Near-infrared spectroscopy-based quantification of substrate and aqueous products in wastewater anaerobic fermentation processes

In this study, a Fourier transform near-infrared （FT-NIR） technique was used for process monitoring of anaerobic fermentation. Orthogonal signal correction （OSC） method was used as NIR spectral pre- processing options. Calibration models were established and the validation of the method was per- formed with the sucrose, ethanol and volatile fatty acids （VFAs） contents determined by the anthrone and gas chromatography methods, respectively. Spectral range and the number of internal latent variables were optimized for the best correlation coefficient. Test set validation of sucrose resulted in excellent measurement of prediction performance and the correlation coefficient of determination is 0.930. Similar prediction statistics for individual VFA and total VFA contents were obtained. These re- sults proved that the NIR spectroscopy technology is able to quantify the contents of both volatile fatty acids and sucrose in wastewater anaerobic fermentation process.     

Study on the FEM design of reinforced earth retaining wall with geogrid

At present, limit equilibrium method is often adopted in the design of reinforced earth retaining wall. Geotechnical engineers at home and abroad have done a lot of work to improve the traditional calculation methods in recent years, while there are lots of defects. This paper first identifies the location of failure surface and safety factor through the finite element program of PLAXIS and then analyses the influencing factors of the stability of reinforced earth retaining wall with geogrid. The authors adopt strength reduction FEM in the design and stability analysis of reinforced earth retaining wall and have achieved some satisfying results. Without any assumptions, the new design method can automatically judge the failure mode of reinforced earth retaining wall; can consider the influence of axial tensile stiffness of the reinforcement stripe on the stability of retaining wall; can identify reasonable distance and length of the reinforcement stripe; can choose suitable parameters of reinforcement stripe, including strength, stiffness and pseudo-friction coefficient which makes the design optimal. It is proved through the calculation examples that this method is more reasonable, reliable and economical in the design of reinforced earth retaining wall.     

网状被动传感器系统优化设计

结合网状被动传感器系统建设的需要,提出了网状被动传感器系统优化设计模型,依据优化设计模型提出了基于误差方差Cramer-Rao下界的优化设计算法。探讨了优化过程中指标集、目标集的确定方法,最后通过仿真计算分析了三种配置结构的性能。     

基于并行混沌和单纯形法的混合全局优化算法

混沌优化算法采用的是串行优化结构,采用并行结构进行,并不断缩小搜索空间,提高了混沌优化在变量取值范围较大情况下的搜索效率。针对混沌在全局最优点附近搜索速度变得很慢、精度较低的缺点,结合单纯形法,提高了收敛的速度和求解精度。仿真结果表明并行混合优化算法可以得到满意的结果。     

一种多粒子群切换决策机制

提出了一种切换决策机制,支持总最佳连接ABC(always best connected).该机制综合考虑接入网络状况、应用需求、用户对接入网络编码制式偏好、用户对接入网络供应商偏好、终端当前运动速率和终端当前剩余电量等因素,通过博弈分析,基于含最优变异的多粒子群优化算法,寻找把N个终端分配到M个接入网络的最佳切换决策方案,使用户和网络提供方效用达到或接近Nash均衡下的Pareto最优.仿真结果表明,该机制是可行和有效的.     

内存数据库查询优化

基于一种广泛接受的内存数据库系统基本概念及内存数据库中的SB-树索引结构，提出了更方便关系代数操作的SB^*-树．在此基础上，提出了优化的关系代数操作的实现算法．根据这些算法和关系代数等价定理，给出了对关系代数查询树进行逻辑优化的规则，并证明了逻辑优化的正确性．最后给出了代价评估模型，实验表明提出的内存数据库查询优化算法相对于传统的优化算法具有更好的时间及空间性能．     

带有全局交叉的修正和声搜索算法

为增强和声搜索算法的全局搜索能力,提出一种带有全局交叉的修正和声搜索(MHSgc)算法.MHSgc算法采用多和声记忆库协同创作,应用邻域学习策略进行调整,取代原有的基音调整,从而增加了种群多样性.同时,提出一种全局交叉操作,并融合到MHSgc算法中,防止算法陷入局部最小.针对几个标准函数进行了实验仿真,数值结果表明,上述算法优于文献报道的8种智能算法(HS,IHS,GHS,NGHS,EHS,ITHS,MPSO,RMDE),具有较好的优化潜力.     

一种改进的和声搜索算法

针对和声搜索算法在处理复杂函数优化问题时容易陷入局部最优、收敛精度低的缺点,提出了一种改进的和声搜索算法,不同于已有的HS算法.整个和声记忆库被划分为一些小的子和声记忆库,每个子库适时地更新内部信息,然后将各子库中的最优解构成一个较优记忆库并进行搜索,这些子记忆库通过重组周期被反复重组,信息在这些子库中被交换,在算法的最后搜索阶段,为了表现一个更好的局部搜索能力,所有和声形成一个和声记忆库.同目前提出的一些HS算法相比,新算法有更好的优化性能.     

基于稀疏栅格优化的蜂窝车联网定位算法

蜂窝车联网(cellular-V2X,C-V2X)中的定位方案是车路协同与车联网业务发展的重要技术途径之一.目前基于基站、卫星等诸多定位方案,在车联网业务以及车路协同场景中常会遇到定位精度、定位处理时延、部署成本等诸多方面的挑战.针对这些问题,文章对已有栅格定位算法进行优化,提出一种基于统计信息网格(statistic...     

双向认知网络物理层安全策略建模及仿真研究

为提高移动通信系统的性能,构建了一个存在主网络干扰和窃听者时双向认知中继网络时物理层安全传输模型,设计了一种中继选择和功率分配的联合优化策略,用于对抗窃听者,保护收发节点的信息传输.通过仿真分析得出,此非凸优化问题限制条件较多,难以保证粒子群优化算法随机生成的初始解的可行性,从而造成求解困难.提出了一种基于可变网格优化...