Module partition method based on hybrid fuzzy clustering

Author(s):  
Xiaoyan Teng ◽  
Taijia Zhang
2011 ◽  
Vol 308-310 ◽  
pp. 273-279
Author(s):  
Yan Hui Chen ◽  
De Jian Zhou

This paper presents a new method of product module partition based on the fuzzy clustering analysis. This method demonstrates the relevant definitions and calculation methods of the initial partition, min-max partition, submodule relevancy and module aggregation etc., and establishes the incidence matrix to respectively carry out the initial partition for the products and calculation of min-max partition according to various incidence relations between parts and components. Taking the submodule as computing unit in each module set, this paper carries out the fuzzy cluster analysis to obtain the module partition results of the products, and finally demonstrates the rationality and effectiveness of this method by taking the example of the working units of the wheel loaders.


2017 ◽  
Vol 166 ◽  
pp. 1-6 ◽  
Author(s):  
Alfonso Pérez-Garrido ◽  
Francisco Girón-Rodríguez ◽  
Andrés Bueno-Crespo ◽  
Jesús Soto ◽  
Horacio Pérez-Sánchez ◽  
...  

2010 ◽  
Vol 139-141 ◽  
pp. 1540-1544
Author(s):  
Li Jing Wang ◽  
Tao Xi ◽  
Yin Feng Zhou ◽  
Run Hua Tan

Most of present product module partition methods are based on product function partition and use fuzzy clustering algorithm, but these methods are not only complex in implementation but also difficult to meet the requirements of product development oriented to product lifecycle. By analyzing interactive effects of product components in product lifecycle, a new method for product module partition is put forward. Firstly, LSSVC which has fast calculation speed and high accuracy is used to illustrate the generating process of modules, so several module partition schemes are obtained. Secondly, module partition schemes which are got by LSSVC and other methods of module partition are evaluated to get the most reasonable module partition scheme. Finally, widely-used speed reducer as an example is provided to illustrate the validity and rationality of the proposed approach.


2012 ◽  
Vol 479-481 ◽  
pp. 1722-1727 ◽  
Author(s):  
Wen Xian Tang ◽  
Chun Yan Wu ◽  
Fei Wang ◽  
Qiu Yun Huang ◽  
Bao Ma

The modular design offers insurement for effective development of serial and diversify products, and hence to satisfy the demand of the market. Through researching the modular design in function and structure, the process schem of modular design characterizing anchor and windlass products was proposed. By studying the module partition method of function and structure, the anchor and windlass was classified as modules of power, modules of driving and changing, modules of control, modules of performing. Based on the program language and three-dimensional software, the computer-aided modular design system for anchor and windlass was established, which satisfies the parametric design for anchor and windlass.


2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Hong Xia ◽  
Qingyi Dong ◽  
Hui Gao ◽  
Yanping Chen ◽  
ZhongMin Wang

It is difficult to accurately classify a service into specific service clusters for the multirelationships between services. To solve this problem, this paper proposes a service partition method based on particle swarm fuzzy clustering, which can effectively consider multirelationships between services by using a fuzzy clustering algorithm. Firstly, the algorithm for automatically determining the number of clusters is to determine the number of service clusters based on the density of the service core point. Secondly, the fuzzy c -means combined with particle swarm optimization algorithm to find the optimal cluster center of the service. Finally, the fuzzy clustering algorithm uses the improved Gram-cosine similarity to obtain the final results. Extensive experiments on real web service data show that our method is better than mainstream clustering algorithms in accuracy.


Sign in / Sign up

Export Citation Format

Share Document