Transformers have achieved promising results on aeroengine remaining useful life (RUL) prediction, but they still have several limitations: 1) Aeroengine domain knowledge, which contains rich information that can reflect the aeroengine's health statue, is largely ignored in modeling process; 2) Traditional transformer ignores the valuable degradation information from other time scales. To address these issues, a novel domain knowledge-augmented multiscale transformer (DKAMFormer) is developed by integrating domain knowledge and multiscale learning to improve the prognostic performance and reliability. First, to obtain rich and professional aeroengine domain knowledge, multiple detail and complete knowledge graphs (KGs) are established based on the working principle of aeroengine, including aeroengine structure, components working characteristics and sensor parameters. Second, the domain knowledge contained in KGs is convert to embedded vector by KG representative learning, which are then utilized to strengthen and enrich the original multidimensional time-series (MTS) monitoring data, aiming to intergrade domain knowledge and monitoring data to train DKAMFormer. Third, to learn rich and complementary degradation features, a novel multiscale time scale-guided self-attention (MTSGSA) mechanism is designed, which maps original MTS into different time-scale feature spaces, and then employs multiple independent self-attention head to extract the degradation features from different time-scale spaces. Finally, through a series of comparative experiments on the public CMAPSS and N-CMAPSS datasets and compared with 17 SOTA methods, the developed DKAMFormer significantly improves the RUL prediction performance under multiple operation conditions and degradation modes.
In contact analysis, reducing the computing time has been an issue under the premise of ensuring calculation accuracy around the region with violent stress changes. To improve computational efficiency for contact analysis in flexible multibody system, this paper proposes an adaptive local mesh refinement and coarsening approach based on analysis-suitable T-splines (ASTS). First, the kinematic model of thin plate is established based on analysis-suitable T-spline surface, and large deformation of flexible thin plate is described by the elastic model created by nonlinear Green–Lagrange strain. Second, to reduce computing time in contact analysis and ensure analysis accuracy, based on contact state and refinement distance, an effective adaptive local element mesh update method is proposed, which only refine locally on subject's refinement region and integrate redundant elements to reduce the degree-of-freedom (DOF) of system. Third, to analyze the system with varying mesh, a new solving algorithm with dynamic variables and geometry update routine is developed. Finally, performance of the proposed method in static and dynamic simulation is validated by four numerical examples. Results and consuming time of ASTS-based varying mesh prove the feasibility of the proposed method in contact problems.
1 January 2025 in Journal of Advanced Manufacturing Science and Technology
In view of the collapse accidents of most high-attached tower cranes under typhoons, this paper makes an in-depth analysis on the strength of tower cranes under typhoons and discusses the maximum attached height of attached tower cranes under typhoons. First, based on the average wind speed profile of typhoon exponential law, the exact calculation formula of the maximum bending moment of the tower crane under ideal conditions and the practical calculation formula under different loads are derived. Secondly, the overhang section of the attached tower crane is equivalent to the elastic support model, and the exact calculation formula of the maximum bending moment subjected to this model and the practical calculation formula under different loads are derived. Finally, the principle of strength equality is proposed, stating that the maximum bending moment of the ideal tower crane equals that of the attached overhang tower crane. This allows us to establish the relationship between the maximum height of the overhang section and the maximum independent working height of the tower crane. The maximum height coefficient of the overhang section is further compared to evaluate the strength and stability of the overhang section in the typhoon state more comprehensively. In general, assuming the axial force is 0, the overhang height ratio under a typhoon can be checked according to the overall stability formula. In the case of different coaxial force coefficients, the ratio of overhead height will be different.