Numerical Prediction Method Research Articles

Applied machine learning in wind speed prediction and loss minimization in unbalanced radial distribution system

ABSTRACT Environmental parameter consideration has always prompted wind power to be used as renewable energy. However, the biggest challenge lies in wind energy integration to the power grid due to wind intermittency. Wind speed or wind power forecasting is one of the approach to manage this intermittency. Numerous prediction methods have been reported in previous literatures over few years. In this work, multivariate wind speed forecasting using Machine Learning framework in a python environment is executed. Several statistical models and neural network models are examined to best predict the wind speed of Surat, India [22.2587° N, 71.1924° E]. The model efficiency is tested in terms of measurements of correlation factors and Mean Absolute Error values. The predicted wind speed value is further considered for the power generation from the wind farm and integrated to the distribution system. Load Impedance Matrix method is implemented for Distribution System Load Flow analysis for being robust and simple with single-step computation. IEEE-19 bus and IEEE-25 bus-unbalanced radial distribution systems are considered for finding the power losses in the branches with wind power as Distributed Generation. An efficient and effective optimization technique, Teaching Learning-Based Optimization, is used to obtain the optimal location and capacity of Distributed Generation to minimize the power loss in the distribution lines.

Effect of spatial distribution of fibers on elastic properties of unidirectional carbon/carbon composites

Carbon/Carbon composites finds its applications in several high temperature applications in the field of Space, Aviation etc. Designing of components or sub systems with carbon/carbon composites is a challenging task. It requires prediction of elastic properties with a very high accuracy. The prediction can be normally done by analytical, numerical or experimental methods. At the design stage the designers resort to numerical predictions as the experimental methods are not feasible during design stage. Analytical methods are complex and difficult to implement. The designers use numerical methods for prediction of elastic properties using Finite Element Modeling (FEM). The spatial distribution of fibers in matrix has an effect on results of prediction of elastic constants. The generation of random spatial distribution of fibers in representative volume element (RVE) challenging. The present work is aimed at study of effect of spatial distribution of fiber in numerical prediction of elastic properties of unidirectional carbon/carbon composites. MATLAB algorithm is used to generate the spatial distribution of fibers in unidirectional carbon/carbon composites. The RVE elements with various random fiber distributions are modeled using numerical Finite element Model using ABAQUS with EasyPBC plugin. The predicted elastic properties have shown significant variation to uniformly distributed fibers.

Deep Learning-Based Weather Prediction: A Survey

Abstract Weather forecasting plays a fundamental role in the early warning of weather impacts on various aspects of human livelihood. For instance, weather forecasting provides decision making support for autonomous vehicles to reduce traffic accidents and congestions, which completely depend on the sensing and predicting of external environmental factors such as rainfall, air visibility and so on. Accurate and timely weather prediction has always been the goal of meteorological scientists. However, the conventional theory-driven numerical weather prediction (NWP) methods face many challenges, such as incomplete understanding of physical mechanisms, difficulties in obtaining useful knowledge from the deluge of observation data, and the requirement of powerful computing resources. With the successful application of data-driven deep learning method in various fields, such as computer vision, speech recognition, and time series prediction, it has been proven that deep learning method can effectively mine the temporal and spatial features from the spatio-temporal data. Meteorological data is a typical big geospatial data. Deep learning-based weather prediction (DLWP) is expected to be a strong supplement to the conventional method. At present, many researchers have tried to introduce data-driven deep learning into weather forecasting, and have achieved some preliminary results. In this paper, we survey the state-of-the-art studies of deep learning-based weather forecasting, in the aspects of the design of neural network (NN) architectures, spatial and temporal scales, as well as the datasets and benchmarks. Then we analyze the advantages and disadvantages of DLWP by comparing it with the conventional NWP, and summarize the potential future research topics of DLWP.

W-Index: An Index for Evaluating Link Prediction considering Only the Role of Wins

With the emergence of numerous link prediction methods, how to accurately evaluate them and select the appropriate one has become a key problem that cannot be ignored. Since AUC was first used for link prediction evaluation in 2008, it is arguably the most preferred metric because it well balances the role of wins (the testing link has a higher score than the unobserved link) and the role of draws (they have the same score). However, in many cases, AUC does not show enough discrimination when evaluating link prediction methods, especially those based on local similarity. Hence, we propose a new metric, called W-index, which considers only the effect of wins rather than draws. Our extensive experiments on various networks show that the W-index makes the accuracy scores of link prediction methods more distinguishable, and it can not only widen the local gap of these methods but also enlarge their global distance. We further show the reliability of the W-index by ranking change analysis and correlation analysis. In particular, some community-based approaches, which have been deemed effective, do not show any advantages after our reevaluation. Our results suggest that the W-index is a promising metric for link prediction evaluation, capable of offering convincing discrimination.

