Journals
  Publication Years
  Keywords
Search within results Open Search
Please wait a minute...
For Selected: Toggle Thumbnails
Bearing Fault Diagnosis Method Based on Down-Sampling in Frequency Domain and CNN
ZHOU Xiangyu, MAO Shanjun, LI Mei
Acta Scientiarum Naturalium Universitatis Pekinensis    2023, 59 (2): 251-260.   DOI: 10.13209/j.0479-8023.2022.098
Abstract290)   HTML    PDF(pc) (3457KB)(120)       Save
In the industrial field, the original fault signals collected during the operation of the equipment have the characteristics of strong noise and multiple working conditions. Most of previous data-driven fault diagnosis methods for bearings have relatively weak anti-noise ability and generalization ability. To solve these problems, a novel bearing fault diagnosis method based on down-sampling in frequency domain and convolutional neural network (CNN), called Ds-CNN, is proposed. Down-sampling in frequency domain consists of maximum down-sampling with bias and noise transverse truncation, which can realize data augmentation, reduce the difference between samples in frequency domain, and reduce the influence of noise on signals in frequency domain. The CNN model based on frequency domain signals can automatically extract fault features from signals after down-sampling and complete the identification and classification of bearing faults. The results of the experiment show that Ds-CNN has higher recognition accuracy than common models under strong noise environment and multiple working conditions. 
Related Articles | Metrics | Comments0
Research on Key Techniques of Smoothing Transition of LOD Texture Blending and Antialiasing Based on Shader
JIANG Zhan, LI Mei, SUN Zhenming, MAO Shanjun
Acta Scientiarum Naturalium Universitatis Pekinensis    2022, 58 (1): 113-122.   DOI: 10.13209/j.0479-8023.2021.051
Abstract506)   HTML    PDF(pc) (21165KB)(93)       Save
Aiming at the problem of LOD (level of detail) popping and aliasing when the LOD Level is switched, this paper proposes a shader-based smooth transition algorithm for LOD texture blending and antialiasing. According to the distance between the 3D model and the viewpoint, the opaque mask algorithm based on Alpha test and weighted adjacent frames antialiasing algorithm are used to generate transition materials between LODs to achieve smooth transition for LOD switching. It can not only improve the texture quality, but also ensure the fluency and authenticity of loading 3D scene. The experimental results show that, compared with the smoothing method of Unreal Engine 4, the average GPU time of the proposed algorithm is reduced by more than 8%, and the frame rate is increased by more than 8%. Compared with the existing texture smooth transition methods, this algorithm can optimize the GPU rendering performance, stabilize and improve the frame rate and maintain good visual effect, effectively solve the popping problem when the LOD Level is switched.
Related Articles | Metrics | Comments0
An Improved Map Matching Algorithm Based on HMM Model
LIU Min, LI Mei, XU Xiaoyu, MAO Shanjun
Acta Scientiarum Naturalium Universitatis Pekinensis    2018, 54 (6): 1235-1241.   DOI: 10.13209/j.0479-8023.2018.038
Abstract1887)   HTML    PDF(pc) (10916KB)(432)       Save

For it’s difficult to guarantee the accuracy and time efficiency of online map matching simultaneously, an improved online map matching algorithm based on the HMM model is proposed. Different from other global or local algorithms, the algorithm introduces reliable point to divide the trajectory, thus reducing the calculation of transition probability and the output delay of matching results. Also, the method of calculating the transmitting probability by combining distance factor and directional factor is proposed. The algorithm is verified using floating car trajectory data in Seattle. Experimental results show that the proposed algorithm outperforms the traditional HMM map matching algorithm both in accuray and time efficiency, which can meet the requirement of online map matching.

Related Articles | Metrics | Comments0
Research on the Platform of Coalmine Safety Information Sharing and Decision-Making on the Web
LIU Qiaoxi,MAO Shanjun,MA Ainai
Acta Scientiarum Naturalium Universitatis Pekinensis   
Abstract652)            Save
Based on the analysis of coalmine safety information management and coalmine workflow, the authors set forth the concept and structure about the platform of coalmine safety information sharing and decision-making on web according to the theories on gray geographical information system and computer supported cooperative work. From the viewpoints of information sharing and collaborative work, the authors put forward gray spatial information sharing model based on dynamically correcting metadata and collaborative model about coalmine safety decision-making on web, and discuss some correlative and key technologies in realization and application of the platform according to data retrieval, data transmission, data storage, data sharing and data processing.The platform is well implemented in Zaozhuang Mining Group Company.
Related Articles | Metrics | Comments0
Using RSIMWEH, Particle System and Metaball to Simulate the Process of Soil Erosion by Water on Hillslopes
SHEN Dayong,MA Ainai,LIN Hui,MAO Shanjun,NIE Xianghui
Acta Scientiarum Naturalium Universitatis Pekinensis   
Abstract591)            Save
Particle system and metaball techniques are introduced from computer graphics to geography. Combined with Remote Sensing In formation Model of Water Erosion on Hillslopes(RSIMWEH), the authors put forward the concept of erosion voxel and dynamically display the process of soil erosion by water on hillslopes in 3D space. This is a new probe for 3D GIS theory and the paper proves its feasibility through application.
Related Articles | Metrics | Comments0
Gray Geographical Information System-the Theory and Technology of Correct Geological Spatial Data Dynamically
MAO Shanjun
Acta Scientiarum Naturalium Universitatis Pekinensis   
Abstract619)            Save
Concept of GGIS(Gray Geological Information System), together with integrated data model and data structure of 2D-GGIS and TIN, of 3D-GGIS and ARTPN(Analogical Right Triangular Prism Network) are therefore put forward in the paper. It lays a basis for realizing geological data dynamic management. Moreover, the key technology of automatically creating TIN and ARTPN is discussed.
Related Articles | Metrics | Comments0