A unified multi-objective physics-informed neural network framework for thermo-mechanical parameter inversion of arch dams with adaptive loss weighting and pareto-based optimization

Abstract

Parameter inversion of arch dams is crucial for safety assessment, but it is usually carried out via finite element (FE) model updating or black-box optimization, which can be computationally expensive, difficult to balance physics- and data-driven objectives, and unstable for multi-parameter thermo-mechanical inversion while underusing short-term quasi-static responses. This study proposes a unified multi-objective inversion framework based on physics-informed neural networks for thermo-mechanical parameter identification in arch dams. Differentiable thermal and mechanical governing equations, boundary conditions and observation misfits are embedded into an end-to-end architecture to enable synchronous inversion of parameters under explicit physical constraints. To better exploit both long-term and short-term components of the monitored responses, a sinusoidal-activation network with prior-informed multi-channel inputs (spatial location, zoning information and dam-related thermal features) is adopted, enhancing physical consistency and high-frequency response representation. An Actor–Critic deep reinforcement learning scheme provides adaptive loss weighting, while the non-dominated sorting genetic algorithm II refines Pareto-optimal solutions. A case study on a concrete arch dam shows that the framework significantly reduces overall and high-frequency errors compared with a traditional separation inversion method. FE verification yields a maximum temperature error of 0.167 °C and a maximum radial displacement error of 0.143 mm, with relative errors below 4.89%.

Graphical abstract

Keywords

arch dams / thermo-mechanical parameters inversion / physics-informed neural networks / deep reinforcement learning / multi-objective optimization

References

[1] Liang C , Chen J Y , Xu Q , Li J . Damage identification of arch dams based on post-earthquake residual displacement gradients. Soil Dynamics and Earthquake Engineering, 2023, 172: 108054

[2] Wang X W , Liu Y R , Tao Z F , Wang W Q , Yang Q . Study on the failure process and nonlinear safety of high arch dam and foundation based on geomechanical model test. Engineering Failure Analysis, 2020, 116: 104704

[3] Gu H , Wu Z R , Huang X F , Song J T , Sun L Z . Dam-zoning modulus reconstruction using adjoint method. European Journal of Environmental and Civil Engineering, 2018, 22(9): 1089–1110

[4] Yang J Q , Wang J T , Pan J W . An intelligent method for temperature load of arch dams. Engineering Structures, 2024, 321: 118918

[5] Zhao E F , Wu C Q , Wang S W , Hu J , Wang W Q . Seepage dissolution effect prediction on aging deformation of concrete dams by coupled chemo-mechanical model. Construction and Building Materials, 2020, 237: 117603

[6] Kang F , Wu Y R , Ma J T , Li J J . Structural identification of super high arch dams using Gaussian process regression with improved salp swarm algorithm. Engineering Structures, 2023, 286: 116150

[7] Huang Y Y , Xie T , Xu Y , Wang R L . Evaluation of the mechanical parameters of a reinforced concrete dam based on multi-source data. Structural Concrete, 2022, 23(2): 652–668

[8] Shao C F , Zheng S , Gu C S , Tian S G , Gu H , Xu Y X , Wang Y . Numerical analysis of temperature deformation characteristics for super-high arch dams considering solar radiation effects. Structural Control and Health Monitoring, 2023, 2023: 2123241

[9] Bukenya P , Moyo P , Beushausen H , Oosthuizen C . Health monitoring of concrete dams: A literature review. Journal of Civil Structural Health Monitoring, 2014, 4(4): 235–244

[10] Wang F , Song R , Yu H , Zhang A , Wang L W , Chen X . Thermal parameter inversion of low-heat cement concrete for Baihetan arch dam. Engineering Applications of Artificial Intelligence, 2024, 131: 107823

[11] Wang S W , Xu C , Liu Y , Xu B . Zonal intelligent inversion of viscoelastic parameters of high arch dams using an HEST statistical model. Journal of Civil Structural Health Monitoring, 2022, 12(1): 207–223

[12] Wang S W , Xu C , Gu C S , Su H Z , Wu B B . Hydraulic-seasonal-time-based state space model for displacement monitoring of high concrete dams. Transactions of the Institute of Measurement and Control, 2021, 43(15): 3347–3359

