Limited Number Of Points Research Articles

Transfer learning in environmental data-driven models: A study of ozone forecast in the Alpine region

Many environmental variables, in particular, related to air or water quality, are measured in a limited number of points and often for a limited time span. This forbids the development of accurate models for interesting locations with missing or insufficient data and poses the question of whether a model developed for another measurement site can be reliably applied. Such a question is particularly critical when the model is entirely data-driven, such as a neural network. In this context, the paper proposes a procedure to evaluate the expected performance of an existing neural network model applied to a new unmonitored station. This transferability assessment is exemplified by the problem of forecasting ozone concentrations in different environmental settings around the Alpine Arc. Long Short-Term Memory (LSTM) neural network models are applied for predicting hourly concentrations in 20 stations of different types (urban, rural, and mountain). The analysis of the results allows us to determine the expected performance of such models in new cases and reduce the transferability uncertainty when the existing models can be partitioned into clusters. The LSTM models demonstrate the possibility of high accuracy in ozone forecasting at all sites. Given the significant impacts of this gas on human health and the environment, this can contribute to better decision-making and mitigation strategies for air pollution control.

Multi-objective sensor placement optimization in SHM systems with Kriging-based mode shape interpolation

Sensor location optimization plays a key role in the application and development of structural integrity monitoring methodologies, especially in large mechanical structures. Given the existence of an effective damage detection and identification procedure, the question of how many and where to place the acquisition points (sensors) so that the monitoring system operates at peak efficiency arises. In this study, an innovative methodology is proposed in order to maximize the quality of modal information and minimize the number of sensors in the SHM system. To maximize the quality of modal information, it considered the reconstruction of mode shapes using Kriging interpolation. The study was carried out on plate-type composite material structures for initial validation and later applied and validated on a main rotor blade of the AS-350 helicopter. The initial modal information (modal deformation) was obtained through the finite element method, and the multi-objective Lichtenberg algorithm was used in the complex optimization process. The proposed method presented in this work allows for the best possible distribution of a minimum and sufficient number of acquisition points in a structure in order to obtain more modal information for a better modal reconstruction from kriging interpolation of these minimum points. Numerical examples and test results show that the proposed method is robust and effective for distributing a reduced number of sensors in a structure and at the same time guaranteeing the quality of the information obtained, even in noisy situations. Numerical results considering both the simple and complex geometric cases show that the proposed combined FS-kriging method is effective in distributing a finite number of sensors on the structures and at the same time guaranteeing the quality of the modal information obtained. The results also indicate that the layout of sensors obtained by multi-objective optimization does not become trivial and symmetrical when a set of modes is considered in the objective function formulation. The proposed strategy is an advantage in modal testing as it is only necessary to acquire signals at a limited number of points, saving time and operating costs in vibration-based processes.

Impact of building types and CHP plants on air quality (2019–2021) in central-eastern European monocentric agglomeration

The quality of city air is influenced by many factors, including the density of buildings, the roughness of the terrain, the presence of street canyons, the heat sources in buildings, the types of industry, the topography, and meteorological conditions. Official air quality monitoring systems measure a very limited number of points, making local analysis impossible without the use of mathematical modeling programs. Here, we present an analysis of local air quality in an urban agglomeration. Data were collected over three years (2019, 2020, 2021), using commercial sensors located throughout the area of investigation. Dense downtown buildings equipped with individual heat sources were not found to have any impact on local air quality. The local municipal combined heat and power (CHP) plants contributed <1 ‰ of the measured concentration of particulate matter. Land height and the density of single-family housing were found to significantly affect air quality. We also took into account the influence of weather conditions, wind speed, and wind direction on the concentrations of particulate matter. High concentrations of particulate matter occurred only during heating periods when wind speeds were moderate. Wind direction did not have a direct impact on air quality, despite the expected benefits of ventilation through air corridors.

