基于全局和声搜索优化的平版印刷油墨预置  

Global harmony search optimization based ink preset for offset printing

作　　者：林剑[1,2] 许力[1] 

机构地区：[1]浙江大学电气工程学院,杭州310027 [2]杭州电子科技大学新闻出版学院,杭州310018

出　　处：《仪器仪表学报》2010年第10期2248-2253,共6页Chinese Journal of Scientific Instrument

基　　金：国家863计划项目(2006AA040202-03)资助

摘　　要：采用最小二乘支持向量机建立平版印刷油墨预置模型,并提出一种全局和声搜索的优化算法来对该模型的参数进行优化。和声搜索算法是一种模拟乐队调音获得完美和声过程的元启发优化算法。将粒子群优化算法中基于种群最优来修正个体的思想与常规的和声搜索算法相结合,以实现全局和声搜索,并应用于油墨预置模型参数的优化。根据实际印刷机数据进行了仿真实验,结果验证了所提算法的有效性。An ink preset model for offset printing was established by using the least square support vector machine, and a global harmony search （GHS） algorithm was proposed for the optimizing of the model parameters. The harmony search （HS） algorithm is a meta-heuristic algorithm by analogy to the music improvisation process to obtain better harmony. In the particle swarm optimization （PSO） algorithm, individuals were modified by referring to the best in the population. This idea was borrowed and combined with the regular HS algorithm to perform globally harmony search, and was further employed to optimize the parameters of the ink preset model. Simulations were conducted with data from the Roland printing machine, and results are presented to show the effectiveness of the proposed scheme.

关 键 词：和声搜索 油墨预置 墨键开度 最小二乘支持向量机 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]