PSO-RBFNN模型及其在岩土工程非线性时间序列预测中的应用

岩土工程受力变形演化是一个典型的非线性问题,其演化的高度非线性和复杂性,很难用简单的力学、数学模型描述,但可用粒子群优化径向基神经网络对岩土工程应力、位移非线性时间序列进行动态实时预测。网络径向基层的单元数通过均值聚类法确定后,所有其它参数：中心位置、形状参数、网络权值,均通过粒子群优化算法在全局空间优化确定。工程实例应用表明,该模型预测结果准确、精度高,有良好的应用前景。     

基于Mindlin理论Winkler地基参数的Kalman滤波识别

提出了Winkler地基参数的Kalman滤波识别方法。引入Mindlin理论后,推导了Winkler地基上板的控制微分方程。利用Fourier变换,推求了Winkler地基上简支板的Fourier闭式解。推导了Winkler地基参数的Kalman滤波方程,研究了Winkler地基参数的Kalman滤波识别的具体计算步骤。研究表明,运用Kalman滤波理论进行Winkler地基参数的识别,能有效地估计Winkler地基参数;Winkler地基参数的滤波收敛速度和精度受地基参数初始值选取和位移实测资料的影响,且Kalman滤波理论也可用于其他地基模型地基参数的识别。     

不同年度收存的褶皱臂尾轮虫休眠卵的孵化及滞育胚胎“激活”效果的研究

比较了 1 985— 1 998年间由同一褶皱臂尾轮虫 (Brachionusplicatilis)种群在实验室内历年产生并收存的休眠卵的孵化效果 ,并初步探讨了胚胎“激活”处理对孵化的影响。结果表明 ,休眠卵的收存年度或时间对其在一定孵育期内的总孵化率 (THR)和日孵化率 (DHR)有极显著影响 (x2 检验 ,p<0 .0 0 1 ) ,但THR和DHR与储存时间间均无显著的相关性 ,而是呈极大波动。同时发现“激活”作用具有促使孵化高峰提前到来、增强孵化同步性的双重功效 ,从而可明显缩短孵育时间 ,提高孵化速率。但“激活”前后多数卵样的总孵化率无显著变化 (p >0 .0 5 )。详细讨论了可能导致孵化率波动的多种因素 ,例如遗传变异、培养历史、储存及孵化的环境条件等。     

基于雷达海冰图像互相关的冰漂流场测量

我国的渤海和黄海北部在每年冬季都出现不同程度的冰情,它直接影响结冰海区的石油平台、船舶以及港口等设施的正常作业,对海冰的观测与预报随着上述海区的开发利用越发显得重要,利用航海雷达连续准确地跟踪海冰漂移运动轨迹,是当今观测、调查、研究冰漂流移动规律的有效方法之一,我国的科研人员利用雷达成像技术对渤海的冰情进行了长期的雷达海冰观测、研究,在由雷达海冰图像对海冰的物理特征的识别、分类以及冰漂流场测量方面取得了一些进展[1-4]。     

Neural Network-Based Active Control for Offshore Platforms

1 .IntroductionWiththedevelopmentofoceantechnology ,moreandmoreextremelylargeandlongflexibleoff shoreplatformsusedforoilexplorationanddrillingoperationarebuiltinhostileoceanenvironments .Ingeneral,thiskindofplatformsisanonlineardistributedparametersystemanditsnaturalfrequencyfallsclosertothedominantwavefrequencieswiththeincreaseofwaterdepth .Besides ,itsstructureisverycomplexandtheexternalwaveforceontheplatformisuncertain .Thus ,theseplatformsarepronetoexcessivewave inducedoscillationsunderbot…     

Influence of Wind on Dynamics and Flushing of Shallow Estuaries

Observations of two small estuaries in Cape Cod, U.S.A. indicate large variations in salinity structure that are forced by variations in along-estuary wind stress. During onshore winds, the estuarine circulation is reduced, and the along-estuary salinity gradient increases as freshwater accumulates. During offshore winds, the surface outflow is enhanced, freshwater is flushed out of the estuary, and the along-estuary salinity gradient becomes weak. Constrictions block the wind-induced flushing, resulting in strong salinity fronts across the constrictions. The residence time of one of the estuaries varies by more than a factor of three in response to variations in wind-induced flushing. The other estuary has little variation of flushing associated with winds, due to a constriction at the mouth that inhibits the wind-induced exchange. The strong influence of winds on the flushing of these estuaries is due in part to their shallow depths, which accentuates the influence of wind stress relative to the effects of the horizontal density gradient. In addition, the residence times of the estuaries are comparable to the time scale of wind forcing, allowing large changes in water properties during wind events.     

季节内SST和混合层深度变化的模式研究

用Niiler—Kraus类型的混合层积分模式,对TOGA—COARE强化观测期间由《实验3号》科学考察船观测资料得到的混合层深度和SST在季节内时间尺度的变化进行了模式研究。指出:1.混合层耗散参数与较长时间尺度过程风应力的变化存在着比较好的对应关系;2.模式可以较好的对风场和热通量场在季节内时间尺度的变化作出响应,模拟出季节内时间尺度SST的变化;3.Niiler,-Kraus模式在考虑耗散作用后,可用于海洋季节内时间尺度变化的模式研究。     

Estuary/ocean exchange and tidal mixing in a Gulf of Maine Estuary: A Lagrangian modeling study

A Lagrangian particle method embedded within a 2-D finite element code, is used to study the transport and ocean–estuary exchange processes in the well-mixed Great Bay Estuarine System in New Hampshire, USA. The 2-D finite element model, driven by residual, semi-diurnal and diurnal tidal constituents, includes the effects of wetting and drying of estuarine mud flats through the use of a porous medium transport module. The particle method includes tidal advection, plus a random walk model in the horizontal that simulates sub-grid scale turbulent transport processes. Our approach involves instantaneous, massive [O(500,000)] particle releases that enable the quantification of ocean–estuary and inter-bay exchanges in a Markovian framework. The effects of the release time, spring–neap cycle, riverine discharge and diffusion strength on the intra-estuary and estuary–ocean exchange are also investigated.The results show a rather dynamic interaction between the ocean and the estuary with a fraction of the exiting particles being caught up in the Gulf of Maine Coastal Current and swept away. Three somewhat different estimates of estuarine residence time are calculated to provide complementary views of estuary flushing. Maps of residence time versus release location uncover a strong spatial dependency of residence time within the estuary that has very important ramifications for local water quality. Simulations with and without the turbulent random walk show that the combined effect of advective shear and turbulent diffusion is very effective at spreading particles throughout the estuary relatively quickly, even at low (1 m²/s) diffusivity. The results presented here show that a first-order Markov Chain approach has applicability and a high potential for improving our understanding of the mixing processes in estuaries.     

珠江三角洲网河区水位变化趋势研究

根据珠江三角洲网河区29个验潮站的实测验潮记录,应用傅氏变换与最平滤波器串联的方法来消除月均序列的周期波动对确定水位变化趋势的影响,由低通序列一元线性回归分析确定各站水位的变化趋势;结果表明,周期波动对确定海平面变化趋势的影响是显著的.应用经验正交函数(EOF)对网河区的水位变化场进行分解,由相互独立的时间函数和空间特征函数表征网河区区域的水位变化特征;应用时间特征函数计算区域水位的平均变化率为0.02mm/a.根据验潮站的水位变化趋势,探讨网河区水位变化与河床冲淤的关系.