In situ U-value measurement of building envelopes through continuous low-cost monitoring

Accurate characterization of the transmittance parameter (U-value) is decisive when assessing energy demand or energy consumption of buildings. The U-value of building envelopes is usually estimated using commercial devices. These commercial solutions present several drawbacks such as their high cost, the fact that they require contact sensors to measure the surface temperature on a limited number of points and the difficulty to meet the standards for ideal measuring conditions. To solve these issues, a Hyper Efficient Arduino Transmittance-meter (HEAT) was recently proposed by the authors for monitoring the U-value based on low-cost technology. This system is now completed with the recording of time series of data based on Internet of Things protocols and the use of non-contact temperature sensors. Studying the applicability, in this case HEAT is used for characterizing walls and ceiling of a building room. To check its performance, the obtained results were compared with those derived from standards and conventional building modelling tools. The results show that the developed system is reliable according to the U-values obtained and expected for the monitored elements, with an acceptable range of uncertainty compared to the ones presented in the literature.

Directional Pareto Front and Its Estimation to Encourage Multi-Objective Decision-Making

This work introduces the following concepts of directional and estimated directional Pareto front to encourage multi-objective decision making, especially when the Pareto front exists in limited regions in the objective space. The general output of multi-objective optimization is a set of non-dominated solutions to approximate the Pareto front. When the Pareto front exists in limited regions, few solutions are obtained and presented to the decision maker. The limited output representing the objective value trade-off is a barrier to multi-objective decision making. The directional Pareto front introduced in this study is a superset of the Pareto front and supplements the objective value trade-off between the Pareto fronts. The estimated directional Pareto front is a response surface that represents the directional Pareto front using a limited number of points, which are objective vectors of the obtained solutions. The experimental results show that the directional Pareto front and the estimated front provide an objective value trade-off even in areas where the Pareto front does not exist and enhances the explanation of the objective space of the target problem.

Deep neural network adaptation to different environmental contexts: A case study of ozone forecast

Many environmental variables, in particular, related to air or water quality, are measured in a limited number of points and often for a limited time span. This forbids the development of accurate models for those locations due to an insufficient number of data and poses the question of whether a model developed for another measurement station can be reliably applied. Such a question is particularly critical when the model is constituted by a neural network, i.e., by an approach fully based on local data. In this context, the paper discusses the results of the application of a model to forecast ozone concentrations trained on stations with various characteristics in different environmental settings.

An Integrated InSAR and GNSS Approach to Monitor Land Subsidence in the Po River Delta (Italy)

Land subsidence affects many areas of the world, posing a serious threat to human structures and infrastructures. It can be effectively monitored using ground-based and remote sensing techniques, such as the Global Navigation Satellite System (GNSS) and Interferometric Synthetic Aperture Radar (InSAR). GNSS provides high precision measurements, but in a limited number of points, and is time-consuming, while InSAR allows one to obtain a very large number of measurement points, but only in areas characterized by a high and constant reflectivity of the signal. The aim of this work is to propose an approach to combine the two techniques, overcoming the limits of each of them. The approach was applied in the Po River Delta (PRD), an area located in Northern Italy and historically affected by land subsidence. Ground-based GNSS data from three continuous stations (CGNSS) and 46 non-permanent sites (NPS) measured in 2016, 2018, and 2020, and Sentinel-1 and COSMO-SkyMed SAR data acquired from 2016 to 2020, were considered. In the first phase of the method, InSAR processing was calibrated and verified through CGNSS measurements; subsequently, the calibrated interferometric data were used to validate the GNSS measurements of the NPS. In the second phase, the datasets were integrated to provide an efficient monitoring system, extracting high-resolution deformation maps. The results showed a good agreement between the different sources of data, a high correlation between the displacement rate and the age of the emerged surfaces composed of unconsolidated fine sediments, and high land subsidence rates along the coastal area (up to 16–18 mm/year), where the most recent deposits outcrop. The proposed approach makes it possible to overcome the disadvantages of each technique by providing more complete and detailed information for a better understanding of the ongoing phenomenon.

