A NEW APPROACH TO IMPROVE THE NUMERICAL WEATHER PREDICTION

For certain particular initial values, the results of numerical weather prediction are extremely sensitive to the errors in some parameters of the models, which probably is one of the reasons to bring about serious failures to the forecast. In this paper, this speculation is shown by an example of numerical experiments, and it is suggested that the parameters in the models be modified with the information provided by observational data of the recent atmospheric evolution to avoid failures of the forecast. The numerical simulation tests with a simple quasi-geostrophic barotropic model have obtained satisfying results.     

Interactions of peptide mimics of hyaluronic acid with the receptor for hyaluronan mediated motility (RHAMM)

Summary Using the hyaluronic acid (HA) binding region of the receptor for hyaluronan-mediated motility (RHAMM) as a model, a molecular perspective for peptide mimicry of the natural ligand was established by comparing the interaction sites of HA and unnatural peptide–ligands to RHAMM. This was accomplished by obtaining a series of octapeptide–ligands through screening experiments that bound to the HA binding domains of RHAMM (amino acids 517–576) and could be displaced by HA. These molecules were computationally docked onto a three-dimensional NMR based model of RHAMM. The NMR model showed that RHAMM(517–576) was a set of three helices, two of which contained the HA binding domains (HABDs) flanking a central groove. The structure was stabilized by hydrophobic interactions from four pairs of Val and Ile side chains extending into the groove. The presence of solvent exposed, positively charged side chains spaced 11Å apart matched the spacing of negative charges on HA. Docking experiments using flexible natural and artificial ligands demonstrated that HA and peptide–mimetics preferentially bound to the second helix that contains HABD-2. Three salt bridges between HA carboxylates and Lys548, Lys553 and Lys560 and two hydrophobic interactions involving Val538 and Val559 were predicted to stabilize the RHAMM-HA complex. The high affinity peptides and HA utilized the same charged residues, with additional contacts to other basic residues. However, hydrophobic contacts do not contribute to affinity for peptide ligand-RHAMM complexes. These results offer insight into how selectivity is achieved in the binding of HA to RHAMM, and how peptide competitors may compete for binding with HA on a single hyaladherin.     

基于禁忌算法优化神经网络的海洋船舶流量预测

为了有效缓解船舶交通拥堵和提高通航效率,对海洋港口和航道管理提供一个更可靠的数据,设计了一种基于禁忌算法优化神经网络的海洋船舶流量预测的方法。首先,建立了基于禁忌算法优化神经网络的海洋船舶流量预测模型。然后,设计了三层的脊波神经网络结构,提出了采用禁忌算法优化脊波神经网络结构参数的具体方法,从而得到一个初始化的脊波神经网络流量预测模型。然后,采用有标签的训练样本数据集对网络进行训练,将满足误差阈值的训练模型作为最终的预测模型。以某港口为例进行仿真实验,结果表明文中得到的预测结果与真实值较为接近,且与其它方法相比,具有更好的预测效果。     

Towards a model for electric vehicle noise emission in the European prediction method CNOSSOS-EU

The CNOSSOS-EU method is recommended in Europe for environmental noise prediction. In regards to road traffic, it includes vehicle noise emission models implicitly referring to internal combustion vehicles. The development of electrically driven vehicles calls for the future consideration of these vehicles in prediction models. On the basis of experimental data, the study reported in this paper proposes a noise emission model for extending CNOSSOS-EU to light electric vehicles. Correction terms to be applied to the propulsion noise component are determined. Investigations on a sample of tyres with good rolling resistance performance, which is a main tyre selection criterion on these vehicles, indicated that no correction is required for the rolling noise component. Differences between the noise emission from conventional vehicles and electric vehicles are discussed for several road surfaces. Owing to the limited vehicle sample as well as transitional statements, this new model for electric vehicles running at constant speed over 20 km/h should be considered as a first step towards the definition of this vehicle technology in CNOSSOS-EU.     

