Fault tolerant routing algorithm based on the artificial potential field model in Network-on-Chip

Recent advances in semiconductor technology enable the VLSI chips to integrate hundreds of intellectual properties with complex functionality. However, as the chip scales, the probability of faults is increasing, making fault tolerance a key concern in designing the large scale chips. The fault tolerant routing algorithms can guarantee sustained communication even the faults exist. It is an efficient technique to achieve fault tolerance in Networks-on-Chip. In this paper, we propose a new model based on the theory of artificial potential field (APF) to design various fault tolerant routing algorithms. In our model, the faults are considered as the poles of the repulsive potential fields while the destinations as the poles of the attractive potential fields. Messages are attracted to destinations and repelled by faults in the combined artificial potential field. The parameters used in the proposed APF based model are optimized through theoretical analysis and simulation experiments. They can support flexible fault tolerant routing algorithms. Finally, we evaluate the performance of the proposed fault tolerant routing algorithm based on the APF model in 2D-mesh NoCs with random faults. The simulation results show that the proposed APF based model is feasible and the routing algorithm can maintain good network performance.     

Gas mixture effects on ion‐beam flux characteristics from dense plasma focus device: Investigation by the Vlasov–Maxwell equations and experiments

The Vlasov–Maxwell equations were numerically solved to calculate the ion‐beam flux from the plasma of argon and the plasma of mixtures of argon and neon. Some experiments were performed to measure the ion beam from the Amirkabir plasma focus (APF) device. The calculations have shown that the argon ion‐beam flux peaked up to 1.928 × 10³⁰ ions m^?2 s^?1 at the optimum pressure of 1.866 mbar while the neon‐argon mixture's ion‐beam flux reached a maximum of 4.301 × 10³⁰ ions m^?2 s^?1 for 15% neon admixture at the optimum pressure of 1.866 mbar. The calculated kinetic energy of the ion beam has shown a maximum value of 708.7 J for the mixture of 85% argon‐15% neon at the mentioned optimum pressure.     

FLUIDIZATION OF CARBON NANOTUBES

Carbon nanotubes (CNTs) can be fluidized in the form of fluidlike agglomerates made of many three-dimensional sub-agglomerates, having a multi-stage agglomerate (MSA) structure and containing large amounts of twisting CNTs of micrometer magnitude.     

谐波抑制方法研究

由于电力电子技术在电力系统中的大量应用 ,使得电力系统中的电流电压波形畸变越来越严重 .该文简介了电力系统谐波的来源、危害和检测 ,阐述了一种利用有源滤波器进行谐波抑制的方法及相应的实验电路与实验结果     

单相UPQC控制系统的数学建模和仿真

针对目前广泛研究的有源电力滤波器存在不能同时补偿市电谐波电流和负载侧谐波电压的问题,设计一种单相UPQC,分析其工作原理,建立该装置的控制系统的数学模型,并进行了仿真实验研究.仿真结果表明:采用这种单相UPQC,不仅可以抑制非线性负载所产生的谐波,而且可有效提高负载侧的供电质量.图8,表1,参8.     

能降低端电压和容量的有源电力滤波器并联混合补偿拓扑结构研究

提出了一种能够同时降低端电压和容量的有源电力滤波器（ＡＰＦ）并联混合补偿拓扑结构．理论分析和仿真结果表明,ＡＰＦ并联混合补偿拓扑结构是合理可行的,能大大降低ＡＰＦ的端电压和容量,使得ＡＰＦ技术在电力系统中和铁道电气化供电系统中应用成为可能．     

有源电力滤波器的自适应滞环电流跟踪策略

针对有源电力滤波器(APF)传统滞环电流跟踪方式逆变器的开关频率高且频率范围大的问题,提出了一种可变环宽的滞环电流跟踪控制策略.该策略根据系统开关频率、直流侧电压、主电路电感等参数动态的调节滞环比较器的环宽,不仅能够有效的降低开关频率和频率变化范围,而且在跟踪精度和动态性能上也比传统滞环比较法优越.通过MATLAB仿真表明,该策略的电流跟踪精度满足有源电力滤波器的要求.     

基于DSP控制的电力系统有源滤波器硬件设计研究

探讨了DSP控制的有源电力滤波器硬件设计,以DSP芯片TMS320LF2407为主要核心控制芯片,配合采样周期信号发生电路以及A/D、D/A等辅助外围电路设计了有源电力滤波器的控制系统;给出了主逆变电路的实现方案;对设计出的样机进行了实验,实验结果表明:补偿电流几乎实时跟踪谐波电流,系统的补偿速度和精度大大提高,体现了数字控制系统的优势,同时,随着采样周期的减小,补偿效果会更加明显.本文提出的以TMS320LF2407为核心的有源电力滤波器在技术上是可行的.     

基于ELM的单相有源电力滤波器的内模控制

为了克服单相有源电力滤波器的非线性特性,在基于数据的基础上,由极限学习机建立有源滤波器非线性系统的内模和逆模,实现基于极限学习机的内模控制.根据内模控制系统稳态误差,评估内模控制系统的性能,并将极限学习机的内模控制系统与神经网络、核岭回归、支持向量机等方法进行比较分析.试验仿真表明,基于极限学习机的内模控制系统具有系统稳态误差小、鲁棒性强等特点.     

基于PSO和人工势场的机器人路径规划

文章提出了一种变形Gaussian函数作为势场模型,它能更准确地反映势场环境;通过分析震荡现象产生的原因,以及局部极小值点的特点,将粒子群算法引入到路径规划过程中,用于绕过障碍物或逃逸局部极小值;仿真结果表明,该方法能有效消除运动路径的震荡现象,极大地降低了陷入局部极小值的概率。