[13] Xiong F J , Wei B W , Xu F G . Identification of arch dam mechanical parameters based on sensitivity analysis and Hooke–Jeeves algorithm optimization. Structures, 2022, 46: 88–98

[14] Bo C , Qingyi W , Weinan C , Hao G . Optimized inversion method for thermal parameters of concrete dam under the insulated condition. Engineering Applications of Artificial Intelligence, 2023, 126: 106898

[15] Zhao C J , Li Z P , Wang F , Zhou Y H , Zhou H W , Liang Z P , Gong P . Study on thermal parameter inversion considering various material zones in concrete dam blocks. Materials Today Communications, 2025, 42: 111501

[16] Zhao C , Wang F , Jiang X X , Song R , Zhang A , Liu X L . Thermal parameter identification of concrete dams based on hybrid particle swarm optimization using distributed optical fiber monitoring data. Swarm and Evolutionary Computation, 2024, 87: 101582

[17] Salazar F , Vicente D J , Irazábal J , De-Pouplana I , San Mauro J . A review on thermo-mechanical modelling of arch dams during construction and operation: Effect of the reference temperature on the stress field. Archives of Computational Methods in Engineering, 2020, 27(5): 1681–1707

[18] Jeong H , Bai J S , Batuwatta-Gamage C P , Rathnayaka C , Zhou Y , Gu Y T . A physics-informed neural network-based topology optimization (PINNTO) framework for structural optimization. Engineering Structures, 2023, 278: 115484

[19] Li Y X , Ni P , Sun L M , Xia Y . Finite element model-informed deep learning for equivalent force estimation and full-field response calculation. Mechanical Systems and Signal Processing, 2024, 206: 110892

[20] Zhang K , Gu C S , Zhu Y T , Li Y T , Shu X S . A mathematical-mechanical hybrid driven approach for determining the deformation monitoring indexes of concrete dam. Engineering Structures, 2023, 277: 115353

[21] Karniadakis G E , Kevrekidis I G , Lu L , Perdikaris P , Wang S F , Yang L . Physics-informed machine learning. Nature Reviews Physics, 2021, 3(6): 422–440

[22] Raissi M , Perdikaris P , Karniadakis G E . Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 2019, 378: 686–707

[23] Zheng J L , Li F , Huang H M . T-phPINN: Physics-informed neural networks for solving 2D non-Fourier heat conduction equations. International Journal of Heat and Mass Transfer, 2024, 235: 126216

[24] Lorenzen F, Zargaran A, Janoske U. Utilizing physics-informed neural networks for modeling 3D fluid flows incorporating parametric boundary conditions. In: Proceedings of the 14th International Conference on Computational Heat and Mass Transfer (ICCHMT 2023). Düsseldorf: Springer, 2023, 180–190

[25] Cuomo S , Di Cola VS , Giampaolo F , Rozza G , Raissi M , Piccialli F . Scientific machine learning through physics-informed neural networks: Where we are and what’s next. Journal of Scientific Computing, 2022, 92(3): 88

[26] Zhang Z L , Pan Q J , Yang Z H , Yang X L . Physics-informed deep learning method for predicting tunnelling-induced ground deformations. Acta Geotechnica, 2023, 18(9): 4957–4972

[27] Li X Y. Enhancing Infrastructure and Dynamic Systems Modeling Through the Synergy of Physics-Based Models and Machine Learning. Dissertation for the Doctoral Degree. East Lansing: Michigan State University, 2024

[28] Teloli R D O , Tittarelli R , Bigot M , Coelho L , Ramasso E , Moal P L , Ouisse M . A physics-informed neural networks framework for model parameter identification of beam-like structures. Mechanical Systems and Signal Processing, 2025, 224: 112189

[29] Mai J , Li Y , Long L , Huang Y , Zhang H L , You Y C . Two-dimensional temperature field inversion of turbine blade based on physics-informed neural networks. Physics of Fluids, 2024, 36(3): 037114

[30] Gao Z R , Fu Z J , Wen M J , Guo Y , Zhang Y M . Physical informed neural network for thermo-hydral analysis of fire-loaded concrete. Engineering Analysis with Boundary Elements, 2024, 158: 252–261

[31] Kurata K, Sato G, Tsunokuni I, Ikeda Y. Noise-robust estimation of early-part room impulse responses based on physics-informed neural network with dynamic pulling method. In: 2024 Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC). Macau: IEEE, 2024, 1–5

