Uncertain Feature Research Articles

Immersion-and-Invariance Fuzzy Adaptive Image-Based Visual Servoing of Omnidirectional Mobile Robots Considering Uncertain Feature Point Dynamics

Immersion-and-Invariance Fuzzy Adaptive Image-Based Visual Servoing of Omnidirectional Mobile Robots Considering Uncertain Feature Point Dynamics

Uncertainty-Aware Denoising Network for Artifact Removal in EEG Signals.

The electroencephalogram (EEG) is extensively employed for detecting various brain electrical activities. Nonetheless, EEG recordings are susceptible to undesirable artifacts, resulting in misleading data analysis and even significantly impacting the interpretation of results. While previous efforts to mitigate or reduce the impact of artifacts have achieved commendable performance, several challenges in this domain still persist: 1) due to black-box skepticism, deep-learning-based automatic EEG artifact removal methods have been impeded from being applied in clinical environments. How to support reliable denoised EEG signals with high accuracy is important; and 2) effectively exploring valuable local and global information from contaminated contexts remains challenging. On the one hand, feature extraction and aggregation in prior works are often performed blindly and assumed to be accurate, which is not always the case. On the other hand, global contextual information is gradually modeled by local fixed single-scaled convolutional filters layer by layer, which is neither efficient nor effective. To address the above challenges, we propose an Uncertainty-aware Denoising Network (UDNet) with multi-scaled pooling attention for efficient context capturing. Specifically, we predict the aleatoric and epistemic uncertainty existing during the denoising process to assist in finding and reducing the uncertain feature representation. We further propose a simple yet effective architecture to capture local and global contexts at multiple scales. Our proposed method can serve as an effective metric for identifying low-confidence epochs that warrant deferral to human experts for further inspection and assessment. Experimental results on two public datasets show that the proposed model outperforms state-of-the-art baselines.

Controls on Subtropical Cloud Reflectivity during a Waterbelt Scenario for the Cryogenian Glaciations

Abstract Waterbelt climate states with an ice-free tropical ocean provide a straightforward explanation for the survival of advanced marine species during the Cryogenian glaciations (720–635 million years ago). Previous work revealed that stable waterbelt states require the presence of highly reflective low-level mixed-phase clouds with a high abundance of supercooled liquid in the subtropics. However, the high uncertainty associated with representing mixed-phase clouds in coarse-scale general circulation models (GCMs) that parameterize atmospheric convection has prohibited assessment of whether waterbelt states are a robust feature of Earth’s climate. Here we investigate whether resolving convective-scale motion at length scales of hectometers helps us to assess the plausibility of a waterbelt scenario. First, we show that substantial differences in cloud reflectivity among GCMs do not arise from the resolved atmospheric circulation. Second, we conduct a hierarchy of simulations using the Icosahedral Nonhydrostatic (ICON) modeling framework, ranging from coarse-scale GCM simulations with parameterized convection to large-eddy simulations that explicitly resolve atmospheric convective-scale motions. Our hierarchy of simulations supports the existence of highly reflective subtropical clouds if we apply moderate ice nucleating particle (INP) concentrations. Third, we test the sensitivity of cloud reflectivity to the INP concentration. In the presence of high but justifiable INP concentrations, cloud reflectivity is strongly reduced. Hence, the existence of stable waterbelt states is controlled by the abundance of INPs. We conclude that explicitly resolving convection can help to constrain Cryogenian cloud reflectivity, but limited knowledge concerning Cryogenian aerosol conditions hampers strong constraints. Thus, waterbelt states remain an uncertain feature of Earth’s climate. Significance Statement The purpose of this study is to assess the impact of atmospheric convection and small airborne ice nucleating particles on the reflectivity of mixed-phase clouds over a subtropical ice margin. This is important as these clouds can determine whether the Cryogenian Earth (720–635 million years ago) was in a hard “snowball” state with a fully ice-covered ocean or a habitable waterbelt state with an ice-free tropical ocean. Our results indicate a clear impact of convection but neither confirm nor deny the existence of a waterbelt state since cloud reflectivity depends critically on the abundance of ice nucleating particles. Therefore, a Cryogenian waterbelt scenario remains uncertain, which calls for more comprehensive Earth system modeling approaches in future studies.

A spatiotemporal grammar network model with wide attention for short-term traffic flow prediction

Short-term traffic flow prediction is of great significance in intelligent transportation. In recent years, with the development of information collection technology and deep learning algorithms, neural network models have become increasingly popular in traffic flow prediction research. However, accurate and fast prediction is a challenge because of the uncertain feature of traffic flow and limitations of the model structure. Motivated by this issue, this paper uses a dual-branch grammar model to extract the deep spatio-temporal features of historical traffic information. Each branch combines the grammar structure with the gated convolution operation to realize the interaction between the implicit features of different traffic parameters. Moreover, scaled exponential linear units (Selu) are used as an activation function for gated convolution operation to enhance the convergence effect of network training. And then, a wide attention module is designed to weigh the extracted deep spatio-temporal features to increase the model's prediction accuracy with a slight increase in computational cost. Finally, actual traffic data from Caltrans Performance Measurement System (PeMS) is used to evaluate the prediction performance with the result that the proposed prediction method outperforms other methods in terms of prediction accuracy. In addition, this paper proves the Selu function's importance by analysing the training error's convergence effect and explains the role of wide attention in the prediction task through visualization and statistical analysis operations.

