It is crucial to recognize unknown faults for the intelligent fault diagnosis of bogie mechanical transmission components. This is essential to ensure the safety of trains in long-term service. However, the unknown faults are hardly represented accurately because of the influence of vibration signals multipath propagation. This results in a challenge to efficiently construct open-set decision boundaries (OSDB). An open-set fault diagnosis (OSFD) method based on multichannel feature-enhanced placeholder learning is proposed to improve diagnostic reliability. It utilizes the multichannel vibration signals collected by sensors located at different positions to construct a placeholder learning network based on the association of multichannel features. It can effectively improve the association of unknown class presentations and boost the effect of unknown fault identification. Besides, the placeholder learning network based on dummy classifier enhancement is proposed to strengthen the representations of known classes and ensure the generalization of OSDB. The proposed approach is validated with case studies constructed by high-speed train bogie fault datasets and the data challenge public datasets of PHM-Beijing 2024. The results demonstrate that the proposed method can enhance the effectiveness of OSFD.

According to the property of numerical control lathe in material removal simulation, this paper presents the idea of space division, convert the moving relation between tools and workpiece into logical estimation of space position, and define relative rules to avoid complex computation. In this paper, the algorithm is studied in details, based on which we developed the simulation system of numerical control lathe under the open architecture. The practice of lots of instances shows that the idea has these characteristics of good efficiency and real time, and can be applied in the development of numerical control lathe system.

A GPGPU-based collision detection algorithm is proposed. Firstly, the information of OBB hierarchy tree and triangles of tested objects are mapped into some data textures designed for GPGPU-based calculation, such as triangle vertex textures, bounding box size texture, tree node relationship texture, etc., then these textures are downloaded to GPU to complete the data preparation. Secondly, the whole collision detection is executed on GPU, in which three key contents are fulfilled: reading necessary data from related textures correctly by order coordinate method and index coordinate method, detecting the intersection between triangle and OBB, triangle and triangle through a collision detection index array. Lastly, collision detection results are outputted to a texture by FBO technology and read back to CPU for post-processing. The data transmission between CPU and GPU is only twice, which reduce the time to read data. Testing results show that the detecting speed of the algorithm proposed in this paper is faster than the similar CPU-based algorithm obviously with the increasing complexity of tested objects, while keeps the same precision.