机器人修磨中融合先验知识的适应学习建模方法

针对机器人修磨磨削量建模中处理突变因素的难题,本文首先从机器学习建模方法的角度指出该问题与统计学习的不同点,并把问题形式化,然后在此基础上提出了融合先验知识的适应学习建模方法.该方法基于半经验公式生成虚拟样本,不但弥补了适应学习建模中新样本不足的问题,而且把半经验公式中的信息更充分地融合到学习机模型中.实验结果证明,该...     

IDENTIFIABILITY IN CAUSAL BAYESIAN NETWORKS: A GENTLE INTRODUCTION

In this article we describe an important structure used to model causal theories and a related problem of great interest to semi-empirical scientists. A causal Bayesian network is a pair consisting of a directed acyclic graph (called a causal graph) that represents causal relationships and a set of probability tables, that together with the graph specify the joint probability of the variables represented as nodes in the graph. We briefly describe the probabilistic semantics of causality proposed by Pearl for this graphical probabilistic model, and how unobservable variables greatly complicate models and their application. A common question about causal Bayesian networks is the problem of identifying causal effects from nonexperimental data, which is called the identifability problem. In the basic version of this problem, a semi-empirical scientist postulates a set of causal mechanisms and uses them, together with a probability distribution on the observable set of variables in a domain of interest, to predict the effect of a manipulation on some variable of interest. We explain this problem, provide several examples, and direct the readers to recent work that provides a solution to the problem and some of its extensions. We assume that the Bayesian network structure is given to us and do not address the problem of learning it from data and the related statistical inference and testing issues.     

Surface soil moisture estimation over dense crop using Envisat ASAR and Landsat TM imagery: an approach

This study focuses on developing a new method of surface soil moisture estimation over wheat fields using Environmental Satellite Advanced Synthetic Aperture Radar (Envisat ASAR) and Landsat Thematic Mapper (TM) data. The Michigan Microwave Canopy Scattering (MIMICS) model was used to simulate wheat canopy backscattering coefficients from experiment plots at incidence angles of 23° (IS2) and 43.9° (IS7). Based on simulated data, the scattering characteristics of a wheat canopy were first investigated in order to derive an optimal configuration of polarization (HH) and incidence angle (IS2) for soil moisture estimation. Then a parametric model was developed to describe wheat canopy backscattering at the optimal configuration. In addition, direct backscattering and two-way transmissivity of wheat crowns were derived from the TM normalized difference vegetation index (NDVI); direct ground backscattering was derived from surface soil moisture and TM NDVI; and backscattering from double scattering was derived from total backscattering. A semi-empirical model for soil moisture estimation was derived from the parametric model. Coefficients in the semi-empirical model were obtained using a calibration approach based on measured soil moisture, ASAR, and TM data. A validation of the model was performed over the experimental area. In this study, the root mean square error (RMSE) for the estimated soil moisture was 0.041 m³ m^?3, and the correlation coefficient between the measured and estimated soil moisture was 0.84. The experimental results indicate that the semi-empirical model could improve soil moisture estimation compared to an empirical model.     

Development of parameter sets for semi-empirical MO calculations of transition metal systems: iron parameters for iron-sulfur proteins

A semi-empirical parameter set for iron has been developed which is appropriate for the study of iron-sulfur proteins having a single iron atom, by fitting to density-functional theory (DFT) calculations obtained for a series of small models of iron-containing proteins. These parameters are obtained using a modified BFGS optimisation procedure previously used to obtain semi-empirical parameters for the main group elements. The modifications to this procedure for obtaining parameters for transition metal atoms are outlined. In addition to modifications to the semi-empirical core repulsion function, which yield significant improvements in the calculation of molecular structures, compared to the standard core repulsion function, are outlined. The reported parameters are then tested on a set of model complexes containing a variety of ligands and show good agreement with both DFT and experimental data for these species.     

