and expander selections for organic Rankine cycle[J]. Renewable and Sustainable Energy Reviews, 2013, 24: 325-342.
9	He C , Liu C , Zhou M , et al . A new selection principle of working fluids for subcritical organic Rankine cycle coupling with different heat sources[J]. Energy, 2014, 68: 283-291.
10	Bricks J L , Kachkovskii A D , Slominskii Y L , et al . Molecular design of near infrared polymethine dyes: a review[J]. Dyes and Pigments, 2015, 121: 238-255.
11	Gholivand M B , Khodadadian M , Ahmadi F . Computer aided-molecular design and synthesis of a high selective molecularly imprinted polymer for solid-phase extraction of furosemide from human plasma[J]. Analytica Chimica Acta, 2010, 658(2): 225-232.
12	Struebing H , Ganase Z , Karamertzanis P G , et al . Computer-aided molecular design of solvents for accelerated reaction kinetics[J]. Nature Chemistry, 2013, 5(11): 952-957.
13	Papadopoulos A I , Stijepovic M , Linke P . On the systematic design and selection of optimal working fluids for organic Rankine cycles[J]. Applied Thermal Engineering, 2010, 30(6/7): 760-769.
14	Palma-Flores O , Flores-Tlacuahuac A , Canseco-Melchor G . Optimal molecular design of working fluids for sustainable low-temperature energy recovery[J]. Computers & Chemical Engineering, 2015, 72: 334-349.
15	Palma-Flores O , Flores-Tlacuahuac A , Canseco-Melchorb G . Simultaneous molecular and process design for waste heat recovery[J]. Energy, 2016, 99: 32-47.
16	White M T , Oyewunmi O A , Haslam A J , et al . Industrial waste-heat recovery through integrated computer-aided working-fluid and ORC system optimization using SAFT-γMie[J]. Energy Conversion and Management, 2017, 150: 851-869.
17	Joback K , Reid R . Estimation of pure-component properties from group-contributions[J]. Chemical Engineering Communications, 1987, 57(1/2/3/4/5/6): 233-243.
18	Constantinou L , Gani R . New group contribution method for estimating properties of pure compounds[J]. AIChE Journal, 1994, 40: 1697-1710.
19	Marrero-Morejón J , Pardillo-Fontdevila E . Estimation of pure compound properties using group-interaction contributions[J]. AIChE Journal, 1999, 45(3): 615-621.
20	Lazzús J A . ρ-T-P prediction for ionic liquids using neural networks[J]. Journal of the Taiwan Institute of Chemical Engineers, 2009, 40(2): 213-232.
21	苏文, 赵力, 邓帅 . 基于基团拓扑的遗传神经网络工质临界温度预测[J]. 化工学报, 2016, 67(11): 4689-4695.
	Su W , Zhao L , Deng S . Prediction of refrigerant critical temperature with genetic neural network based on group topology[J]. CIESC Journal, 2016, 67(11): 4689-4695.
22	Mondejar M E , Cignitti S , Abildskov J , et al . Prediction of properties of new halogenated olefins using two group contribution approaches[J]. Fluid Phase Equilibria, 2017, 433: 79-96.
23	Barbieri E S , Morini M , Pinelli M . Development of a model for the simulation of organic Rankine cycles based on group contribution techniques[C]// ASME 2011 Turbo Expo: Turbine Technical Conference and Exposition. American Society of Mechanical Engineers, 2011.
24	Su W , Zhao L , Deng S . Developing a performance evaluation model of organic Rankine cycle for working fluids based on the group contribution method[J]. Energy Conversion and Management, 2017, 132: 307-315.
25	Lemmon E W , Huber M L , Mclinden M O . NIST standard reference database 23: reference fluid thermodynamic and transport properties-REFPROP. 9.1[DB]. 2013.
26	Deng S , Su W , Zhao L . A neural network for predicting normal boiling point of pure refrigerants using molecular groups and a topological index[J]. International Journal of Refrigeration, 2016, 63: 63-71.
27	Taskinen J , Yliruusi J . Prediction of physicochemical properties based on neural network modelling[J]. Advanced Drug Delivery Reviews, 2003, 55(9): 1163-1183.
28	Moosavi M , Soltani N . Prediction of hydrocarbon densities using an artificial neural network-group contribution method up to high temperatures and pressures[J]. Thermochim Acta, 2013, 556: 89-96.
29	Moosavi M , Sedghamiz E , Abareshi M . Liquid density prediction of five different classes of refrigerant systems (HCFCs, HFCs, HFEs, PFAs and PFAAs) using the artificial neural network-group contribution method[J]. International Journal of Refrigeration, 2014, 48: 188-200.