Hierarchical Disentangling Network for Building Extraction from Very High Resolution Optical Remote Sensing Imagery

Building extraction using very high resolution (VHR) optical remote sensing imagery is an essential interpretation task that impacts human life. However, buildings in different environments exhibit various scales, complicated spatial distributions, and different imaging conditions. Additionally, with the spatial resolution of images increasing, there are diverse interior details and redundant context information present in building and background areas. Thus, the above-mentioned situations would create large intra-class variances and poor inter-class discrimination, leading to uncertain feature descriptions for building extraction, which would result in over- or under-extraction phenomena. In this article, a novel hierarchical disentangling network with an encoder–decoder architecture called HDNet is proposed to consider both the stable and uncertain feature description in a convolution neural network (CNN). Next, a hierarchical disentangling strategy is set up to individually generate strong and weak semantic zones using a newly designed feature disentangling module (FDM). Here, the strong and weak semantic zones set up the stable and uncertain description individually to determine a more stable semantic main body and uncertain semantic boundary of buildings. Next, a dual-stream semantic feature description is built to gradually integrate strong and weak semantic zones by the designed component feature fusion module (CFFM), which is able to generate a powerful semantic description for more complete and refined building extraction. Finally, extensive experiments are carried out on three published datasets (i.e., WHU satellite, WHU aerial, and INRIA), and the comparison results show that the proposed HDNet outperforms other state-of-the-art (SOTA) methods.

Forecasting Carbon Dioxide Emissions for Singapore using Grey Model with Cramer’s Rule

This article analyses and forecasts carbon dioxide () emissions in Singapore for the 2012 to 2016 period. The study analysed the data using grey forecasting model with Cramer’s rule to calculate the best SOGM(2,1) model with the highest accuracy of precision compared to conventional grey forecasting model. According to the forecasted result, the fitted values using SOGM(2,1) model has a higher accuracy precision with better capability in handling information to fit larger scale of uncertain feature compared to other conventional grey forecasting models. This article offers insightful information to policymakers in Singapore to develop better renewable energy instruments to combat the greater issues of global warming and reducing the fossil carbon dioxide emissions into the environment.

Optimizing the Segmentation of a High-Resolution Image by Using a Local Scale Parameter

Image segmentation is a critical procedure in object-based identification and classification of remote sensing data. However, optimal scale-parameter selection presents a challenge, given the presence of complex landscapes and uncertain feature changes. This study proposes a local optimal segmentation approach that considers both intersegment heterogeneity and intrasegment homogeneity, uses the standard deviation and local Moran's index to explore each optimal segment across different scale parameters, and combines the optimal segments into a single layer. The optimal segment is measured by using high-spatial-resolution images. Results show that our approach out-performs and generates less error than the global optimal segmentation approach. The variety of land cover types or intrasegment homogeneity leads to segment matching with the geo-objects on different scales. Local optimal segmentation demonstrates sensitivity to land cover discrepancy and provides good performance on cross-scale segmentation.

Bayesian network for dynamic variable structure learning and transfer modeling of probabilistic soft sensor

Bayesian network is a frequently-used uncertainty reasoning method, which systematically describes relations between random variables. Dynamic Bayesian network is an extension of Bayesian network, which contains the relations between variables at different times. Soft sensor is an important industrial application, in which feature variables are selected to predict the value of the target variables. For industrial soft sensor applications, dynamics is still a tough problem, particularly together with the uncertain feature of process data. In this article, DBN is employed for dynamic variable structure learning and transfer modeling to some strong regression models to build a soft sensor for dynamic industrial processes. At the beginning, a series of dynamic Bayesian networks are constructed on the training set with a sliding window. From these network structures we can find variables related to the quality variables. Then, in each time period, a sequence of data is compared with training data to select the most similar sequence by Dynamic Time Warping. Therefore, the structure of variables is built, i.e., the related feature variables of the quality variables can be determined by the network structure. In the regression step, the dynamic variable structure is transferred to some strong regressors, like Support Vector Regression and Adaboost for further regression. In case study, we use the debutanizer process and a low temperature transformer case to confirm the quality of the soft sensor method. The result reveals that, the prediction accuracy of the new method is much higher than the original commonly-used regression methods.

