Application of soft sensor modeling based on SSA-CNN-LSTM in solar thermal power collection subsystem

Abstract

To address the stochasticity and nonlinearity of solar collector power systems, a soft sensor prediction model with a hybrid convolutional neural network (CNN) and long short-term memory network (LSTM) was constructed, and the hyperparameter optimization of the hybrid neural network (CNN-LSTM) was carried out by using the sparrow search algorithm (SSA). The model utilized the powerful feature extraction and non-linear mapping capabilities of deep learning to effectively handle the complex relationship between input and target variables. The batch normalization technique was used to speed up the training and improve the stability of the soft-sensing model, and the random discard technique was used to prevent the soft-sensing model from overfitting. Finally, the mean absolute error (MAE) was used to assess the accuracy of the soft sensor model predictions. This study compared the proposed model with soft sensor prediction models like Bp, Elman, CNN, LSTM, and CNN-LSTM, using dynamic thermal performance data from the solar collector field of the molten salt linear Fresnel photovoltaic demonstration power plant. The deep learning-based soft sensor model outperformed the other models according to the experimental data. Its coefficients of determination (namely R²) are higher by 6.35%, 8.42%, 5.69%, 6.90%, and 3.67%, respectively. The accuracy and robustness have been significantly improved.

Keywords

soft sensor modeling / linear Fresnel collector subsystem / collector field outlet temperature / deep learning / sparrow search algorithm

References

[1] ZHOU J Z, JI W H, CAO X L, et al. A current perspective on the renewable energy hydrogen production process. Journal of Thermal Science, 2023, 32(2): 542-596.

[2] PARANDEH K, BAGHERI A, JADID S. Optimal day-ahead dynamic pricing of grid-connected residential renewable energy resources under different metering mechanisms. Journal of Modern Power Systems and Clean Energy, 2023, 11(1): 168-178.

[3] YANG X X, CAI B, XUE Y S. Review on optimization of nuclear power development: a cyber-physical-social system in energy perspective. Journal of Modern Power Systems and Clean Energy, 2022, 10(3): 547-561.

[4] HU Z G, JIN Z Y. Opportunities and challenges of solar thermal power generation in China under the background of an energy transformation strategy. Solar Energy, 2019(11): 11-17.

[5] WANG R D, MA J, WANG C L, et al. Progress of linear Fresnel concentrator heat collection system. Infrared and Laser Engineering, 2021, 50(11): 20210452.

[6] GAO X B, LIAO C H, LI G. Study on control strategy and variable load rate of linear Fresnel solar power station. Thermal Power Generation, 2024, 53(5): 47-55.

[7] SU T, ZHAO J B, PEI Y S, et al. Probabilistic physics-informed graph convolutional network for active distribution system voltage prediction. IEEE Transactions on Power Systems, 2023, 38(6): 5969-5972.

[8] LI J J, ZHANG C H, SUN B. Two-stage hybrid deep learning with strong adaptability for detailed day-ahead photovoltaic power forecasting. IEEE Transactions on Sustainable Energy, 2023, 14(1): 193-205.

[9] ALIBERTI A, FUCINI D, BOTTACCIOLI L, et al. Comparative analysis of neural networks techniques to forecast global horizontal irradiance. IEEE Access, 2021, 9: 122829-122846.

[10] ZHAO J Y, CHI Y, ZHOU Y T. Short-term load forecasting based on SSA-LSTM model. Advanced Technology of Electrical Engineering and Energy, 2022, 41(6): 71-79.

[11] JI Y Q, YAN Y B, HE P, et al. CNN-LSTM short-term load forecasting based on the K-Medoids clustering and grid method to extract load curve features. Power System Protection and Control, 2023, 51(18): 81-93.

[12] DE ARAÚJO ELIAS T, COSTA MENDES P R DA, NORMEY-RICO J E. Hybrid predictive controller for overheating prevention of solar collectors. Renewable Energy, 2019, 136: 535-547.

[13] KUMAR R, NADDA R, KUMAR S, et al. Influence of artificial roughness parametric variation on thermal performance of solar thermal collector: an experimental study, response surface analysis and ANN modelling. Sustainable Energy Technologies and Assessments, 2022, 52: 102047.

[14] BIAZETTO P H F, DE ANDRADE G A, NORMEY-RICO J E. Development of an optimal control strategy for temperature regulation and thermal storage operation of a solar power plant based on Fresnel collectors. IEEE Transactions on Control Systems Technology, 2023, 31(3): 1149-1164.

[15] LU X J, LIANG Z P. Application of model predictive control based on Kalman filter in solar collector field of solar thermal power generation. Energy Engineering, 2021, 118(4): 1171-1183.

[16] ZHANG Z Y, LU X J, KONG L G, et al. Predicting molten salt temperature at the circuit outlet of linear Fresnel heat collector using K-means combined with RBF neural network. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(3): 213-222.

[17] SHAH A, AHMED K, HAN X. A novel prediction error-based power forecasting scheme for real PV systems using the PVUSA model: a grey box-based neural network approach. IEEE Access, 2021, 9: 87196-87206.

[18] LU X J, SUN P, FAN D J. Application of sliding mode predictive control in a solar thermal power generation collector subsystem. Journal of Applied Science and Engineering, 2023, 26(10): 1481-1489.

[19] LIU W, XUE F F, GAO Y S, et al. Wind-speed forecasting model based on DBN-Elman combined with improved PSO-HHT. Global Energy Interconnection, 2023, 6(5): 530-541.

[20] PATARO I M L, GIL J D, AMERICANO DA COSTA M V, et al. A stochastic nonlinear predictive controller for solar collector fields under solar irradiance forecast uncertainties. IEEE Transactions on Control Systems Technology, 2024, 32(1): 99-111.

[21] XUE J K, SHEN B. A novel swarm intelligence optimization approach: sparrow search algorithm. Systems Science & Control Engineering, 2020, 8(1): 22-34.