节点文献

油漆调色系统的算法改进和功能扩展

【作者】 富丽

【导师】 余隋怀;

【作者基本信息】 西北工业大学 , 管理科学与工程, 2002, 硕士

【摘要】 根据现有计算机辅助工业设计系统中色彩设计和色彩生产难以协调统一的问题,以及油漆厂的实际需求,本文以提高BP网络的学习精度和速度为研究目标,在原有的“基于神经网络的油漆调色系统”的基础上,进行BP网络学习算法的改进,并完成系统功能扩展。 本文围绕研究目标,从理论基础、技术基础、算法改进和功能扩展等几个方面进行了详细的论述:阐明了色彩、色度的基本理论知识;创建了BP网络学习样本数据库,经过实际的算法分析与测试,找到了一种适用于实际应用的有效的BP改进算法;实现了油漆调色功能的网络化,做到资源共享。 研究测试的结果表明:BP改进算法能够弥补原有算法精度速度不够理想的不足,解决了电脑配色的问题。

【Abstract】 In accordance of the inconsistent status between color design and color production in current Computer-Aided Industrial Design System and the practical requirement of oil paint factory, the paper investigates and puts forward an new BP algorithm to improve the studying precision and studying speed, and extends the system function on the basis of the original "Neural Network-based oil paint dividing system".The paper gives a detailed discuss on its theory, technique, algorithm improvement and function extending. It not only expounds the basic knowledge on color and chroma, builds the sample database for the studying of BP network, but also finds an effective BP algorithm adapted to actual requirements after high analysis and test. It also utilizes some internet technique to realize the network edition of the dividing function and fulfill the sharing of resource.The testing result shows that the improved BP algorithm can supply a gap of the old BP algorithm and solve the problem of computer dividing.

  • 【分类号】TQ630
  • 【被引频次】3
  • 【下载频次】140
节点文献中: 

本文链接的文献网络图示:

本文的引文网络