Application of artificial intelligence techniques in automatic hull form generation

The paper describes a methodology for automatic hull form generation of ships with some desired performances using artificial intelligence techniques. The whole implementation process is divided into three main components. First of all the half-breadth weight matrices are generated that would provide population with pre-fitness by examining the relationship of the hull form with the principal dimensions of a number of existing vessels. Secondly, breadth and draft are adjusted using Neural Network Concepts. Breadth, draft, length, displacement and speed of the ship are very related terms and relations among them create some constraints. Neural Networks solve these constraints and adjust the parameters. Finally, Genetic Algorithm is used for searching the exact solution by examining several generations. For this, the algorithms need to measure fitness for every population in every generation. Unfortunately, GA doesn't guarantee fairness of the surface of the hull form, which can't be ignored. So, for every population especially for newborn population fairing techniques would be used. However, being complex geometric shape, hull form surface can't be faired using least-square or other single equation fitting techniques. Fairness is done taking pair-wise points using B-spline functions [CAD 20(1988)].     

Temporal significant wave height estimation from wind speed by perceptron Kalman filtering

Significant wave height estimates are necessary for many applications in coastal and offshore engineering and therefore various estimation models are proposed in the literature for this purpose. Unfortunately, most of these models provide simultaneous wave height estimations from wind speed measurements. However, in practical studies, the prediction of significant wave height is necessary from previous time interval measurements. This paper presents a dynamic significant wave height prediction procedure based on the perceptron Kalman filtering concepts. Past measurements of significant wave height and wind speed variables are used for training the adaptive model and it is then employed to predict the significant wave height amounts for future time intervals from the wind speed measurements only. The verification of the proposed model is achieved through the dynamic significant wave height and wind speed time series plots, observed versus predicted values scatter diagram and the classical linear significant wave height models. The application of the proposed model is presented for a station in USA.     

Sub-pixel classification of SPOT-VEGETATION time series for the assessment of regional crop areas in Belgium

Global time series of low resolution images are available with high repeat frequency and at low cost, but their analysis is hampered by the presence of mixed pixels and the difficulty in locating detailed spatial features. This study examined the potential of sub-pixel classification for regional crop area estimation using time series of monthly NDVI-composites of the 1 km resolution sensor SPOT-VEGETATION. Belgium was selected as test zone, because of the availability of ample reference data in the form of a vectorial GIS with the boundaries and cover type of the large majority of agricultural fields. Two different methods were investigated: the linear mixture model and neural networks. Both result in area fraction images (AFIs), which contain for each 1 km pixel the estimated area proportions occupied by the different cover types (crops or other land use). Both algorithms were trained with part of the reference data and validated with the remainder. Validation was repeated at three different levels: the 1 km pixel, the municipality and the agro-statistical district. In general, the neural network outperformed the linear mixture model. For the major classes (winter wheat, maize, forest) the obtained acreage estimates showed good agreement with the true values, especially when aggregated to the level of the municipality (R² ≈ 85%) or district (R² ≈ 95%). The method seems attractive for wide-scale, regional area estimation in data-poor countries.     

基于神经网络的似大地水准面模型精化方法研究

似大地水准面的精化与高程异常的求解是同一个概念,为提高似大地水准面精化的精度,采用格网＋神经网络方法,借助神经网络BP方法,建立高程异常与坐标之间的函数关系;用训练好的神经网络模型和固定间隔的格网,建立格网模型;在已建好的格网模型中,内插出给定点的高程异常值。结合江苏省C级GPS水准网进行试验,拟合效果有明显的改进,克服了目前研究中存在的问题。     

Artificial Neural Networks for the analysis of spread⿿mooring configurations for floating production systems

This work presents the development of Artificial Neural Networks for the analysis of any arbitrarily defined spread-mooring configuration for floating production systems (FPS), considering a given scenario characterized by the water depth, metocean data, characteristics of the platform hull, and the riser layout. The methodology is applied to recent designs of deepwater semi-submersible platforms connected to a large number of risers with asymmetrical layout. In such cases, the design variables may include values for the azimuthal spacing and mooring radius varying along the corners of the platform, besides the pretension and material of the lines. The results of the case study indicated that, given any mooring configuration characterized by the combination of all these design variables, the ANNs provide fairly accurate values for the parameters of the response that are required for the design of mooring systems (typically platform offsets and line tensions).     