One-class classification for oil spill detection

SAR oil spill classification is a challenging topic, which is tackled by semi-empirical ad hoc approaches supported by very qualified experts. In all cases, the feature space is empirically defined, and two-class classification approaches are used. Although this approach allows achieving acceptable operational results, there is still room for improving both the comprehension of the physical phenomenon and the performance of classification techniques. In this paper, we propose a novel approach to oil-spill classification based on the paradigm of one-class classification. A classifier is trained using only examples of oil spills, instead of using oil spills and look-alikes, as in two-class approaches. Further, since the feature space is empirically defined, we also propose an objective technique to select the most powerful one that is suited for the oil-spill detection task at hand. Results on two case study datasets are reported to validate the proposed approach.     

Toward a heuristic optimum design of rolling schedules for tandem cold rolling mills

Scheduling for tandem cold mills refers to the determination of inter-stand gauges, tensions and speeds of a specified product. Optimal schedules should result in maximized throughput and minimized operating cost. This paper presents a genetic algorithm based optimization procedure for the scheduling of tandem cold rolling mills. The optimization procedure initiates searching from a logical staring point — an empirical rolling schedule — and ends with an optimum cost. Cost functions are constructed to heuristically direct the genetic algorithm’s searching, based on the consideration of power distribution, tension, strip flatness and rolling constraints. Numerical experiments have shown that the proposed method is more promising than those based on semi-empirical formulae. The results generated from a case study show that the proposed approach could significantly improve empirically derived settings for the tandem cold rolling mills.     

Soil moisture retrieval over agricultural fields from multi-polarized and multi-angular RADARSAT-2 SAR data

The aim of this study was to estimate soil moisture from RADARSAT-2 Synthetic Aperture Radar (SAR) images acquired over agricultural fields. The adopted approach is based on the combination of semi-empirical backscattering models, four RADARSAT-2 images and coincident ground measurements (soil moisture, soil surface roughness and vegetation characteristics) obtained near Saskatoon, Saskatchewan, Canada during the summer of 2008. The depolarization ratio (χ_v), the co-polarized correlation coefficient (ρ_vvhh) and the ratio of the absolute value of cross polarization to crop height (Λ_vh) derived from RADARSAT-2 data were analyzed with respect to changes in soil surface roughness, crop height, soil moisture and vegetation water content. This sensitivity analysis allowed us to develop empirical relationships for soil surface roughness, crop height and crop water content estimation regardless of crop type. The latter were then used to correct the semi-empirical Water-Cloud model for soil surface roughness and vegetation effects in order to retrieve soil moisture data. The soil moisture retrieved algorithm is evaluated over mature crop fields (wheat, pea, lentil, and canola) using ground measurements. Results show average relative errors of 19%, 10%, 25.5% and 32% respectively for the retrieval of crop height, soil surface roughness, crop water content and soil moisture.     

Generalized semi-empirical modelling of wheat radar response

Two semi-empirical models, simulating the backscattering coefficient from different crops, were tested on the same wheat canopy. The physics of the semi-empirical models and the limitations of these models when scattering effects are discussed. The results show that both models can be included in a generalized formulation adapted to a large range of measurement conditions. This conclusion confirms that the radar configuration (waveband, polarization, incidence angle) plays a role similar to that of the canopy structure in the control of the radar backscattering coefficient.     

Polarimetric contrast classification of agricultural fields using MAESTRO 1 AIRSAR data

In this paper, we show the results of a polarimetric contrast classifier applied to the MAESTRO I JPL-AIRSAR data of agricultural fields. For the Dutch Flevoland agricultural site, we first compare the classifier's performance to results given in earlier studies, where polarimetric information was not available. This comparison shows that the availability of the measurements of the full polarimetric covariance matrix greatly enhances single-date classification accuracy. In the next step, we illustrate the use of frequency diverse classification results for further separation within specific classes. Furthermore, we will show the use of the classifierin two unsupervised approaches. In the first approach, we define the training clases from theoretical and semi-empirical models and run the classifier on a fully calibrated Fievoland dataset. In the second approach, we take covariance matrices extracted from the Flevoland data as the training set and use these in an unsupervised classification of the U.K. Feltwell agricultural site.     