[32] Hng Lim W , Sfarra S , Hsiao T Y , Yao Y . Physics-informed neural networks for defect detection and thermal diffusivity evaluation in carbon fiber-reinforced polymer using pulsed thermography. IEEE Transactions on Instrumentation and Measurement, 2025, 74: 4500910

[33] Amini D , Haghighat E , Juanes R . Inverse modeling of nonisothermal multiphase poromechanics using physics-informed neural networks. Journal of Computational Physics, 2023, 490: 112323

[34] Tian Z Y , Zhou S H , Lee A , Zhao Y , Gong Q M . A Bayesian-based approach for inversion of earth pressures on in-service underground structures. Acta Geotechnica, 2024, 19(4): 1911–1928

[35] Tian Z Y , Zhou S H , Lee A , Shan Y , Detmann B . How to identify earth pressures on in-service tunnel linings: Insights from Bayesian inversion to address non-uniqueness. Transportation Geotechnics, 2024, 48: 101344

[36] Tian Z Y , Lee A , Zhou S H . Adaptive tempered reversible jump algorithm for Bayesian curve fitting. Inverse Problems, 2024, 40(4): 045024

[37] Yang D G , Karimi H R , Pawelczyk M . A new intelligent fault diagnosis framework for rotating machinery based on deep transfer reinforcement learning. Control Engineering Practice, 2023, 134: 105475

[38] Luo J G , Chen C , Xie J C . Multi-objective immune algorithm with preference-based selection for reservoir flood control operation. Water Resources Management, 2015, 29(5): 1447–1466

[39] Li W Y , Wu Z Y . A methodology for dam parameter identification combining machine learning, multi-objective optimization and multiple decision criteria. Applied Soft Computing, 2022, 128: 109476

[40] Castiglione F , Corrente S , Greco S , Bianucci P , Sordo-Ward A , Garrote L , Foti E , Musumeci R E . Interactive multiobjective evolutionary optimization model for dam management support. Journal of Hydrology, 2025, 647: 132304

[41] Li J R , Chen C , Wu Z Y , Chen J K . Multi-source data-driven unsaturated seepage parameter inversion: Application to a high core rockfill dam. Journal of Hydrology, 2023, 617: 129171

[42] Li Z F , Wu Z Y , Chen J K , Li Y L , Huang H B , Lu Y , Lu X , Li J R . Inverse analysis of deformation moduli for high arch dams using the displacement reconstruction technique and multi-objective optimization. Computer-Aided Civil and Infrastructure Engineering, 2024, 39(10): 1507–1529

[43] Zhang D M , Zhou Y L , Huang H W , Zhang J Z . A physics and information dual-driven intelligent diagnosis method for longitudinal mechanical behavior of long-distance shield tunnels. Rock and Soil Mechanics, 2023, 44(10): 2997–3010

[44] Lu Y , Mei G . A deep learning approach for predicting two-dimensional soil consolidation using physics-informed neural networks (PINN). Mathematics, 2022, 10(16): 2949

[45] Sukumar N , Srivastava A . Exact imposition of boundary conditions with distance functions in physics-informed deep neural networks. Computer Methods in Applied Mechanics and Engineering, 2022, 389: 114333

[46] Lu L , Meng X H , Mao Z P , Karniadakis G E . DeepXDE: A deep learning library for solving differential equations. SIAM Review, 2021, 63(1): 208–228

[47] Son H , Cho S W , Hwang H J . Enhanced physics-informed neural networks with augmented Lagrangian relaxation method (AL-PINNs). Neurocomputing, 2023, 548: 126424

[48] Juntunen M , Stenberg R . Nitsche’s method for general boundary conditions. Mathematics of Computation, 2009, 78(267): 1353–1374

[49] Samaniego E , Anitescu C , Goswami S , Nguyen-Thanh V M , Guo H , Hamdia K , Zhuang X , Rabczuk T . An energy approach to the solution of partial differential equations in computational mechanics via machine learning: Concepts, implementation and applications. Computer Methods in Applied Mechanics and Engineering, 2020, 362: 112790

[50] Eshaghi M S , Anitescu C , Thombre M , Wang Y Z , Zhuang X Y , Rabczuk T . Variational physics-informed neural operator (VINO) for solving partial differential equations. Computer Methods in Applied Mechanics and Engineering, 2025, 437: 117785