Spatial predictions of geological rock mass properties based on in-situ interpretations of multi-dimensional seismic data

The article presents a statistical approach to characterize and predict engineering geological conditions in the up to 2000 m deep Faido tunnel and Gotthard base tunnel in Switzerland. Seismic investigations were conducted to improve the technology of interpreting seismic tomographic images. Overall, the goal of this study was to predict spacial maps of geological rock mass properties, such as, uniaxial compressive strength or total fracture spacing, by using up to six seismic features in combination, e.g., compression-wave and shear-wave velocities and dynamic Poisson's ratio. Self-Organizing Mapping (SOM), an artificial intelligent method, was used for the purposes of interpreting multi-dimensional geophysical attributes derived from seismic profiles of tomographic images along tunnel sidewalls. The SOM-method was applied in the Faido tunnel to delineate complex physical relations between the geological and seismic parameters. Then, the method was applied to predict geological properties around a segment of the Gotthard base tunnel with unknown geological–geotechnical conditions. The results illuminate that correlation analyses (pairwise parameter classification) are substantially less powerful than the SOM-method (multi-parameter classification) in order to interpret geological features from seismic in-situ data. Moreover, predicted spatial distributions of the total fracture spacing and the uniaxial compressive strength, for example, corresponded well with drill core and tunnel mapping results. The SOM-approach was a helpful tool for practitioners in predicting zones of instabilities and geological complexity during underground excavation processes of the Gotthard base tunnel. It is suggested to use such an interpretation method as decision support for purposes of sub/surface exploration and long-term geophysical monitoring of large-scale geoengineering projects, such as, disposals of nuclear waste and greenhouse gases or geopower plants for renewable energy (geothermal, biosoils).     

Forward Modeling with Forced Neural Networks for Gravity Anomaly Prof?le

In this paper, we introduce a new method called Forced Neural Network (FNN) to find the parameters of the object in geophysical section respect to gravity anomaly assuming the prismatic model. The aim of the geological modeling is to find the shape and location of underground structure, which cause the anomalies, in 2D cross section. At the first stage, we use one neuron to model the system and apply back propagation algorithm to find out the density difference. At the second level, quantization is applied to the density differences and mean square error of the system is computed. This process goes on until the mean square error of the system is small enough. First, we use FNN to two synthetic data, and then the Sivas–Gürün basin map in Turkey is chosen as a real data application. Anomaly values of the cross section, which is taken from the gravity anomaly map of Sivas–Gürün basin, are very close to those obtained from the proposed method.     

Assessment of liquefaction triggering using strain energy concept and ANN model: Capacity Energy

In the present study, an artificial neural network (ANN) model was developed to establish a correlation between soils initial parameters and the strain energy required to trigger liquefaction in sands and silty sands. A relatively large set of data including 284 previously published cyclic triaxial, torsional shear and simple shear test results were employed to develop the model. A subsequent parametric study was carried out and the trends of the results have been confirmed via some previous laboratory studies. In addition, the data recorded during some real earthquakes at Wildlife, Lotung and Port Island Kobe sites plus some available centrifuge tests data have been utilized in order to validate the proposed ANN-based liquefaction energy model. The results clearly demonstrate the capability of the proposed model and the strain energy concept to assess liquefaction resistance (capacity energy) of soils.     

预测工后软土地基沉降

本文简要叙述沉降观测在软土地基施工中的重要性,并提出了基于神经网络的高速公路工后沉降预测方法。结合工程实例运用MATLAB工具箱函数建立网络模型,选择合适的训练函数,使其训练精度和速度达到最优化。     

Neural network modeling for regional hazard assessment of debris flow in Lake Qionghai Watershed, China

This paper presents a neural network (NN) based model to assess the regional hazard degree of debris flows in Lake Qionghai Watershed, China. The NN model was used as an alternative for the more conventional linear model MFCAM (multi-factor composite assessment model) in order to effectively handle the nonlinearity and uncertainty inherent in the debris flow hazard analysis. The NN model was configured using a three layer structure with eight input nodes and one output node, and the number of nodes in the hidden layer was determined through an iterative process of varying the number of nodes in the hidden layer until an optimal performance was achieved. The eight variables used to represent the eight input nodes include density of debris flow gully, degree of weathering of rocks, active fault density, area percentage of slope land greater than 25° of the total land (APL25), frequency of flooding hazards, average covariance of monthly precipitation by 10 years (ACMP10), average days with rainfall >25 mm by 10 years (25D10Y), and percentage of cultivated land with slope land greater than 25° of the total cultivated land (PCL25). The output node represents the hazard-degree ranks (HDR). The model was trained with the 35 sets of data obtained from previous researches reported in literatures, and an explicit uncertainty analysis was undertaken to address the uncertainty in model training and prediction. Before the NN model is extrapolated to Lake Qionghai Watershed, a validation case, different from the above data, is conducted. In addition, the performances of the NN model and the MFCAM were compared. The NN model predicted that the HDRs of the five sub-watersheds in the Lake Qionghai Watershed were IV, IV, III, III, and IV–V, indicating that the study area covers normal hazard and severe hazard areas. Based on the NN model results, debris flow management and economic development strategies in the study are proposed for each sub-watershed.