利用非光敏BCB树脂实现多芯片组件平坦化研究

多层布线和绝缘介质材料的引入使得多芯片组件的芯片表面形貌凸凹不平,采用旋涂介质膜实现芯片表面平坦是常用的平坦化方法.分别介绍了聚酰亚胺、旋涂玻璃膜和苯并环丁烯(BCB)的平坦化特性,论述了非光敏BCB树脂的平坦化工艺原理.在此原理的基础上实现了非致冷红外焦平面的读出电路芯片平坦化,表面台阶从1.39 μm下降到0.097 μm,平坦度达到93%.     

Cascades with coupled map lattices in preferential attachment community networks

In this paper, cascading failure is studied by coupled map lattice （CML） methods in preferential attachment community networks. It is found that external perturbation R is increasing with modularity Q growing by simulation. In particular, the large modularity Q can hold off the cascading failure dynamic process in community networks. Furthermore, different attack strategies also greatly affect the cascading failure dynamic process. It is particularly significant to control cascading failure process in real community networks.     

Topological Effects on the Performance of Transportation Networks

We investigate the influence of the network topology on the performance （characterized by the total system cost and maximal tratffic volume） of transportation networks, where the weights are not static （constant）, but dynamic （a function of the flow on the link）. Four classes of networks are used in the simulation, including regular networks, random networks, small-world networks and scale-free networks. The initial simulation results show that topologies play important roles on the performance of transportation networks, and random networks have better performance than other networks. Also, we find that there are distinct difference of the link flow distribution for various networks in both the distribution function form and the span between the minimum and the maximum of the link flow, explaining the difference of the performance among distinct networks. These findings will be useful in network design problems of transportation systems.     

考虑交通出行惯例的双向行人流模型研究

推广了Baek等人最近提出的一个双向行人流模型,提出了两种改进策略,并从行人平均速度-密度关系、行人空间分布密度和位置分布等方面进行了数值分析. 研究发现,引入的两个新策略不仅可以提高行人流的平均速度,而且可以提高道路系统(尤其是道路中央区域)利用率,减轻拥堵状况,有效避免严重堵塞的发生. 改进的策略对行人的心理特点和行为特性等方面考虑更加全面,而且可以较好地模拟高密度的双向行人流. 关键词： 元胞自动机模型 双向行人流 交通惯例     

Efficiency promotion for an on-ramp system based on intelligent transportation system information

<正>The effect of cars with intelligent transportation systems(ITSs) on traffic flow near an on-ramp is investigated by car-following simulations.By numerical simulations,the dependences of flux on the inflow rate are investigated for various proportions of cars with ITSs.The phase diagrams as well as the spatiotemporal diagrams are presented to show different traffic flow states on the main road and the on-ramp.The results show that the saturated flux on the main road increases and the free flow region is enlarged with the increase of the proportion of cars with ITS.Interestingly,the congested regions of the main road disappear completely when the proportion is larger than a critical value.Further investigation shows that the capacity of the on-ramp system can be promoted by 13%by using the ITS information, and the saturated flux on the on-ramp can be kept at an appropriate value by adjusting the proportion of cars with ITS.     

反馈控制双车道跟驰模型研究

通过引入合理的换道规则，将单车道反馈控制模型扩展成双车道交通流模型.数值仿真表明，与经典的优化速度双车道模型相比，由于引入适当的反馈控制信息，车辆速度波动变小或消失，系统不会出现严重的堵塞现象.这说明该双车道反馈控制模型能有效地抑制或缓解交通流堵塞. 关键词： 跟驰模型 双车道模型 反馈控制 堵塞抑制     

Combined bottleneck effect of on-ramp and bus stop in a cellular automaton model

The combined bottleneck effect is investigated by modeling traffic systems with an on-ramp and a nearby bus stop in a two-lane cellular automaton model. Two cases, i.e. the bus stop locates in the downstream section of the on-ramp and the bus stop locates in the upstream section of the on-ramp, are considered separately. The upstream flux and downstream flux of the main road, as well as the on-ramp flux are analysed in detail, with respect to the entering probabilities and the distance between the on-ramp and the bus stop. It is found that the combination of the two bottlenecks causes the capacity to drop off, because the vehicles entering the main road from the on-ramp would interweave with the stopping (pulling-out) buses in the downstream (upstream) case. The traffic conflict in the former case is much heavier than that in the latter, causing the downstream main road to be utilized inefficiently. This suggests that the bus stop should be set in the upstream section of the on-ramp to enhance the capacity. The fluxes both on the main road and on the on-ramp vary with the distance between the two bottlenecks in both cases. However, the effects of distance disappear gradually at large distances. These findings might give some guidance to traffic optimization and management.     

Multiple flux difference effect in the lattice hydrodynamic model

Considering the effect of multiple flux difference, an extended lattice model is proposed to improve the stability of traffic flow. The stability condition of the new model is obtained by using linear stability theory. The theoretical analysis result shows that considering the flux difference effect ahead can stabilize traffic flow. The nonlinear analysis is also conducted by using a reductive perturbation method. The modified KdV (mKdV) equation near the critical point is derived and the kink-antikink solution is obtained from the mKdV equation. Numerical simulation results show that the multiple flux difference effect can suppress the traffic jam considerably, which is in line with the analytical result.     

改进的GMDH型神经网络及其在混沌预测中的应用

本文提出基于改进自组织方法的GMDH（Group Method of Data Handling）型神经网络并将它应用于混沌顶测。一般的GMDH型神经网络的自组织功能是通过给定一个准则阈值来确定或直接给定数值来实现，但GMDH型神经网络的自组织准则的阈值难以合适确定，由此提出了一种简单的自组织方法来实现真正意义上的自组织功能。这种用改进了自组织方法所构成的GMDH型神经网络可以应用于混沌时间序列预测。通过仿真实验，证明其预测效果明显比基本的GMDH型神经网络好，即改进GMDH型神经网络优于基本的GMDH型神经网络。