基于多agent的多级供应链需求预测研究

用多agent系统模拟一个多级供应链,通过各个agent来运行供应链中的各个实体,构建了基于多agent系统的多级供应链模型;在此基础上利用遗传算法预测需求,计算供应链中各个实体的需求量;并和其它需求预测方法进行对比;对比结果表明,该方法优于移动平均值、指数平滑等需求预测方法。     

A validated instrument measuring students' perceptions on plastinated and three-dimensional printed anatomy tools

Due to the modernization of the medical curriculum and technological advancements, anatomy education has evolved beyond cadaveric dissection alone. Plastination techniques, three-dimensional (3D) modeling, and 3D printing technologies have progressively gained importance. However, there are limited valid and reliable surveys to evaluate students' perceptions of these new anatomy tools. Hence, this study aimed to develop a validated instrument to measure students' learning satisfaction, self-efficacy, humanistic values, and perceived limitations of plastinated and 3D printed models. A 41-item survey (five-point Likert scale, 1 = strongly disagree to 5 = strongly agree) was administered to Year 1 undergraduate medical students following a randomized controlled crossover study that evaluated plastinated and 3D printed cardiac and neck models. Ninety-six responses were received, and a factor analysis was performed with the Kaiser–Meyer–Olkin sampling adequacy of 0.878. The confirmatory factor analysis yielded a 4-factor, 19 items model that had a good fit with the latent constructs of $x$² (147) = 211.568, P < 0.001, root mean square error of approximation = 0.068, root mean square residual = 0.064, comparative fit index = 0.946, and Tucker Lewis index = 0.937. The Cronbach's alpha for the individual factors ranged from 0.74 to 0.95, indicating good internal consistency. This demonstrated a psychometrically valid and reliable instrument to measure students' perceptions toward plastinated and 3D printed models.     

用因子图推断改进智能盘点机器人定位与地图构建

人工智能和图书馆的结合,促进了智慧图书馆的快速发展。图书馆智能盘点机器人的出现解决了图书错架和盘点的需求。盘点机器人在室内工作,自身的定位和导航以及对环境的适应是机器人要解决的重要问题。目前图书馆智能盘点机器人在这方面有不足的地方,文章结合国内外SLAM的研究进展,参考其他行业机器人SLAM实践,用因子图推断方法改进机器人同步定位与地图构建,取得了一些初步的成果。