遗传算法在边坡数值计算中的应用

改进了进化方向的遗传算法与有限元数值法，结合并研制了相应的软件。应用该软件对多类型岩土边坡进行弹性模量、内聚力、内摩擦角等参数反演分析，显示误差很小，收敛速度也很快，这说明改进进化方向遗传算法这种新型的优化算法在多类型岩土参数优化估计中具有独特的优势。     

新疆伽师及邻近地区现今地壳形变的GPS监测与研究

利用高精度数据软件对伽师及邻近地区的GPS监测网观测数据进行计算分析(基线精度达到了 10 - 8～ 10 - 9) ,获得了该地区的现今地壳形变速率图 .结果表明 ,伽师及邻近地区在印度板块的推挤作用下 ,地壳正快速缩短变形 ,主压应力为近南北方向 .各地块运动速率不同 ,平均应变率为 0 .0 4× 10 - 6 /a .     

Application of genetic BP network to discriminating earthquakes and explosions

Introduction Artificial Neural Network (ANN) is an important branch of artificial intelligence. It is proposed on the foundation of the study on modern neural science, is a man-made network that can implement some functions based on the mans comprehensive understanding for cerebral neural network (HAN, WANG, 1997). ANN is a mathematical model of simplified human brain neural network and is used to simulate the structures and functions of human brain neural network. ANN is a complex netw…     

Fuzzy neural network models for liquefaction prediction

Integrated fuzzy neural network models are developed for the assessment of liquefaction potential of a site. The models are trained with large databases of liquefaction case histories. A two-stage training algorithm is used to develop a fuzzy neural network model. In the preliminary training stage, the training case histories are used to determine initial network parameters. In the final training stage, the training case histories are processed one by one to develop membership functions for the network parameters. During the testing phase, input variables are described in linguistic terms such as ‘high’ and ‘low’. The prediction is made in terms of a liquefaction index representing the degree of liquefaction described in fuzzy terms such as ‘highly likely’, ‘likely’, or ‘unlikely’. The results from the model are compared with actual field observations and misclassified cases are identified. The models are found to have good predictive ability and are expected to be very useful for a preliminary evaluation of liquefaction potential of a site for which the input parameters are not well defined.     

ELF propagation parameters for uniform models of the Earth–ionosphere waveguide

Uniform models for the Earth–ionosphere cavity are considered with particular attention to the physical properties of the ionosphere for the extremely low frequency (ELF) range. Two consistent features have long been recognized for the range: the presence of two distinct altitude layers of maximum energy dissipation within the lower ionosphere, and a “knee”-like change in the vertical conductivity profile representing a transition in dominance from ion-dominated to electron-dominated conductivity. A simplified two-exponential version of the Greifinger and Greifinger (1978) technique widely used in ELF work identifies two slopes in the conductivity profile and, providing accurate results in the ELF communication band (45–75 Hz), simulates too flat a frequency dependence of the quality factor within the Schumann resonance frequency range (5–40 Hz). The problem is traced to the upward migration, with frequency increasing, of the lower dissipation layer through the “knee” region resulting in a pronounced decrease of the effective scale height for conductivity. To overcome this shortcoming of the two-exponential approximation and still retain valuable model analyticity, a more general approach (but still based on the Greifinger and Greifinger formalism) is presented in the form of a “knee” model whose predictions for the modal frequencies, the wave phase velocities and the quality factors reasonably represent observations in the Schumann resonance frequency range.     

ARTIFICIAL NEURAL NETWORK CORRECTION OF ROUGH ERRORS OF OBSERVATIONS*

In the context of tower measured radiation datasets.following the correction principle meeting a diagnostic equation in data quality control and in terms of a technique for model construction on data and ANN (artificial neural network) retrieval for BP correction of radiation measurements with rough errors available,a BP model is presented.Evidence suggests that the developed model works well and is superior to a convenient multivariate linear regression model,indicating its wide applications.     

Surface vibration due to a sequence of high speed moving harmonic rectangular loads

The transmission of vibrations over the surface of the ground, due to high-speed moving, vertical harmonic rectangular loads, is investigated theoretically. The problem is three-dimensional and the interior of the ground is modelled as an elastic half-space or a multilayered ground. The transformed solutions are obtained using the Fourier transform on the space variable. A new damping model in the spatial wavenumber domain, presented in Lefeuve-Mesgouez et al. [J. Sound. Vibr. 231 (2000) 1289] is used. Numerical results for the displacements on the surface are presented for loads moving with speeds up to and beyond the Rayleigh wave speed of the half-space.     

泾阳4.8级地震的构造背景

1998年1月5日烃阳Ms 4.8地震是陕西省近40年来发生的最大一次地震。研究了此次地震的构构背景和地球物理场特征。     

青藏高原大气—植被相互作用的模拟试验Ⅱ.植被叶面积指数和净初级生产力

利用大气－植被相互作用模式AVIM对青藏高原上30个站点进行模拟计算，给出了青藏高原上植被叶面积指数和净初级生产力的分布特征。模拟结果表明，高原上叶面积指数、净初级生产力由东南向西北减少，模拟和观测的分布相一致。分析指出，青藏高原净初级生产力的分布决定于水热条件的共同作用，降水充沛、温度适宜的青藏高原东南部净初级生产力较高，而青藏高原西北部由于降水过少或温度过低，初级生产力很低。     

基于神经网络的单元自动机CA及真实和忧化的城市模拟

提出了一种基于神经网络的单元自动机（ＣＡ）。ＣＡ已被越来越多地应用在城市及其它地理现象的模拟中。ＣＡ模拟所碰到的最大问题是如何确定模型的结构和参数。模拟真实的城市涉及到使用许多空间变量和参数。当模型较复杂时，很难确定模型的参数值。本模型的结构较简单，模型的参数能通过对神经网络的训练来自动获取。分析表明，所提出的方法能获得更高的模拟精度，并能大大缩短寻找参数所需要的时间。通过筛选训练数据，本模型还可以进行优化的城市模拟，为城市规划提供参考依据。