作者：
Rong-Ji Wang;Jianbing Li 0004;Fenghua Wang;Xinhua Li;Qingding Wu
期刊：
International Journal of Manufacturing Research,2009年4(3):362-373 ISSN：1750-0591
通讯作者：
Wang, R.-J.(wangrj6623@yahoo.com)
作者机构：
[Qingding Wu; Rong-Ji Wang; Jianbing Li 0004; Fenghua Wang; Xinhua Li] Central South University of Forestry and Technology, College of Mechanical and Electrical Engineering, Changsha 410004, China
通讯机构：
College of Mechanical and Electrical Engineering, Central South University of Forestry and Technology, China
摘要：
The effects of process parameters on density of the part prepared by Selective Laser Sintering (SLS) were modelled, using an Artificial Neural Network (ANN) with a feed forward topology and a back propagation algorithm. The inputs of the ANN are the process parameters, including layer thickness, hatch spacing, laser power, scanning speed, temperature of working environment, interval time and scanning mode. The output of the ANN is the density. The experimental investigation results show that the ANN model may be used to analyse the relationship between the process parameters and the density of the SLS part quantitatively. [Received 12 September 2007; Revised 14 March 2009; Accepted 24 March 2009]
摘要：
Selective laser sintering (SLS) is an attractive rapid prototyping (RP) technology capable of manufacturing parts from a variety of materials. However, the wider application of SLS has been limited, due to their accuracy. This paper presents an optimal method to determine the best processing parameter for SLS by minimizing the shrinkage. According to the nonlinear and multitudinous processing parameter feature of SLS, the theory and the algorithms of the neural network are applied for studying SLS process parameters. The process is modeled and described by neural network based on experiment. Moreover, the optimum process parameters, such as layer thickness, hatch spacing, laser power, scanning speed, work surroundings temperature, interval time, and scanning mode are obtained by adopting the genetic algorithm based on the neural network model. The optimum process parameters will be benefit for RP users in creating RP parts with a higher level of accuracy.
作者机构：
[李新华; 吴庆定; 王荣吉; 周建] College of Mechanical and Electrical Engineering, Institute of Processing Technologies of Materials, Central South University of Forestry and Technology, Changsha 410004, China
通讯机构：
College of Mechanical and Electrical Engineering, Institute of Processing Technologies of Materials, Central South University of Forestry and Technology, China