Direct Dynamic-Simulation Approach to Trajectory Optimization for Rotorcraft Category-A Maneuver Procedures

This paper treats numerical methods for an efficient prediction of the rotorcraft emergency procedures after engine failures. The analytical means of compliance for the Category-A requirements can be used for the type certification of the transport-category rotorcraft when their fidelities are approved by the civil airworthiness authority. However, the most promising trajectory optimization approaches to the Category-A maneuver analyses typically suffer from a dimensionality problem when a high-fidelity math model is adopted. To cope with such difficulties, the paper proposes new techniques, where the system states except the initial ones and all dynamic constraints are removed from the resultant nonlinear programming problem. For these proposes, the controls are parameterized using the Hermit splines with the local support and efficient recursive formulas to predict the constraint-function Jacobians are derived. The efficiency of the proposed techniques is compared with that using the pseudo-spectral collocation method. In addition to an autorotational descent maneuver, four Category-A procedures for the continued takeoff, rejected takeoff, continued landing, and balked landing maneuvers are analyzed with varying the engine-failure conditions and with a suitable consideration on the pilot-delay time to validate the usefulness of the proposed methods.

Numerical simulation of the seismic response and soil\u2013structure interaction for a monitored masonry school building damaged by the 2016 Central Italy earthquake

Despite significant research advances on the seismic response analysis, there is still an urgent need for validation of numerical simulation methods for prediction of earthquake response and damage. In this respect, seismic monitoring networks and proper modelling can further support validation studies, allowing more realistic simulations of what earthquakes can produce. This paper discusses the seismic response of the “Pietro Capuzi” school in Visso, a village located in the Marche region (Italy) that was severely damaged by the 2016–2017 Central Italy earthquake sequence. The school was a two-story masonry structure founded on simple enlargements of its load-bearing walls, partially embedded in the alluvial loose soils of the Nera river. The structure was monitored as a strategic building by the Italian Seismic Observatory of Structures (OSS), which provided acceleration records under both ambient noise and the three mainshocks of the seismic sequence. The evolution of the damage pattern following each one of the three mainshocks was provided by on-site survey integrated by OSS data. Data on the dynamic soil properties was available from the seismic microzonation study of the Visso village and proved useful in the development of a reliable geotechnical model of the subsoil. The equivalent frame (EF) approach was adopted to simulate the nonlinear response of the school building through both fixed-base and compliant-base models, to assess the likely influence of soil–structure interaction on the building performance. The ambient noise records allowed for an accurate calibration of the soil–structure model. The seismic response of the masonry building to the whole sequence of the three mainshocks was then simulated by nonlinear time history analyses by using the horizontal accelerations recorded at the underground floor as input motions. Numerical results are validated against the evidence on structural response in terms of both incremental damage and global shear force–displacement relationships. The comparisons are satisfactory, corroborating the reliability of the compliant-base approach as applied to the EF model and its computational efficiency to simulate the soil–foundation–structure interaction in the case of masonry buildings.

Research on prediction method of pulsating pressure of underwater vehicles

When fluid flows through the underwater vehicles, flow separation occurs, and the flow separation generates eddy currents, which in turn induce the pulsating pressure on the surfaces of underwater vehicles, generating noise and making the underwater vehicle easy to expose. We conducted research on this phenomenon, taking cone-column shell as the research object and based on the Large Eddy Simulation (LES) theory. Additionally, the calculation results of each parameter such as grid scale of the structural surface, grid structure and height of the first layer grid with different values were compared with the experimental values to determine the model simulation parameter values with the highest calculation accuracy. Furthermore, the systematic numerical prediction method of pulsating pressure was supposed by the combination of numerical and experimental methods. Moreover, accuracy of the prediction method was verified by experiments. The prediction method provides a generally good reference basis and a simple tool for engineering applications.