Assessing and projecting surface air temperature conditions required to sustain permafrost in Japan

Permafrost covers a wide area of the Northern Hemisphere, including high-altitude mountainous areas and even at mid-latitudes. There is concern that the thawing of mountain permafrost can cause slope instability and substantially impact alpine ecosystems, and because permafrost in mountainous areas is difficult to observe, detailed analyses have not been performed on its current distribution and future changes. Although previous studies have observed permafrost only at a limited number of points in Japan (e.g., Daisetsu Mountains, Mt. Fuji, and Mt. Tateyama in the Northern Japan Alps), we show that permafrost potentially exists in nine domains in Japan (Daisetsu Mountains, Mt. Fuji, Northern and Southern Japan Alps, Hidaka Mountains, Mt. Shiretokodake, Sharidake, Akandake, and Yotei). In the Daisetsu Mountains and Mt. Fuji, the environmental conditions required for maintaining at least some permafrost are projected to remain in the future if a decarbonized society is achieved (RCP2.6 or RCP4.5). However, if greenhouse gas emissions continue to increase (RCP8.5), the environmental conditions required for sustaining permafrost are projected to disappear in the second half of the twenty-first century. In other domains, the environmental conditions required for maintaining permafrost are either projected to disappear in the next ten years (Hidaka Mountains, Northern Japan Alps) or they have almost disappeared already (Southern Japan Alps, Mt. Shiretokodake, Sharidake, Akandake, and Yotei). Our projections show that climate change has a tremendous impact on Japan's mountain permafrost environment and suggests the importance of monitoring the mountain environment and considering measures for adapting to future climate change.

3D Lattice Boltzmann Method-Discrete-Element Method with Immersed Moving Boundary Scheme Numerical Modeling of Microparticles Migration Carried by a Fluid in Fracture

SummaryBased on the LBM-IMB-DEM (coupled lattice Boltzmann method-discrete-element method with immersed moving boundary scheme) model, the 3D microparticles’ migration carried by a fluid in fracture is numerically investigated both considering and not considering the van der Waals adhesive force. A boundary cells’ tracing method called the layer-by-layer tracing method both applicable to 2D and 3D situations for IMB with high efficiency and accuracy is proposed, and based on its searching results, the contacts in DEM could be detected straightforwardly with a finite amount of computation. In the layer-by-layer tracing method, a limited number of points, including some cell centers and nodes, near the particle boundary are determined whether they are covered by the particle, and then the other cells in one layer could be precisely classified. In the contact detecting process for DEM, based on the boundary cells tracing results in IMB, the cells that are covered by no less than two particles are the potential cells where there may be contacts. For wall boundaries with irregular shapes, the wall solid boundary (WSB) cells are converted to circumcircle (or circumscribed sphere in 3D situation), and then the solid particle-wall interactions could be replaced by the interactions between the solid particles and those circumcircles or circumscribed spheres. Two cases, including single-particle sedimentation in a viscous fluid, two-particle Drafting-Kissing-Tumbling (DKT) simulation for both nonadhesive and adhesive particles, are used to validate the LBM-IMB-DEM method. Besides, multiparticle sedimentation tests for different particle radiuses are conducted to present the advantages of the layer-by-layer tracing method. At last, nonadhesive and adhesive microparticles’ liberation, transport, and retention carried by a fluid in fracture with irregular shapes are simulated. The numerical results show that the adhesive force that plays a dominated role for microparticles has significant effects on the mechanics of solid particles migration. Under the influence of adhesive forces, the microparticles tend to form a stable agglomerate and migrate as a whole, which is different from the situation for nonadhesive particles where they are relatively independent and basically migrate by layer. Besides, even if the fluid velocity is much lower than the critical velocity evaluated through force or torque analysis on a single particle, when the total hydrodynamic forces exerted on the agglomerate overcome the total adhesive force between the wall and the lower layer of particles, the agglomerate can start to move and liberate from the wall.

