文献类型：期刊文献

英文题名：A neural network-based multi-dimensional simulation modeling approach for food microbial growth

作者：Laisheng X.[1]

机构：[1]Network and Education Technology Center, Guangdong Ocean University, Zhanjiang 524088, China

年份：2012

卷号：6

起止页码：400

外文期刊名：Advanced Science Letters

收录：Scopus(收录号：2-s2.0-84861505152)

语种：英文

外文关键词：Food Microbial Growth; Fresh-Cut Lotus Roots; Multi-Dimensional Simulation Modeling; Neural Network

外文摘要：With regard to neural networks applied to simulation and modeling for microbial growth, some scholars carried out beneficial attempts. But up to now there is not a systemic method to be established for guidance. In view of this kind of situation, in the presented paper a neural network-based multi-dimensional simulation modeling approach for food microbial growth was systemically presented and its validity was tested through simulation experiments. At first, a neural network-based multi-dimensional simulation modeling approach is designed by building its basic structure, defining concepts of N-dimensional models, and introducing the principles of its simulation modeling, etc. Next, the validity of the presented simulation modeling approach was validated: (1) Taking data of fresh-cut lotus roots stored in 4 °C, 8 °C, 20 °C as an example, one-dimensional neural network simulation model for microbial growth was establishment. Its performance was simulated and compared with traditional microbial growth model Gompertz. (2) A detailed two-dimensional simulation modeling process was described and at last a 3D simulation graphics for it was given, too. Simulation results showed that the models established by this approach have higher precisions. Thus the approach presented in the paper is scientific and feasible. The models established by this approach need fewer experimental data, but required data models can be built in entire multi-dimensional space that can provide continuous interpolation at any point in the space as predicted values, by this way experimental costs are greatly reduced. Therefore, the approach presented here provides a new way for the description of food microbial growth, and it could be applied to the whole food processing industry, bringing rapid assessment and prediction for its production quality and product safety. ? 2012 American Scientific Publishers. All rights reserved.