The Impact of Coating Ingredients on the Aging Resistance of Topcoat Paints by Model Trees

Topcoat paint is mainly composed of resin and pigment and hence its quality highly depends on the type and proportion of these two ingredients. This study aims at testing the formula of the topcoat paint for finding one that can achieve better quality for anti-aging. Various formulas of paint are applied on boards that will be put into ultraviolet accelerated test machines to simulate weathering tests. The gloss and color, before and after the tests, are collected and numerical prediction method M5P is used to grow model trees for discovering the key factors affecting aging. Based on the structure and the linear regression models in the trees, a better topcoat paint should be composed of a high proportion of resin and generally a low proportion of pigment. Good types of resin and pigment are also identified for keeping color and gloss.

Prediction of vehicle impact speed based on the post-cracking behavior of automotive PVB laminated glass: Analytical modeling and numerical cohesive zone modeling

Targeted investigation of the impact fracture performance of PVB laminated glass, which is a combined analytical and numerical method for impact dynamics prediction, is presented. In this method, the relationship between the impact force and the post-cracking response of laminated glass is developed in the framework of impact dynamics to present a prediction formulation. The formulation is utilized to be applied to predict vehicle impact speed based on the crack patterns propagated on the automotive laminated glass due to the pedestrian head impact. To prepare a numerical set-up, an automotive laminated glass finite element model is proposed by inserting cohesive elements on all common surfaces of glass layers and glass-PVB interfacial. The prediction model is validated well by comparison with numerical crash patterns and ten real vehicle–pedestrian accident cases.

Tropical Cyclone Intensity Change Prediction Based on Surrounding Environmental Conditions with Deep Learning

Tropical cyclone (TC) motion has an important impact on both human lives and infrastructure. Predicting TC intensity is crucial, especially within the 24 h warning time. TC intensity change prediction can be regarded as a problem of both regression and classification. Statistical forecasting methods based on empirical relationships and traditional numerical prediction methods based on dynamical equations still have difficulty in accurately predicting TC intensity. In this study, a prediction algorithm for TC intensity changes based on deep learning is proposed by exploring the joint spatial features of three-dimensional (3D) environmental conditions that contain the basic variables of the atmosphere and ocean. These features can also be interpreted as fused characteristics of the distributions and interactions of these 3D environmental variables. We adopt a 3D convolutional neural network (3D-CNN) for learning the implicit correlations between the spatial distribution features and TC intensity changes. Image processing technology is also used to enhance the data from a small number of TC samples to generate the training set. Considering the instantaneous 3D status of a TC, we extract deep hybrid features from TC image patterns to predict 24 h intensity changes. Compared to previous studies, the experimental results show that the mean absolute error (MAE) of TC intensity change predictions and the accuracy of the classification as either intensifying or weakening are both significantly improved. The results of combining features of high and low spatial layers confirm that considering the distributions and interactions of 3D environmental variables is conducive to predicting TC intensity changes, thus providing insight into the process of TC evolution.

Advanced numerical methods for the strength prediction of hybrid adhesively bonded T-peel joints

ABSTRACT The market requirements impose searching for new joining solutions, which must be simple and effective. Adhesive joints are currently used, for example in the automotive and aeronautical industries, since welding is not possible in many cases and because drilling the base materials is not required. To analyse adhesive joints, the Finite Element Method (FEM) is increasingly used. The Extended Finite Element Method (XFEM) is able to predict crack growth and joint strength in adhesive structures. However, there are still not many case studies about its suitability. On the other hand, hybrid joints, which combine spot-welding with adhesive bonding, are being increasingly used in different industries. This study has the purpose of validating the XFEM to predict the performance of T-peel hybrid joints between DIN C45E steel adherends under tensile loads. A comparison is made with spot-welded and adhesive joints. Three adhesives are tested (Araldite® AV138, Araldite® 2015 and Sikaforce® 7752). Different damage initiation and growth criteria, and also damage law shapes, were evaluated for strength and dissipated energy predictions. The joints’ performance was found to highly vary between adhesives and bonding method, and the XFEM analysis is accurate for specific modelling conditions.

