A geostatistical approach based on ordinary kriging is presented for the evaluation and the augmentation of an existing rain gauge network. The evaluation is based on estimating the percentage of the area that achieve...
详细信息
A geostatistical approach based on ordinary kriging is presented for the evaluation and the augmentation of an existing rain gauge network. The evaluation is based on estimating the percentage of the area that achieves a targeted level of acceptable accuracy. The variances of kriging estimation erros at un-gauged locations were assumed to be normally distributed. Kriging estimation erros with a probability that equals to or exceeds a given threshold value of acceptance probability were assumed to have satisfactory accuracies. The percentage of the area that achieved the targeted probability of acceptance is delineated and used to judge the overall performance of the existing rain gauge network. A study area in northern Oman located in Sohar governorate is selected as the pilot case. The area has 34 rain gauges and it is characterized by a terrain surface that varies from coastal plain to mountains. For a threshold value of 0.85, and 0.90 of acceptance probability, the existing network achieved area of acceptable probability of 88.71 and 77.72 %, respectively. For a success criterion of 80 %, the existing rain gauge network indicated acceptable performance for acceptance probability threshold of 0.85 and inadequate performance is noticed in the case of probability threshold of 0.90, which necessitated further network augmentation. A sequential algorithm for ranking and prioritization of the existing rain gauges is used to classify the existing rain gauges into base and non-base rain gauges. The base rain gauge network for mean annual rainfall comprised about 29 of the existing rain gauges. The identified non-base rain gauges were sequentially relocated to achieve higher levels of percentage of area with acceptable accuracy. The percentage of area with acceptable accuracy increased from 88.71 % for the original rain gauge network to about 94.51 % for the augmented network by adding four rain gages at probability acceptance threshold of 0.85. It also increased from 77.
A D-optimal minimax design criterion is proposed to construct two-level fractional factorial designs, which can be used to estimate a linear model with main effects and some specified interactions. D-optimal minimax d...
详细信息
A D-optimal minimax design criterion is proposed to construct two-level fractional factorial designs, which can be used to estimate a linear model with main effects and some specified interactions. D-optimal minimax designs are robust against model misspecification and have small biases if the linear model contains more interaction terms. When the D-optimal minimax criterion is compared with the D-optimal design criterion, we find that the D-optimal design criterion is quite robust against model misspecification. Lower and upper bounds derived for the loss functions of optimal designs can be used to estimate the efficiencies of any design and evaluate the effectiveness of a search algorithm. Four algorithms to search for optimal designs for any run size are discussed and compared through several examples. An annealing algorithm and a sequential algorithm are particularly effective to search for optimal designs. (C) 2010 Elsevier B.V. All rights reserved.
We introduce an algorithm which, in the context of nonlinear regression on vector-valued explanatory variables, aims to choose those combinations of vector components that provide best prediction. The algorithm is con...
详细信息
We introduce an algorithm which, in the context of nonlinear regression on vector-valued explanatory variables, aims to choose those combinations of vector components that provide best prediction. The algorithm is constructed specifically so that it devotes attention to components that might be of relatively little predictive value by themselves, and so might be ignored by more conventional methodology for model choice, but which, in combination with other difficult-to-find components, can be particularly beneficial for prediction. The design of the algorithm is also motivated by a desire to choose vector components that become redundant once appropriate combinations of other, more relevant components are selected. Our theoretical arguments show these goals are met in the sense that, with probability converging to 1 as sample size increases, the algorithm correctly determines a small, fixed number of variables on which the regression mean, g say, depends, even if dimension diverges to infinity much faster than n. Moreover, the estimated regression mean based on those variables approximates g with an error that, to first order, equals the error which would arise if we were told in advance the correct variables. In this sense, the estimator achieves oracle performance. Our numerical work indicates that the algorithm is suitable for very high dimensional problems, where it keeps computational labor in check by using a novel sequential argument, and also for more conventional prediction problems, where dimension is relatively low.
Accurate knowledge of the heat transfer coefficient (HTC) distribution is of great significance to the control and optimization of the heat transfer system in the air jet impingement. For this purpose, the estimation ...
详细信息
Accurate knowledge of the heat transfer coefficient (HTC) distribution is of great significance to the control and optimization of the heat transfer system in the air jet impingement. For this purpose, the estimation of the temporally-spatially varying HTC in the air jet impingement is studied based on the inverse heat conduction problem. The transient two-dimensional heat conduction model of the target disc under a single cooling air jet is established and utilized as the direct problem. A self-scaling sequential quasi-Newton method (SS-SQNM) is developed for accurately and efficiently solving the inverse problem. As a modification of the sequential quasi -Newton method (SQNM), the SS-SQNM adopts the self-scaling updating equation to accelerate the single-step convergence rate during iteration, and thus attains higher inversion efficiency than the SQNM. A series of nu-merical tests are carried out to investigate the performance of the SS-SQNM and the effects of the regularization parameters on inverse results. The results reveal that the SS-SQNM can give stable and accurate estimates of the HTC when regularization parameters are appropriately selected through the discrepancy principle. The mean absolute percentage error of the estimated HTC distribution is generally less than 10.2% in the presence of measurement noises. Additionally, the iteration number required by the SS-SQNM for inversion is only about 7.81%-27.78% of that of the SQNM under the same conditions, which saves lots of computational time for the SS-SQNM. Therefore, the SS-SQNM developed in this study is a promising inverse algorithm, which can be employed to estimate the HTC distribution in the air jet impingement with reasonable accuracy and efficiency.