Improving semi-empirical equations of ultimate bearing capacity of shallow foundations using soft computing polynomials

This study presents the ultimate bearing capacity of shallow foundations in meaningful ways and improves its semi-empirical equations accordingly. Approaches including weighted genetic programming (WGP) and soft computing polynomials (SCP) are utilized to provide accurate prediction and visible formulas/polynomials for the ultimate bearing capacity. Visible formulas facilitate parameter studies, sensitivity analysis, and applications of pruning techniques. Analytical results demonstrate that the proposed SCP is outstanding in both prediction accuracy and provides simple polynomials as well. Notably, the SCP identifies that the shearing resistance angle and foundation geometry impact on improving the Vesic's semi-empirical equations.     

Simulations of flow-like landslides invading urban areas: a cellular automata approach with SCIDDICA

Different methodologies are used for modelling flow-like landslides. A common critical point concerns the flooding into town areas, which cannot be assimilated straight to a morphology, especially, when the urban tissue is very irregular with narrow streets and odd setting of buildings, so that discretization processes of approximating numerical methods have to be carefully examined in their limits. A semi-empirical approach by the computational paradigm of cellular automata is here considered with SCIDDICA, a competitive (related to PDE approach) cellular automata model for 3-dimensions simulation of flow-like landslides. This paper presents innovations to the transition function of SCIDDICA-SS2, which manage opportunely building data in the cells corresponding to the urban tissue. The novelties of the transition function need a theorem, here demonstrated which regards the Algorithm of Minimization of Differences in the new context of inhomogeneous cells. This progress permits to simulate the complete evolution of landslides, from the detachment area to its exhaustion almost on the same precision level. This is an advantage for hazard and risk analyses in threatened zones. Improved SCIDDICA-SS2 was applied successfully to all the well-known 2009 debris flows of Giampilieri Superiore (Sicily) also in comparison with simulation results of the previous versions.     

Process modeling with the regression network

A new connectionist network topology called the regression network is proposed. The structural and underlying mathematical features of the regression network are investigated. Emphasis is placed on the intricacies of the optimization process for the regression network and some measures to alleviate these difficulties of optimization are proposed and investigated. The ability of the regression network algorithm to perform either nonparametric or parametric optimization, as well as a combination of both, is also highlighted. It is further shown how the regression network can be used to model systems which are poorly understood on the basis of sparse data. A semi-empirical regression network model is developed for a metallurgical processing operation (a hydrocyclone classifier) by building mechanistic knowledge into the connectionist structure of the regression network model. Poorly understood aspects of the process are provided for by use of nonparametric regions within the structure of the semi-empirical connectionist model. The performance of the regression network model is compared to the corresponding generalization performance results obtained by some other nonparametric regression techniques.     

First and second order convex approximation strategies in structural optimization

In this paper, various methods based on convex approximation schemes are discussed, that have demonstrated strong potential for efficient solution of structural optimization problems. First, theconvex linearization method (CONLIN) is briefly described, as well as one of its recent generalizations, themethod of moving asymptotes (MMA). Both CONLIN and MMA can be interpreted as first order convex approximation methods, that attempt to estimate the curvature of the problem functions on the basis of semi-empirical rules. Attention is next directed toward methods that use diagonal second derivatives in order to provide a sound basis for building up high quality explicit approximations of the behaviour constraints. In particular, it is shown how second order information can be effectively used without demanding a prohibitive computational cost. Various first and second order approaches are compared by applying them to simple problems that have a closed form solution.     