[51] Abadi M, Barham P, Chen J M, Chen Z F, Davis A, Dean J, Devin M, Ghemawat S, Irving G, Isard M, et al. TensorFlow: A system for Large-Scale machine learning. In: 12th USENIX Symposium on Operating Systems Design and Implementation. Savannah: USENIX Association, 2016, 265–283

[52] Fang Q , Mou X K , Li S B . A physics-informed neural network based on mixed data sampling for solving modified diffusion equations. Scientific Reports, 2023, 13(1): 2491

[53] Haghighat E , Raissi M , Moure A , Gomez H , Juanes R . A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics. Computer Methods in Applied Mechanics and Engineering, 2021, 379: 113741

[54] Alkhadhr S, Almekkawy M. Multifrequency encoding in PINNs for precision wave equation modeling in inhomogeneous media. In: 2024 IEEE Ultrasonics, Ferroelectrics, and Frequency Control Joint Symposium (UFFC-JS). Taipei, China: IEEE, 2024, 1–4

[55] Sharafati A , Yaseen Z M , Shahid S . A novel simulation–optimization strategy for stochastic-based designing of flood control dam: A case study of Jamishan dam. Journal of Flood Risk Management, 2021, 14(1): e12678

[56] Mooselu M G , Nikoo M R , Bakhtiari P H , Rayani N B , Izady A . Conflict resolution in the multi-stakeholder stepped spillway design under uncertainty by machine learning techniques. Applied Soft Computing, 2021, 110: 107721

[57] Liu R , Ma G , Kong F H , Ai Z T , Xiong K , Zhou W , Wang X M , Chang X L . Pareto-guided active learning for accelerating surrogate-assisted multi-objective optimization of arch dam shape. Engineering Structures, 2025, 326: 119541

[58] Xu Y P , Wu Z Y . Parameter identification of unsaturated seepage model of core rockfill dams using principal component analysis and multi-objective optimization. Structures, 2022, 45: 145–162

[59] Tian Y T , Cao X H , Huang K , Fei C , Zheng Z , Ji X W . Learning to drive like human beings: A method based on deep reinforcement learning. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(7): 6357–6367

[60] Li Y H , Wang Y P , Zhao X , Chen Z . A deep reinforcement learning-based intelligent fault diagnosis framework for rolling bearings under imbalanced datasets. Control Engineering Practice, 2024, 145: 105845

[61] Xiong G L , Mao J Q , Cai H B , Zhang H Q . Optimal reservoir operation for transport timescales using an integrated methodology. Environmental Research Letters, 2023, 18(11): 114025

[62] Ji B , Yuan X H , Yuan Y B . Orthogonal design-based NSGA-III for the optimal lockage co-scheduling problem. IEEE Transactions on Intelligent Transportation Systems, 2017, 18(8): 2085–2095

[63] Zheng G K , Li Y S , Zhou Z , Yan R Q . A remaining useful life prediction method of rolling bearings based on deep reinforcement learning. IEEE Internet of Things Journal, 2024, 11(13): 22938–22949

[64] Li J L , Cao X , Chen R X , Zhang X , Huang X Z , Qu Y Z . Graph neural network architecture search for rotating machinery fault diagnosis based on reinforcement learning. Mechanical Systems and Signal Processing, 2023, 202: 110701

[65] Qin Y , Liu X W , Mao Y F . Multilabel decoupling diagnosis for compound faults by improved deep Q-network. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 1–10

[66] Jin F , Chen Z , Wang J T , Yang J . Practical procedure for predicting non-uniform temperature on the exposed face of arch dams. Applied Thermal Engineering, 2010, 30(14–15): 2146–2156

[67] Li H H , Fu S J , Li G , Hu G F . FEA of effects induced by diurnal temperature variation on downstream surface of Xiaowan arch dam. Advances in Civil Engineering, 2021, 2021(1): 6300387

[68] Gu C S , Cui X R , Gu H , Yang M . A new hybrid monitoring model for displacement of the concrete dam. Sustainability, 2023, 15(12): 9609

[69] Deng Z H , Gao Q , Huang M S , Wan N , Zhang J W , He Z . From data processing to behavior monitoring: A comprehensive overview of dam health monitoring technology. Structures, 2025, 71: 108094

RIGHTS & PERMISSIONS

Higher Education Press