Robust Suspicious Action Recognition Approach Using Pose Descriptor

In the current era of technological development, human actions can be recorded in public places like airports, shopping malls, and educational institutes, etc., to monitor suspicious activities like terrorism, fighting, theft, and vandalism. Surveillance videos contain adequate visual and motion information for events that occur within a camera’s view. Our study focuses on the concept that actions are a sequence of moving body parts. In this paper, a new descriptor is proposed that formulates human poses and tracks the relative motion of human body parts along with the video frames, and extracts the position and orientation of body parts. We used Part Affinity Fields (PAFs) to acquire the associated body parts of the people present in the frame. The architecture jointly learns the body parts and their associations with other body parts in a sequential process, such that a pose can be formulated step by step. We can obtain the complete pose with a limited number of points as it moves along the video and we can conclude with a defined action. Later, these feature points are classified with a Support Vector Machine (SVM). The proposed work was evaluated on the benchmark datasets, namely, UT-interaction, UCF11, CASIA, and HCA datasets. Our proposed scheme was evaluated on the aforementioned datasets, which contained criminal/suspicious actions, such as kick, punch, push, gun shooting, and sword-fighting, and achieved an accuracy of 96.4% on UT-interaction, 99% on UCF11, 98% on CASIA and 88.72% on HCA.

Post-buckling path and free vibration of a symmetric laminated plate vertically coupled with fluid under in-plane load

In this paper, the effect of compressive in-plane load on the equilibrium paths and vibration characteristics of a vertical symmetric laminated plate coupled with fluid is studied theoretically both in the pre- and post-buckling state. The Third-order shear deformation theory and von Karman nonlinear strain displacement relationship are utilized to describe the nonlinear deformation of the plate consists of a number of fiber-reinforced layers. The Hamilton's principle is introduced to formulate the plate-fluid coupled governing equations. A two-step theoretical approach is developed. Firstly, based on the Cylindrical arc-length method, we solve the nonlinear static equations to trace the snap-back behavior and pass the limit points along the complex equilibrium paths obtaining the stable buckling deflection. Then the coupled vibration characteristics are calculated using the tangent stiffness at the stable positions. The present theoretical results are validated by the numerical results. The effects of added mass and hydrostatic load are compared and analyzed. It is noted that, in the post-buckling state, the number of limit points along the static unstable equilibrium path increases with the compressive in-plane load getting larger. The variation trends of the natural frequencies and modes under different in-plane loads in pre- and post-buckling state are investigated.

Gradient-free Stein variational gradient descent with kernel approximation

Stein variational gradient descent (SVGD) has been shown powerful as a nonparametric variational inference algorithm for general purposes. However, the standard SVGD necessitates calculating the gradient of the target density and therefore cannot be used where the gradient is unavailable or too expensive to evaluate. A gradient-free variant (GF-SVGD) has been proposed to substitute the gradient with a surrogate, however, the major computational challenge of computing the forward model still prohibits the use of SVGD in inferencing complex distributions. In this paper, we address this issue by evaluating the forward model at only a limited number of points and build its approximation using pre-calculated kernels to keep the computational cost as low as possible. Our approximation method is then combined with an adaptation strategy that automatically refines the model by selecting particles at critical local locations, increasing precision at a low cost. We observe significant computational gains over the original SVGD and GF-SVGD algorithms.

An approximate solution of Riccati’s differential equation using fuzzy linguistic model

Looking for offer, an approximate solution of the Riccati’s differential equations in this research work is based on a fuzzy linguistic model (FLM). It is shown by the FLM that it is possible to approximate continuous functions which are satisfied at the interpolation points. To obtain the function as the approximation of the solution of the RDEs, fuzzy linguistic model (FLM) was used. In numerical approaches, in order to find an approximation for a function that exists on a particular interval, just a limited number of points are utilized. However, the distinctive feature of the FLM in comparison with other approaches is that in this method all points in the arbitrary interval are used. Five examples along with their approximate solutions have been presented to show the suitability, convenience, simplicity and performance of the method. The outcomes demonstrate that the advantage of this approach compared to other methods is that there are no limiting conditions for nonlinear differential equations of order one or two with initial conditions or boundary conditions.