Efficient learning of a one-dimensional density functional theory

Density functional theory underlies the most successful and widely used numerical methods for electronic structure prediction of solids. However, it has the fundamental shortcoming that the universal density functional is unknown. In addition, the computational result---energy and charge density distribution of the ground state---is useful for electronic properties of solids mostly when reduced to a band structure interpretation based on the Kohn-Sham approach. Here, we demonstrate how machine learning algorithms can help to free density functional theory from these limitations. We study a theory of spinless fermions on a one-dimensional lattice. The density functional is implicitly represented by a neural network, which predicts, besides the ground-state energy and density distribution, density-density correlation functions. At no point do we require a band structure interpretation. The training data, obtained via exact diagonalization, feeds into a learning scheme inspired by active learning, which minimizes the computational costs for data generation. We show that the network results are of high quantitative accuracy and, despite learning on random potentials, capture both symmetry-breaking and topological phase transitions correctly.

Generalized Numerical Differentiation Method for Stability Calculation of Periodic Delayed Differential Equation: Application for Variable Pitch Cutter in Milling

This paper proposes a generalized numerical differentiation method for stability prediction of the non-autonomous delayed differential equations (DDEs) with periodic coefficients and discrete delays. Firstly, the periodic DDE is described in state-space form and the period of a system is equally discretized. Then, the discrete first derivatives versus time are approximated by a linear combination of the state function values at multiple neighboring sampling grid points based on the finite-difference formulas. Such that, the original DDE is approximated as a series of algebraic equations and the Floquet transition matrix can be constructed on one period. At last, the system stability is determined according to the Floquet theory by checking the eigenvalues. The delayed damped Mathieu equation is regarded as a typical case to verify the effectiveness and efficiency of the presented method. The stability diagrams and rate of convergence are computed in comparison with those via the benchmark algorithm (the semi-discretization method). As an application, the presented method is used to predict the stability of milling with variable pitch cutter, and the computational result agrees well with the experimentally verified example.

Precipitation forecasting by large-scale climate indices and machine learning techniques

Global warming is one of the most complicated challenges of our time causing considerable tension on our societies and on the environment. The impacts of global warming are felt unprecedentedly in a wide variety of ways from shifting weather patterns that threatens food production, to rising sea levels that deteriorates the risk of catastrophic flooding. Among all aspects related to global warming, there is a growing concern on water resource management. This field is targeted at preventing future water crisis threatening human beings. The very first stage in such management is to recognize the prospective climate parameters influencing the future water resource conditions. Numerous prediction models, methods and tools, in this case, have been developed and applied so far. In line with trend, the current study intends to compare three optimization algorithms on the platform of a multilayer perceptron (MLP) network to explore any meaningful connection between large-scale climate indices (LSCIs) and precipitation in the capital of Iran, a country which is located in an arid and semi-arid region and suffers from severe water scarcity caused by mismanagement over years and intensified by global warming. This situation has propelled a great deal of population to immigrate towards more developed cities within the country especially towards Tehran. Therefore, the current and future environmental conditions of this city especially its water supply conditions are of great importance. To tackle this complication an outlook for the future precipitation should be provided and appropriate forecasting trajectories compatible with this region’s characteristics should be developed. To this end, the present study investigates three training methods namely backpropagation (BP), genetic algorithms (GAs), and particle swarm optimization (PSO) algorithms on a MLP platform. Two frameworks distinguished by their input compositions are denoted in this study: Concurrent Model Framework (CMF) and Integrated Model Framework (IMF). Through these two frameworks, 13 cases are generated: 12 cases within CMF, each of which contains all selected LSCIs in the same lead-times, and one case within IMF that is constituted from the combination of the most correlated LSCIs with Tehran precipitation in each lead-time. Following the evaluation of all model performances through related statistical tests, Taylor diagram is implemented to make comparison among the final selected models in all three optimization algorithms, the best of which is found to be MLP-PSO in IMF.

Methods for Numerical Prediction of Relaxation and Reduction Properties of Polymer Textile Materials

Methods of numerical prediction of relaxation and reduction properties of polymer textile materials are considered. Numerical prediction is performed based on integration of the governing Boltsmann–Volterra equations applied to relaxation-reduction processes of the materials under investigation.