Given simple polygons P and Q, their separation, denoted by sigma(P, Q), is defined to be the minimum distance between their boundaries. We present a parallel algorithm for finding a closest pair among all pairs (p, q...
详细信息
Given simple polygons P and Q, their separation, denoted by sigma(P, Q), is defined to be the minimum distance between their boundaries. We present a parallel algorithm for finding a closest pair among all pairs (p, q), p is an element of P and q is an element of Q. The algorithm runs in O(log n) time using O(n) processors on a CREW PRAM, where n = vertical bar P vertical bar + vertical bar Q vertical bar. This algorithm is time-optimal and improves by a factor of O(log n) on the time complexity of previous parallel methods. The algorithm can be implemented serially in Theta(n) time, which gives the first optimal sequential algorithm for determining the separation of simple polygons. Our results are obtained by providing a unified treatment of the separation and the closest visible vertex problems for simple polygons.
We present a high-level algorithm description language which is translated to Event-B specifications for simulation, model checking and proof. Rather than trying to recover the program structure from a lower-level Eve...
详细信息
ISBN:
(纸本)9783319336008;9783319335995
We present a high-level algorithm description language which is translated to Event-B specifications for simulation, model checking and proof. Rather than trying to recover the program structure from a lower-level Event-B specification, we start with a high-level description of the algorithm. Our goals are more tractable code generation and more convenient modelling, while keeping the power of the Event-B method in terms of proof and refinement. We present various examples of algorithm descriptions and show that our translation ensures that they can be completely proven within Rodin while achieving a high-level of automatic proof.
In this paper, we rederive a particle filtering based algorithm for single-channel blind separation of co-frequency MPSK signals. A simplified implement approach is then proposed to reduce its complexity using an appr...
详细信息
ISBN:
(纸本)9781424414468
In this paper, we rederive a particle filtering based algorithm for single-channel blind separation of co-frequency MPSK signals. A simplified implement approach is then proposed to reduce its complexity using an approximation and a sequential algorithm similar to the one utilized in decoding convolutional codes. Computer simulations show that the novel algorithm is superior compared to the existing algorithms.
In this paper, we propose a novel floorplanning algorithm based on simulated annealing on GPUs. Simulated annealing is an inherently sequential algorithm, far from the typical programs suitable for Single Instruction ...
详细信息
ISBN:
(纸本)9781612849140
In this paper, we propose a novel floorplanning algorithm based on simulated annealing on GPUs. Simulated annealing is an inherently sequential algorithm, far from the typical programs suitable for Single Instruction Multiple Data (SIMD) style concurrency in a GPU. We propose a fundamentally different approach of exploring the floorplan solution space, where we evaluate concurrent moves on a given floorplan. We illustrate several performance optimization techniques for this algorithm on GPUs. Compared to the sequential algorithm, our techniques achieve 6-160X speedup for a range of MCNC and GSRC benchmarks, while delivering comparable or better solution quality.
Genetic algorithms(GAs) are suitable for parallel computing since population members fitness maybe evaluated in parallel. Most past parallel GA studies have exploited this aspect, besides resorting to different algori...
详细信息
ISBN:
(纸本)9781424481262
Genetic algorithms(GAs) are suitable for parallel computing since population members fitness maybe evaluated in parallel. Most past parallel GA studies have exploited this aspect, besides resorting to different algorithms, such as island, single-population master-slave, fine-grained and hybrid models. A GA involves a number of other operations which, if parallelized, may lead to better parallel GA implementation than those currently existing. In this paper, we parallelize binary and real-coded genetic algorithms using CUDA API's with C. Although, objective and constraint violations evaluations are embarassingly parallel, other algorithmic and code optimizations have been proposed and tested. The bottlenecks in a parallel GA implementation are identified and modified suitably. The results are compared with the sequential algorithm on accuracy and clock time for varying problems by studying the effect of a number of parameters, namely: (i) population sizes, (ii) number of threads, (iii) problem sizes, and (iv) problems of differing complexities. Significant speed-ups have been observed over the sequential GA.
Light Detection and Ranging (LiDAR) data has been used to model earth surface in an easy and economic way. As technology is developed the application of LiDAR data is also widely expanded to various areas, such as hyd...
详细信息
ISBN:
(纸本)9780769534718
Light Detection and Ranging (LiDAR) data has been used to model earth surface in an easy and economic way. As technology is developed the application of LiDAR data is also widely expanded to various areas, such as hydrological modeling, telecommunication service and urban planning. Finding accurate road networks is one of the common applications from massive LiDAR data. A novel algorithm to extract road points has been developed based on both the intensity and height information of data points. First the robustness of the sequential algorithm has been verified with real data points. Then a parallel algorithm has been developed by applying smart area partitioning. The performance of a parallel algorithm showed us a close linear speedup with the use of up to four processors. Experimental results from the parallel algorithm are presented in this paper in detail and demonstrate the robustness of the proposed method.
暂无评论