Analytical Prediction for Nonlinear Buckling of Elastically Supported FG-GPLRC Arches under a Central Point Load.

In this paper, we present an analytical prediction for nonlinear buckling of elastically supported functionally graded graphene platelet reinforced composite (FG-GPLRC) arches with asymmetrically distributed graphene platelets (GPLs). The effective material properties of the FG-GPLRC arch are formulated by the modified Halpin–Tsai micromechanical model. By using the principle of virtual work, analytical solutions are derived for the limit point buckling and bifurcation buckling of the FG-GPLRC arch subjected to a central point load (CPL). Subsequently, the buckling mode switching phenomenon of the FG-GPLRC arch is presented and discussed. We found that the buckling modes of the FG-GPLRC arch are governed by the GPL distribution pattern, rotational restraint stiffness, and arch geometry. In addition, the number of limit points in the nonlinear equilibrium path of the FG-GPLRC arch under a CPL can be determined according to the bounds of successive inflexion points. The effects of GPL distribution patterns, weight fractions, and geometric configurations on the nonlinear buckling behavior of elastically supported FG-GPLRC arches are also comprehensively discussed.

Guided Waves Measurement Techniques in Pitch–Catch Configuration

Abstract Guided waves (GWs) testing allows a long-range screening in pipes of different types and represents an effective and powerful nondestructing technique for defect detections using a limited number of points of measures. After the characterization through a general theoretical analysis, the focus is set to a real steam discharge pipe with a high mechanical complexity used for many years in a research plant now decommissioned. The experimental method applied here is the pitch–catch configuration of two magnetostrictive sensors. This paper aims to establish a strong theoretical background to pave the way for a robust experimental investigation. Preliminary experimental results are consistent with the theoretical analysis.

Human body reconstruction from limited number of points

AbstractWe propose a novel approach for reconstructing plausible three‐dimensional (3D) human body models from small number of 3D points which represent body parts. We leverage a database of 3D models of humans varying from each other by physical attributes such as age, gender, weight, and height. First we divide the bodies in database into seven semantic regions. Then, for each input region consisting of maximum 40 points, we search the database for the best matching body part. For the matching criterion, we use the distance between novel point‐based features of input points and body parts in the database. We then combine the matched parts from different bodies into one body, with the help of Laplacian deformation, which results in a plausible human body. To evaluate our results objectively, we pick points from each part of the ground‐truth human body models, then reconstruct them using our method and compare the resulting bodies with the corresponding ground‐truths. Also, our results are compared with registration‐based results. In addition, we run our algorithm with noisy data to test the robustness of our method and run it with input points whose body parts are manually edited, which produces plausible human bodies that do not even exist in our database. Our experiments verify qualitatively and quantitatively that the proposed approach reconstructs human bodies with different physical attributes from a small number of points using a small database.

Probabilistic learning on manifolds constrained by nonlinear partial differential equations for small datasets

A novel extension of the Probabilistic Learning on Manifolds (PLoM) is presented. It makes it possible to synthesize solutions to a wide range of nonlinear stochastic boundary value problems described by partial differential equations (PDEs) for which a stochastic computational model (SCM) is available and which depend on a vector-valued random control parameter. The cost of a single numerical evaluation of this SCM is assumed to be such that only a limited number of points can be computed for constructing the training dataset (small data). Each point of the training dataset is made up of realizations from a vector-valued stochastic process (the stochastic solution) and the associated random control parameter on which it depends. The presented PLoM constrained by PDE allows for generating a large number of learned realizations of the stochastic process and its corresponding random control parameter. These learned realizations are generated so as to minimize the vector-valued random residual of the PDE in the mean-square sense. Appropriate novel methods are developed to solve this challenging problem. Three applications are presented. The first one is a simple uncertain nonlinear dynamical system with a nonstationary stochastic excitation. The second one concerns the 2D nonlinear unsteady Navier–Stokes equations for incompressible flows in which the Reynolds number is the random control parameter. The last one deals with the nonlinear dynamics of a 3D elastic structure with uncertainties. The results obtained make it possible to validate the PLoM constrained by stochastic PDE but also provide further validation of the PLoM without constraint.