30	Su W , Zhao L , Deng S . Group contribution methods in thermodynamic cycles: physical properties estimation of pure working fluids[J]. Renewable and Sustainable Energy Reviews, 2017, 79: 984-1001.
31	Peng D Y , Robinson D B . A new two-constant equation of state[J]. Minerva Ginecologica, 1976, 12(11/12): 3069–3078.

[1]	Manzheng ZHANG, Meng XIAO, Peiwei YAN, Zheng MIAO, Jinliang XU, Xianbing JI. Working fluid screening and thermodynamic optimization of hazardous waste incineration coupled organic Rankine cycle system [J]. CIESC Journal, 2023, 74(8): 3502-3512.
[2]	Linqi YAN, Zhenlei WANG. Multi-step predictive soft sensor modeling based on STA-BiLSTM-LightGBM combined model [J]. CIESC Journal, 2023, 74(8): 3407-3418.
[3]	Gang YIN, Yihui LI, Fei HE, Wenqi CAO, Min WANG, Feiya YAN, Yu XIANG, Jian LU, Bin LUO, Runting LU. Early warning method of aluminum reduction cell leakage accident based on KPCA and SVM [J]. CIESC Journal, 2023, 74(8): 3419-3428.
[4]	Yuying GUO, Jiaqiang JING, Wanni HUANG, Ping ZHANG, Jie SUN, Yu ZHU, Junxuan FENG, Hongjiang LU. Water-lubricated drag reduction and pressure drop model modification for heavy oil pipeline [J]. CIESC Journal, 2023, 74(7): 2898-2907.
[5]	Yanhui LI, Shaoming DING, Zhouyang BAI, Yinan ZHANG, Zhihong YU, Limei XING, Pengfei GAO, Yongzhen WANG. Corrosion micro-nano scale kinetics model development and application in non-conventional supercritical boilers [J]. CIESC Journal, 2023, 74(6): 2436-2446.
[6]	Yuan YU, Weiwei CHEN, Junjie FU, Jiaxiang LIU, Zhiwei JIAO. Study and prediction of flow field in the annular region of geometrically similar turbo air classifier [J]. CIESC Journal, 2023, 74(6): 2363-2373.
[7]	Xuejin GAO, Yuzhuo YAO, Huayun HAN, Yongsheng QI. Fault monitoring of fermentation process based on attention dynamic convolutional autoencoder [J]. CIESC Journal, 2023, 74(6): 2503-2521.
[8]	Yurong DANG, Chunlan MO, Kerui SHI, Yingcong FANG, Ziyang ZHANG, Zuoshun LI. Comprehensive evaluation model combined with genetic algorithm for the study on the performance of ORC system with zeotropic mixture [J]. CIESC Journal, 2023, 74(5): 1884-1895.
[9]	Cheng YUN, Qianlin WANG, Feng CHEN, Xin ZHANG, Zhan DOU, Tingjun YAN. Deep-mining risk evolution path of chemical processes based on community structure [J]. CIESC Journal, 2023, 74(4): 1639-1650.
[10]	Xinyuan WU, Qilei LIU, Boyuan CAO, Lei ZHANG, Jian DU. Group2vec: group vector representation and its property prediction applications based on unsupervised machine learning [J]. CIESC Journal, 2023, 74(3): 1187-1194.
[11]	Jiahui CHEN, Xinze YANG, Guzhong CHEN, Zhen SONG, Zhiwen QI. A critical discussion on developing molecular property prediction models: density of ionic liquids as example [J]. CIESC Journal, 2023, 74(2): 630-641.
[12]	Kenian SHI, Jingyuan ZHENG, Yu QIAN, Siyu YANG. Two-stage stochastic programming of steam power system based on Markov chain [J]. CIESC Journal, 2023, 74(2): 807-817.
[13]	Xuejin GAO, Kun CHENG, Huayun HAN, Huihui Gao, Yongsheng QI. Fault diagnosis of chillers using central loss conditional generative adversarial network [J]. CIESC Journal, 2022, 73(9): 3950-3962.
[14]	Jing YANG, Zhenkang LIN, Jun TANG, Cheng FAN, Kening SUN. A review of fault characteristics, fault diagnosis and identification for lithium-ion battery systems [J]. CIESC Journal, 2022, 73(8): 3394-3405.
[15]	Xinjie ZHOU, Jianlin WANG, Xingcong AI, Enguang SUI, Rutong WANG. IDPC-RVM based online prediction of quality variables for multimode batch processes [J]. CIESC Journal, 2022, 73(7): 3120-3130.