Experimental and Analytical Investigation of Sand Production in Weak formations for Multiple Well Shut-Ins

Sand production is a severe problem in oilfields and is majorly associated with weak and unconsolidated formations. Therefore, identifying the expected amount and rate of sand has been approached by different analytical and numerical prediction methods. Within them the former requires less computational time and therefore being widely used. However, despite some of the analytical models proved to be fairly predictive for sand production during steady-state flow, they cannot be used directly for the wells with complex production profiles. The current study focuses on the experimental investigation of the sand production pattern during well production and shut-in periods. The laboratory work was conducted on artificial sandstone replicating weak Cretaceous formations in Kazakhstan using customized High Pressure Consolidation System (HPCS) for sand production experiments. HPCS accommodates a sandstone sample of a large diameter to diminish boundary effect when the fluid is injected. The tests were conducted at different overburden stresses as well as pressure drawdowns in several well shut-in runs. The paper presents a modification of the classical steady-state flow analytical solution incorporating field and fluid conditions as well as the dependence on the well shut-in runs. The model showed good convergence with the experimental and well data.

A numerical external pitting damage prediction method of buried pipelines

Abstract The work introduces a numerical external damage prediction method for buried pipelines. The external pitting initiation and corrosion rate of oil or gas pipelines are affected by pipeline age, physicochemical properties of soils and cathodic protection performance as well as coating conditions. Before developing the damage prediction model, the influencing factors were weighed by grey relational analysis, and then the relationship among the pitting depth and the influencing factors of external corrosion was established for corrosion damage prediction through artificial neural network (ANN). Subsequently, the established ANN was applied to predict corrosion damage and corrosion rate for some selected cases, and the neural network prediction model was analyzed and compared to another corrosion rate prediction models. Through the analysis and comparison, a few opinions were proposed on the external corrosion damage prediction and pipeline integrity management.

Fast buckling load numerical prediction method for imperfect shells under axial compression based on POD and vibration correlation technique

Vibration correlation technique (VCT) is an effective non-destructive buckling experimental technique for shell structures. In this study, VCT is studied from the point-of-view of being a buckling load numerical prediction method by numerically simulating the experimental procedure of VCT. Firstly, the formulas of VCT are introduced for axially loaded cylindrical shells and conical shells under the clamped–clamped boundary condition. According to the VCT formulas, the numerical procedure of VCT is provided. In order to accelerate the repeated eigenvalue analysis of VCT, the proper orthogonal decomposition (POD) method is integrated into VCT, and the POD-VCT is developed. Extensive examples are presented to verify the effectiveness of the proposed method, including unstiffened cylindrical shell with real measured imperfection, unstiffened conical shell with single perturbation load imperfection, composite cylindrical shell with eigenmode imperfection, and hierarchical stiffened cylindrical shell with combined imperfection. In comparison to buckling test results, high-fidelity explicit dynamic method and VCT method, the high prediction accuracy and efficiency of the proposed POD-VCT are fully demonstrated. Additionally, example results indicate the strong applicability of the proposed POD-VCT for various types of structural configurations, materials and imperfections. Above all, the POD-VCT is verified to be a fast buckling load numerical prediction method for imperfect shells.

A FEM Multiscale Homogenization Procedure using Nanoindentation for High Performance Concrete

This paper aims to develop a numerical multiscale homogenization method for prediction of elasto-viscoplastic properties of a high performance concrete (HPC). The homogenization procedure is separated into two-levels according to the microstructure of the HPC: the mortar or matrix level and the concrete level. The elasto-viscoplastic behavior of individual microstructural phases of the matrix are identified from nanoindentation data using an inverse identification method. The micromechanical results are then used as input parameters for numerical elasto-viscoplastic homogenization at microscale. The mortar level is analyzed with numerical homogenization by using the finite element simulation to predict the overall elasto-viscoplastic properties of HPC. The results are compared with macroscopic experimental and analytical results from the literature showing a good agreement.

Solar irradiance resource and forecasting: a comprehensive review

With the increase in demand for energy, penetration of alternative sources of energy in the power grid has increased. Photovoltaic (PV) energy is the most common and popular form of energy sources which is widely integrated into the existing grid. As solar energy is intermittent in nature, to ensure uninterrupted and reliable power supply to the prosumers, it is essential to forecast the solar irradiance. Accurate solar forecasting is necessary to facilitate large-scale modelling and deployment of PV plants without disrupting the quality and reliability of the power grid as well as to manage the power demand and supply. There are various methods to predict the solar irradiance such as numerical weather prediction methods, satellite-based approaches, cloud-image based methodologies, data-driven methods, and sensor-network based approaches. This study gives an overall review of the different resources and methods used for forecasting solar irradiance in different time horizons and also gives an extensive review of the sensor networks that are used for determining solar irradiance. The various error metrics and accessible data sets available for the sensor networks are also discussed that can be used for validation purposes.