Molecular dynamics simulations with many-body potentials on multiple GPUs—The implementation,package and performance

Molecular dynamics (MD) is an important research tool extensively applied in materials science. Running MD on a graphics processing unit (GPU) is an attractive new approach for accelerating MD simulations. Currently, GPU implementations of MD usually run in a one-host-process-one-GPU (OHPOG) scheme. This scheme may pose a limitation on the system size that an implementation can handle due to the small device memory relative to the host memory. In this paper, we present a one-host-process-multiple-GPU (OHPMG) implementation of MD with embedded-atom-model or semi-empirical tight-binding many-body potentials. Because more device memory is available in an OHPMG process, the system size that can be handled is increased to a few million or more atoms. In comparison with the serial CPU implementation, in which Newton’s third law is applied to improve the computational efficiency, our OHPMG implementation has achieved a 28.9x–86.0x speedup in double precision, depending on the system size, the cut-off ranges and the number of GPUs. The implementation can also handle a group of small simulation boxes in one run by combining the small boxes into a large box. This approach greatly improves the GPU computing efficiency when a large number of MD simulations for small boxes are needed for statistical purposes.     

Performance assessment of decentralized control systems: An iterative approach

In this work, an efficient approach for performance assessment of decentralized control systems based on a general quadratic performance index involving both system states and inputs is proposed. The performance assessment problem is formulated as an optimization problem subject to constraints in the form of linear/bilinear matrix inequalities which explicitly take the block-diagonal structural constraint on decentralized control systems into account. In order to solve the optimization problem efficiently, an iterative approach based on the original optimization problem and an equivalent transformation of the original one is proposed. Specifically, the proposed approach under the assumption that the full state feedback is available is first presented; and then the approach is extended to the case that only output feedback is available. The proposed approach solves for both the best achievable performance and the corresponding controller (and observer) gains. The application of the proposed approach to two examples including a reactor–separator chemical process example illustrates the applicability and effectiveness of the proposed approach.     

An approach to precisely specifying the problem domain of design patterns

The problem domain of a design pattern describes the problem context in which the pattern can be applied. In general, determining the applicability of a pattern to a particular problem heavily relies on the knowledge and experience the developer has with the pattern. This significantly limits the use of patterns. To address this issue, we propose an approach for rigorously specifying the problem domain of patterns. This approach systematically guides one to develop rigorous specifications of a pattern's problem domain using a precise notation. The resulting specifications can be used to develop tool support for automatic evaluation of pattern applicability. We describe the approach using the Visitor pattern, and show how the resulting specification can be used to evaluate pattern applicability for a particular problem model. We also demonstrate tool support for the approach.     

Integration of soil moisture as an auxiliary parameter for the anchor pixel selection process in SEBAL using Landsat 8 and Sentinel - 1A images

ABSTRACT

The sensible heat flux component calculation in the Surface Energy Balance Algorithm for Land (SEBAL) completely relies upon the anchor pixel locations. The fragmented agriculture regions with diverse crops make the selection process more challenging with less probability to correlate with ground realities. The conventional methods for anchor pixel selection have limitations in capturing the soil moisture variations beneath the canopy as they utilize only the thermal remote sensing images and ancillary datasets like land cover and crop type maps. The current research attempts to reduce the existing gap between the ground realities and simulated results by adding soil moisture as a supplementary parameter. The prospects of semi-empirical Water Cloud Model (WCM) to estimate the soil moisture content was tested in a fragmented agricultural region for different time periods. The concurrent satellite data obtained from Sentinel - 1A and Landsat 8 were utilized to supply the inputs for inversion modelling based on the Levenberg-Marquardt approach. The current research introduced ‘virtual Normalized Difference Vegetation Index’ concept to refine the WCM and yielded reliable soil moisture output for supplementing the anchor pixel selection process. The robustness of the approach was justified by considering the available energy at anchor pixel locations. The research outcomes revealed that the anchor pixel selection with and without soil moisture criterion has a significant impact on actual evapotranspiration estimation. The research also explores the scope of the synergetic use of optical and Synthetic Aperture Radar (SAR) inputs in SEBAL.     

Decomposition of a linear-quadratic optimal control problem with fast and slow variables

A linear-quadratic optimal control problem for a discrete different time-scale system is studied. The decomposition of the boundary value problem for the maximum principle is based on the geometric approach using the properties of invariant manifolds of slow and fast motions. This approach aids in constructing a transformation for reducing the initial problem to a boundary-value problem for slow variables and two initial-value problems for fast variables. The transformation is expressed as an asymptotic siries in powers of a small parameter.