Fast multi-feature image segmentation

Multi-feature segmentation has demonstrated its superiority against one-dimensional feature approaches based on only grayscale information. Mean shift (MS) is an algorithm that has been used commonly for multi-feature segmentation. In spite of its interesting results, MS maintains a computational cost that is prohibitive for segmentation scenarios where the feature map consists of multi-dimensional features. In this paper, a new competitive segmentation algorithm for grayscale images is introduced. The proposed approach considers a two-dimensional feature map that includes the grayscale value and the local variance for each pixel in the image. To reduce the computational cost, the Mean shift (MS) algorithm is modified to operate with a very limited number of points from all available data. Under such conditions, two sets of elements are differentiated: involved data (the reduced dataset considered in the MS operation) and not involved data (the rest of the available data). Different from the classical MS, which employs Gaussian functions, in our approach, the process of estimating the feature map is carried out using a more accurate approach such as the Epanechnikov kernel function. Once the MS results are obtained, they are generalized to include the not involved data. Therefore, each unused element is assigned to the same cluster of the closest used data. Finally, clusters with the fewest elements are fused with other neighboring clusters. The proposed segmentation method has been compared with other state-of-art algorithms considering the full number of images from the Berkeley dataset. Experimental results confirm that the proposed scheme produces segmented images with a 50% better quality of visual perception approximately two times (≈ 1.8 − 2) faster than its competitors.

Non-Destructive Testing of the Longest Span Soil-Steel Bridge in Europe-Field Measurements and FEM Calculations.

The article describes interdisciplinary and comprehensive non-destructive diagnostic tests of final bridge inspection and acceptance proposed for a soil-steel bridge made of corrugated sheets, being the European span length record holder (25.74 m). As an effect of an original concept a detailed and precise information about the structure short-term response was collected. Periodic diagnostics of bridge deformations was done one year after it was built. Load test design was based on numerical simulations performed by means of finite element method (FEM). In situ measurements were done with the aid of: inductive sensors, optical total station, and terrestrial laser scanner. The results produced by terrestrial laser scanning were used to build a precise image of structure deformation in 3D space during the tests. The accuracy of laser mapping was significantly increased using the information coming from total station and inductive sensors. These have higher accuracy and therefore can be used as reference. Thus, new quality in measurements is introduced. Good correspondence between in situ values and FEM estimations was achieved. Therefore, such a combination of testing methods can be used in non-destructive diagnostics of structures and is an interesting alternative for the standard approach, in which the measurements are done in limited number of points.

How much Influences Image Quality the Form Error in Industrial X-Ray CT?

For dimensional metrology using X-ray computed tomography (CT), the surface of an object is of importance. Measurements are performed on it, for example, to evaluate the distance between two points, or the specific form. In CT, the determination process of the surface is performed on voxel data. This voxel data is afflicted by noise and sharpness of the system. While sphere centres are computed using a huge number of voxels, the elements used to establish the surface are limited. Hence, noise becomes more of an issue for evaluating forms, or measuring between just a limited number of points. As the surface is determined by following the edge, it is afflicted by sharpness, too: the less sharp the CT system is, the more the curve from the material to the outside diverges from an ideal edge, the more voxels are within the range of the critical values to determine the surface, and therefore the influence of noise on the surface is increased. In this paper, measurements and simulations are performed to determine the relationship of those two measures in order to get closer to a prediction of measurement uncertainties.