Homology Priority Task Scheduling in μC/OS-Ⅱ Real-Time Kernel

μC/OS-Ⅱ is an open source real-time kernel adopting priority preemptive schedule strategy. Aiming at the problem of μC/OS-Ⅱ failing to support homology priority tasks scheduling, an approach for solution is proposed. The basic idea is adding round-robin scheduling strategy in its original scheduler in order to schedule homology priority tasks through time slice roundrobin. Implementation approach is given in detail. Firstly, the Task Control Block （TCB） is extended. And then, a new priority index table is created, in which each index pointer points to a set of homology priority tasks. Eventually, on the basis of reconstructing μC/OS-Ⅱ real-time kernel, task scheduling module is rewritten. Otherwise, schedulability of homology task supported by modified kernel had been analyzed, and deadline formula of created homology tasks is given. By theoretical analysis and experiment verification, the modified kernel can support homology priority tasks scheduling, meanwhile, it also remains preemptive property of original μC/OS-Ⅱ.     

正交双向网格图及方向优先算法

目前建筑工程的工程量计算基本上采取的是“统筹法”和“公式法”,这两种方法对工程的概算、预算的精确度及工作效率都产生了那样或这样的问题。本文从建筑图的正向分析开始,进而讨论了工程量计算模型的定义,在定义基础上给出了轴线的方向函数以及方向优先算法和具体实现过程。     

具有优先级的深度确定性策略梯度算法在自动驾驶中的应用

深度确定性策略梯度(deep deterministic policy gradient,DDPG)算法在自动驾驶领域中应用广泛,但DDPG算法因采用均匀采样而导致低效率策略比例较高、训练效率低、收敛速度慢等.提出了基于优先级的深度确定性策略梯度(priority-based DDPD,P-DDPG)算法,通过优先级采样代替均匀采样来提升采样利用率、改善探索策略和提高神经网络训练效率,并且提出新的奖励函数作为评价标准.最后,在开源赛车模拟(The Open Racing Car Simulator,TORCS)平台上对P-DDPG算法的性能进行了测试,结果表明相对于DDPG算法,P-DDPG算法的累积奖励在25回合之后就有明显提升而DDPG在100回合之后训练效果才逐渐显现,提升约4倍.P-DDPG算法不仅训练效率提升且收敛速度加快.     

中断实验的设计及外部中断源扩展的实现

针对授课过程中学生对中断优先权以及中断屏蔽等概念不理解的情况设计了一款简单实用的由8031组成的中断实验系统．通过该实验，可以直观地看到中断、中断优先权、中断屏蔽、外中断源扩展的过程，使学生较容易地理解掌握单片机控制系统中中断的处理方法，提高了实验效果．     

China'''' s Fundamental Research in Energy Utilizations and Environment

The progress in the science of energy utilizations will act crucial effect on the developments of energy science and technology, which will then promote social and economical developments and fulfill requirements for the national strategic objectives. For the sake of sustainable development, a harmonious blend of energy utilizations and environment considerations will become one of the vital topics in the future research area of energy science. It is suggested that clean and high-efficiency utilization of traditional or fossil energy resources, fundamental investigations on the energy and environment theory, renewable energy utilizations, and the development of nuclear energy are selected as priority research areas during the period of the Tenth Five-year Plan of China, according to the development trend of the world energy science and the research background of Chinese energy science. It is expected to promote the interdisciplinary investigations in the science of energy u-tilizations and provide scien     

不耐烦等待信元的优先权排队

研究具有两类信元的带优先权的M／M／1排队系统，两类信元到达为相互独立泊松过程，两类信元分别在各自有限的缓冲区中排队，第一类信元较第二类信元具有强占优先权，同时第一类信元是不耐烦的．笔者采用矩阵分析的方法给出了两类信元各自的稳态分布，并作了相应的性能分析。     

IEEE802.4令牌总线网实时应用性能分析算法

对符合ＩＥＥＥ８０２．４标准的令牌总线网的性能指标评估给出了１种新的计算方法。重点分析了ＩＥＥＥ８０２．４标准的优先权机制，并考虑了令牌总线网络中的２种计时器ＴＨＴ，ＴＲＴ控制和信包缓冲区容量有限的性能指标评在问题，克服了以往不能分析优先权机制令牌网性能的局限。     

快速公交车辆平面交叉口信号优先实现方法

提出以等待延误大小为约束条件在设置快速公交（BRT）专用进口道的十字平面交叉口给予BRT车辆有条件信号优先．有条件优先含义是只给予满足特定条件的BRT车辆信号优先．具体做法是利用检测系统检测BRT辆抵达交叉口遇红灯开始排队待行时刻，依据所有待行BRT车辆的最大等待延误判断是否应提供信号优先服务．信号优先实现方法包括实时调整优先相位、优先相位分隔多次出现和压缩非优先相位．与定时信号控制相比，感应控制根据BRT车辆等待延误大小进行信号调整，提高BRT系统服务质量．     

嵌入式RTOS中任务调度问题研究

针对嵌入式实时系统任务的管理和调度的特点,多数采用的是基于优先级的可抢占式调度策略,任务在运行的时候可能被更高优先级的任务中断。分析了嵌入式 RTOS 中的最高就绪任务的查找算法,指出该调度算法存在的不同优先级的反转问题,并且针对这一问题进行研究,并给出了相应的解决方案。     

灰色聚类协调分析模型及其应用

以灰色聚类方法为基础，利用协调优先指标与非协调优先指标的性质，建立了一种新的综合评判模型。实例分析结果表明，该模型具有科学合理、计算简便、分辨率高等特点。