基于遗传神经网络的旋转机械故障预测方法研究

许多大型旋转机械运行工况恶劣,非平稳、非线性特征明显,以及各种突发性、偶然性因素的影响,给基于振动信号处理的状态预测和状态维护分析带来困难。神经网络以其强大的处理非线性系统的能力在故障预测中得到广泛的应用,但由于其在追求高精度训练目标时易陷入局部极值,且收敛速度慢甚至发散。针对这个问题,提出了采用遗传算法对神经网络连接权值和阈值进行优化,这样不仅发挥了神经网络广泛的映射特性也使遗传算法的全局搜索优势尽显无疑。通过组合这两种算法,在提升网络学习的准确度方面,优点尤其突出,最终提高对旋转机械故障预测和寿命估计的性能,这在某环境模拟试验系统动力风机的轴承磨损故障预测中得到了验证。     

含有误差校正的小波神经网络交通流量预测

交通流量的准确预测对于高速路管理者进行决策至关重要。建立了小波神经网络(WNN)交通流量预测模型,并通过预测训练误差和测试误差校正预测结果来提高预测精度。首先构建WNN模型对交通流量进行初步预测,然后利用经验模态分解(EMD)和WNN模型对训练误差和测试误差进行预测。分别用训练误差预测值、测试误差预测值和两种误差预测值的加权对流量初步预测结果进行修正得到最终预测值。采用四川省成灌高速路交通流量数据进行了仿真对比实验,仿真结果表明含有误差校正的小波神经网络模型能有效提高交通流量预测精度,并且利用两种误差加权修正模型的预测精度高于利用测试误差的修正模型和利用训练误差的修正模型。     

Evaluation of an amide‐based stationary phase for supercritical fluid chromatography

A relatively new stationary phase containing a polar group embedded in a hydrophobic backbone (i.e., ACE ® C18‐amide) was evaluated for use in supercritical fluid chromatography. The amide‐based column was compared with columns packed with bare silica, C₁₈ silica, and a terminal‐amide silica phase. The system was held at supercritical pressure and temperature with a mobile phase composition of CO₂ and methanol as cosolvent. The linear solvation energy relationship model was used to evaluate the behavior of these stationary phases, relating the retention factor of selected probes to specific chromatographic interactions. A five‐component test mixture, consisting of a group of drug‐like molecules was separated isocratically. The results show that the C₁₈‐amide stationary phase provided a combination of interactions contributing to the retention of the probe compounds. The hydrophobic interactions are favorable; however, the electron donating ability of the embedded amide group shows a large positive interaction. Under the chromatographic conditions used, the C₁₈‐amide column was able to provide baseline resolution of all the drug‐like probe compounds in a text mixture, while the other columns tested did not.     

乳腺癌术前灰阶超声与彩色多普勒超声联合应用对预后的预测价值

目的探究乳腺癌术前灰阶超声与彩色多普勒超声联合应用对预后的预测价值。方法选取温州医科大学附属第一医院2013年5月—2015年5月期间收治的乳腺癌患者83例,治疗前均行灰阶超声与彩色多普勒超声检查,分析患者超声检查结果、临床征象与术后随访2年期间患者预后的关系,从而分析术前灰阶超声与彩色多普勒超声联合应用对乳腺癌患者预后的预测价值。结果乳腺癌肿块边界清晰、边界不清晰、边界恶性晕的患者2年无瘤生存率依次降低,边界恶性晕与边界不清晰患者死亡率均显著高于边界清晰患者(P0.05);肿瘤≥2 cm患者2年无瘤生存率显著低于2 cm患者;淋巴结转移患者2年无瘤生存率显著低于无转移患者(P0.05);血流分级Ⅱ级患者2年无瘤生存率明显低于0级(P0.05),Ⅲ级患者2年无瘤生存率显著低于0级和Ⅰ级(P0.05)。结论乳腺癌患者术前联合应用灰阶超声与彩色多普勒超声检查可确定肿瘤边界、大小及淋巴转移等征象,其与患者无瘤生存率、死亡率等预后指标密切相关,可作为预测乳腺癌患者预后的重要方法。