针对算力网络中的度量与建模问题,从算力网络度量体系的狭义和广义定义出发,阐述了当前算力网络度量与建模的研究进展,包括算力网络资源度量和算力网络需求度量两大部分.站在算力网络服务的角度,提出了以服务为中心的算力网络度量与建模方案的思考,并对该方案当前仍存在的问题和所面临的技术挑战进行了归纳和讨论.

为了更准确的对图像中的目标物体进行识别,科学家研发了目标检测技术,其是计算机视觉的重要组成.最近几年深度学习越来越普及,基于深度学习的目标检测技术具有更高的准确率和检测效率,可以应用在很多领域,然而应用过程中还存在一些影响深度学习目标检测的发展的问题急需解决,基于此,本文对目标检测概念进行了简单介绍,并对深度学习目标检测算法进行了归纳总结,对深度学习目标检测中存在的问题和未来发展进行了探究,希望可以为目标检测实现更好发展提供助力.

机械零件几何要素的形位误差是零件的主要质量指标之一,影响机器整体的质量.直线度又是形位误差中重要的组成部分,本文主要研究基于Labview的小尺寸零件直线度误差的检测系统,该系统可以采用不同方法评定直线度误差,具有准确,快速,可靠等特点.

针对突发事件发生后,商业企业需要调整采购决策,重新进行供应商选择问题,建立了突发事件供应商评价指标体系,该指标体系中既有供应商的基本指标,又存在反映突发事件特性的指标.将层次分析法、模糊综合评判和全排列多边形图示指标法运用到突发事件供应商选择当中,建立了突发事件供应商选择模型.该模型能较好地解决评价指标难以准确统计和量化的问题,合理地解决了单级指标和多级指标同时作为评价指标的权重问题,使选择结果更准确客观.

The experimental teaching and social practice is become more and more important in the process of course-teaching. In response to the shortcomings in traditional experiment of Test technology, hierarchical experimental teaching structure system is introduced. By combing with the characteristics of the course, the experiment are divided into multiple levels to meet different needs of students. The purpose is to mobilize the enthusiasm and initiative of the students, improve the quality of teaching.

Presents a micro-assembly control system architecture based on "master computer+slave controller+mutil-sensors" which could assemble micro-plate parts automatically and more reliably.Absorbing micro-gripper was designed and presented to grip the plate parts instead of the jaw-type gripper.Automatic optical alignment system based on single camera was designed complete the accurate alignment between the part and the substrate.The presented micro-assembly system could serve the purpose of assembling micro-plate structure successfully.

针对北京联合大学注重学以致用的教学目标,对"电机及驱动技术"课程进行了重新的审视。传统电机教学方法注重理论教学,希望学生毕业后能进行进一步的深造。而北京联合大学的教学目标希望学生毕业后能迅速适应自己的工作环境。因此,该课程的教材改革方向力求在深度和广度上做到"少而精"和"理论联系实际"。基础理论以必需、够用为度,同时注重电力拖动中的工程实际问题,以培养技术应用型人才为目标。

A detection model of mine-used infrared gas sensor was established.Through calculation and curve fitting,the relationship between gas concentration and output voltage of infrared gas sensor was gotten and inversion test for the detection model was(done.) The test result showed that measuring precision of the detection model meets field's requirement.