Title :
The security analysis of complicated mechanical structure based on PSO-SVM
Author :
Shan, Xianlei ; Zhang, Tian ; Ma, Tianran ; Wang, Yan ; Xing, Yan
Author_Institution :
Sch. of Mech. & Electr. Eng., China Univ. of Min. & Technol., Xuzhou, China
Abstract :
The loading condition of complicated structure is very sophisticated. The finite element method in doing mechanical loading analysis possesses certain limits because of the difficulty in modeling and mesh generation. The use of particle swarm optimization to improve the parameters of vector machine can avoid the blindness caused by men in selecting parameters, thus solving the problem produced by traditional and uncertain methods and promoting the training speed and the ability to predict and popularize of the model. Based on in depth research, this paper comes up with a new security analysis. The new method adopts finite element method and gets various groups of loading effects, then establishes the security analysis of complicated mechanical structure based on PSO-SVM. So the structural security analysis can be directly conducted by taking the results of the infinite element analysis as the training sample sets. Lastly, this paper does example analysis by adopting this method. By doing so, the security analysis of complicated structure is greatly simplified. The effectiveness of this method is proved through the comparison of analysis results and the actual ones.
Keywords :
finite element analysis; particle swarm optimisation; structural engineering computing; support vector machines; PSO-SVM; finite element method; mechanical loading analysis; mechanical structure; particle swarm optimization; structural security analysis; Analytical models; Deformable models; Kernel; Optimization; Security; Support vector machines; PSO-SVM; complicated structure; infinite element analysis; loading effect; security analysis;
Conference_Titel :
Computer, Mechatronics, Control and Electronic Engineering (CMCE), 2010 International Conference on
Conference_Location :
Changchun
Print_ISBN :
978-1-4244-7957-3
DOI :
10.1109/CMCE.2010.5610272