Haiyan Li

Thermomechanical coupling effects significantly influence the dynamic response of flexible multibody systems operating in thermal environments. Accurate simulation using conventional projection-based reduced-order models remains challenging due to strong nonlinearity and the time-varying nature of thermal fields. This work proposes a projection-free sparse-solving framework for thermomechanical coupling dynamics. The elastic and thermal fields are represented using sparse POD coefficients, and the governing equations are directly sampled online without Galerkin projection. The sparse coefficients are recovered via l₁ norm optimization at each time step. To ensure stable recovery under thermal-mechanical coupling, a unit-norm tight frame-based preconditioner is introduced to reduce the coherence of the underdetermined system matrix. The proposed method is validated through three numerical examples. Results show that the method maintains stable accuracy over long-time simulations, reduces computational time by over 40%, and exhibits improved robustness compared with the discrete empirical interpolation method in thermomechanical coupling problems. The adaptive basis selection capability and the necessity of unit norm tight frame preconditioning are confirmed. The method offers an efficient and reliable alternative for thermomechanical multibody dynamics simulation.

To address the limitations of traditional convolutional neural networks (CNNs) in high‐voltage transmission‐line fault diagnosis, where fixed local receptive fields are difficult to represent broadband transient multiscale features and purely data‐driven paradigms lack physical consistency, a physics‐guided Fourier operator convolutional network (PG‐FCNN) is proposed. First, a frequency‐domain convolution module based on a Fourier neural operator is designed to extract multiscale transient features within the effective frequency band using a global receptive field. Second, a physics decoder based on the telegrapher’s equations is incorporated in parallel, with the residual of the distributed‐parameter transmission‐line model used as a physics‐based regularization constraint. A dataset containing 127,450 samples is constructed using real fault‐recording data from 87,500 kV and 220 kV transmission lines in a regional power grid in China from 2015 to 2025. Experimental results show that PG‐FCNN achieves a diagnostic accuracy of 98.82%, with macroaveraged precision, recall, and F1 score of 96.94%, 97.61%, and 97.26%, respectively. In the ablation experiment, the macroaveraged F1 scores of the baseline CNN, CNN + physics, F‐CNN, and PG‐FCNN are approximately 94.3%, 95.7%, 96.5%, and 97.2%, respectively. Compared with F‐CNN, PG‐FCNN further improves macroaveraged precision, recall, and F1 score by approximately 0.8, 0.8, and 0.7 percentage points, respectively. Across five independent runs, the accuracy and macroaveraged F1 score of PG‐FCNN are 98.82% ± 0.02% and 97.26% ± 0.03%, respectively. After removing the physics decoder during inference, the computational cost is only 0.33 GFLOPs. Compared with CNN‐LSTM, ResNet‐18, and transformer, PG‐FCNN achieves the highest accuracy and macroaveraged F1 score and maintains diagnostic accuracies of 93.61%–95.33% under nonideal operating conditions, including noise, CT saturation, high‐resistance grounding, and channel loss. The results can provide technical support for high‐voltage transmission‐line fault diagnosis under complex operating conditions.

The telescopic boom forklift is a very efficient engineering equipment because of its distinctive boom system. However, to speed up product development and reduce costs, surrogate models should be constructed to replace computationally expensive co-simulation models for design optimization. In this paper, telescopic boom forklift design is investigated using deep transfer learning and is coupled with a mode pursuing sampling algorithm to determine and propose optimal solutions based on observed boom amplitude. A workflow that combines uncertainty analysis-based active data development, deep surrogate modeling, and global optimization is proposed. It considers the common characteristics of the existing product simulation data, the personality characteristics of the current product, and the learning characteristics of the neural network when searching for the best surrogate. We demonstrated that tuned Multiplayer perceptron (MLP) can make accurate predictions and be used to create an acceptable surrogate model. The correlation coefficient (R) indicator can reach 9.34 when only 25 groups of data are used. Additionally, a 38.28% reduction in Root mean square error (RMSE) and a 26.25% reduction in Maximum absolute error (MAE) can be achieved when compared to benchmarks. Generated designs attain the maximal boom amplitude that is lower by more than 46% and the fluctuation range from (3-35) MPa to (8-19) MPa. Overall, we have demonstrated that the proposed framework has merit and can be used as a viable methodology in telescopic boom forklift design optimization. Moreover, it can also provide useful references for the simulation optimization of other complex electromechanical products.

The one-dimensional velocity interferometer system for any reflector (VISAR) is widely used in inertial confinement fusion (ICF). Compressed ultrafast photography (CUP) is recently combined with VISAR to enable two-dimensional imaging, in which, the time-varied two-dimensional fringe images reconstruction from one image is a challenging problem. In this Letter, a new l ₁ norm based sparse reconstruction approach is presented, to the best of our knowledge, in which: (1) fringe images are sparsely represented with a few coefficients over the discrete cosine transformation (DCT) basis; (2) the image reconstruction is formulated as an l ₁ norm based sparse coefficients optimization problem; and (3) the two-step iterative shrinkage/threshold algorithm combined with a soft-thresholding operator is proposed to efficiently solve such a problem. Finally, the results show that, compared with the usual method, reconstructed fringe images with clear boundaries and good continuity are obtained. Additionally, the maximum relative error of the velocity is reduced from 14% to 8%, which is a reduction of approximately half.

Simulation and optimization methods have been widely used in forklift design due to their cost-effectiveness. However, this type of method involves challenges such as the accuracy of the simulation model and the simulation solution time. These challenges reduce the stability and precision of the surrogate model and hence generate further optimization errors. In this paper, a multi-objective surrogate modeling (MSM) method for telescopic boom forklifts based on closed-loop transfer learning is proposed in order to solve these challenges. The MSM consists of the following two steps: to pre-train an initial deep neural network model (deep model) with a large amount of existing simulation data from the same type of forklift and to transfer the model with a small amount of measurement data collected on the current forklift. A general framework for deep neural network (DNN) training is introduced to improve the approximation ability of the initial model. Moreover, a novel uncertainty-analysis-based sampling method is suggested for measurement data development, and combined with transfer learning to form a closed-loop mode to improve the stability of the final model. The superiority of MSM is demonstrated through comparative studies with the fine-tuning method on a telescopic boom forklift with two objectives. The experimental results show that the Correlation coefficient (R) of the deep model can reach 0.9971 by using only 80 sets of training data. In addition, it can also achieve an improvement of at least a 13.25% reduction in Root Mean Squared Error (RMSE) and a 9.19% reduction on average in Maximum Absolute Error (MAE), as well as stronger robustness compared to the benchmarks. Furthermore, it will provide a valuable reference for the simulation optimization of complex electromechanical products.