No abstract

No abstract

A cognitive modelling based new inversion method, the successive differential evolution (DE-S) algorithm, is proposed to estimate the Q factor and velocity from the zero-offset vertical seismic profile (VSP) record for oil-gas reservoir exploration. The DE algorithm seeks optimal solutions by simulating the natural species evolution processes and makes the individuals become optimal. This algorithm is suitable for the high-dimensional nonseparable model space where the inversion leads to recognition and prediction of hydrocarbon reservoirs. The viscoelastic medium is split into layers whose thicknesses equal to the space between two successive VSP geophones, and the estimated parameters of each layer span the related subspace. All estimated parameters span to a high dimensional nonseparable model space. We develop bottom-up workflow, in which the Q factor and the velocity are estimated using the DE algorithm layer by layer. In order to improve the inversion precision, the crossover strategy is discarded and we derive the weighted mutation strategy. Additionally, two kinds of stopping criteria for effective iteration are proposed to speed up the computation. The new method has fast speed, good convergence and is no longer dependent on the initial values of model parameters. Experimental results on both synthetic and real zero-offset VSP data indicate that this method is noise robust and has great potential to derive reliable seismic attenuation and velocity, which is an important diagnostic tool for reservoir characterization. Index Terms-Successive differential evolution algorithm, VSP data, high dimensional data, velocity and Q inversion I. INTRODUCTION The exploration targets are turning from conventional to unconventional reservoirs with the development of oil-gas exploration technology [1]. How to finely describe the medium structure, lithology and saturation of fluids is a critical

Material recognition plays an important role in the interaction between robots and the external environment. For example, household service robots need to replace humans in the home environment to complete housework, so they need to interact with daily necessities and obtain their material performance. Images provide rich visual information about objects; however, it is often difficult to apply when objects are not visually distinct. In addition, tactile signals can be used to capture multiple characteristics of objects, such as texture, roughness, softness, and friction, which provides another crucial way for perception. How to effectively integrate multi-modal information is an urgent problem to be addressed. Therefore, a multi-modal material recognition framework CFBRL-KCCA for target recognition tasks is proposed in the paper. The preliminary features of each model are extracted by cascading broad learning, which is combined with the kernel canonical correlation learning, considering the differences among different models of heterogeneous data. Finally, the open dataset of household objects is evaluated. The results demonstrate that the proposed fusion algorithm provides an effective strategy for material recognition.This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

This paper proposes a novel k-nearest neighbor algorithm to predict soil moisture in maize field. In order to estimate soil moisture in maize field accurately without any destruction to root and soil, this paper uses biological characteristics of maize to estimate soil moisture, including plant height, leaf area, stem diameter, dry weight and fresh weight, all the values of which are non-negative. So a novel k-nearest neighbor based on I-divergence (ID_KNN) is proposed. ID_KNN uses I-divergence as the distance metric instead of Euclidean distance, which is more effective when the data is positive. The proposed method is tested on datasets in six growth stages of maize, and the experimental results show that ID_KNN is more effective in accuracy and macro F 1 measure than traditional k-nearest neighbor algorithm.