Path Tracking and Direct Yaw Moment Coordinated Control Based on Robust MPC With the Finite Time Horizon for Autonomous Independent-Drive Vehicles

It is a striking fact that the characteristics of parametric uncertainties, external disturbance, time-varying and nonlinearities are available in the constructed model of autonomous independent-drive vehicles; therefore, in this paper, the robust model predictive control (MPC) with the finite time horizon is proposed to realize the coordinated path tracking and direct yaw moment control (DYC) for autonomous four in-wheel motor independent-drive electric vehicles (AMIDEV). Firstly, considering the time-varying and uncertain feature of the tire cornering stiffness and the vehicle velocity in the state space equation constructed by 2 degrees of freedom (DoF) vehicle model and the path tracking preview model, the linear parameter varying (LPV) discrete model with four polytypic vertexes is constructed. Then, based on the linear matrix inequality (LMI) method, the novel robust MPC theory with the finite time horizon is put forward to solve the min-max optimization problem after updating four polytypic vertexes in real time, which could deal with the inevitable model mismatch problem caused by the time-varying, uncertain vehicle dynamic characteristics and external disturbance. Finally, the simulation and experimental results have verified that the proposed novel robust MPC theory could emerge from the stranglehold exercised by the conservativeness of the traditional robust MPC theory with the infinite time horizon, which strengthens the robustness of this control system as well as achieves better path tracking accuracy and handling ability of AMIDEV.

A robust multi-objective model for supporting agricultural water management with uncertain preferences

Sustainable management of agricultural water resources is essential for promoting regional development and restoring ecological environment. Due to the imperfection of knowledge and imprecision of expression, the preferences of objectives are uncertain in nature. However, the uncertain feature of objective preferences is often neglected in solving multi-objective water management problems. Also, random information is prevalent in in the optimizaiton efforts of agricultural water management systems. Recognizing the needs to tackle uncertainty existing in both preferences over objectives and parameters of constraints, a new mathematical programming method named multi-objective chance-constrained programming approach for planning problems with uncertain weights (MCUW) was proposed. It could handle uncertain weights of objectives without known distributions, and quantify the risks of objective unattainability arising from such uncertainty. It could also deal with uncertain parameters with known probability distributions, generating solutions with varied risks of constraint violation. The proposed MCUW method was applied to a case study of agricultural water resources management problem in Northwest China to demonstrate its applicability. Multiple sets of optimal solutions under different combinations of weights fluctuation ranges, protection levels, and surface water availabilities were obtained, providing management options for stakeholders with different risk appetites. Results indicate that the objective value would increase with higher risks of objective unattainability or water-shortage. The MCUW method was compared to two potential alternatives and deemed effective in balancing multiple objectives and tackling complex uncertainties. Monte Carlo simulation showed that the results of MCUW were more densely distributed than those obtained from the model with deterministic weights, verifying that the developed MCUW model could provide robust solutions when faced with uncertain weights.

Design And Implementation Of Ls-Pmsg For Small Scale Hydro Power Plant

- In conventional power system, synchronous generator which characterized by high-speed operating condition has been widely implemented. However, those high-speed synchronous machines are not suitable for a power plant powered by renewable energy (RE) due to uncertain feature of renewable resources. To overcome this problem, an electrical machine with low-speed characteristic is required for ensuring stable operation and maintaining output power of the RE based distribution generation (DG) unit. In this paper, a low-speed permanent magnet synchronous generator (LS-PMSG) is designed and implemented. To realize low-speed operation capability, the multi-stages permanent magnet synhcronous machine is proposed. The multi-stages machine equipped with two stators and three rotors construction. From the experimental results, it was monitored that output voltage of single rotor of the designed machine was 35 V. Moreover, to increase the output voltage to 50 V, those three rotors can be connected in series.

Inexact fuzzy chance-constrained programming for community-scale urban stormwater management

Abstract Due to frequent and serious waterlog and environmental pollution in cities in recent years, urban stormwater management (USM) has become an essential issue of urban sustainable development in China. Low Impact Development (LID) technologies are effective and popular measures of USM, and can be used to reduce urban waterlog and the associated environmental pollution. In this paper, in order to identify an optimal strategy of LID technologies based on economic efficiencies and environmental performances, the construction costs and environmental benefits of the four LID technologies (including grass swales, bioretention cells, green roofs, and permeable pavement) were analyzed. Then, due to the uncertain feature of rainfall, concentration of pollutants and many social-economic factors, an inexact fuzzy chance-constrained programming (IFCCP) model was developed. Multiple uncertainties that can be expressed as interval parameters, fuzzy sets, and stochastic distribution can be addressed effectively and incorporated directly into the modeling process. Based on the analysis of the precipitation probability distribution and land use data, the developed model was applied to a university campus with relatively high building density and plot ratio in Beijing. Then, optimal solutions of multiple precipitation probabilities ( p m ) under varying USM goals were obtained. The four LID technologies were chosen to meet multiple USM goals based on multiple p m values. When p m = 0.1 , 0.05 , 0.02 and 0.015 , the total construction investments would be 1.46 to 4.66, 1.46 to 5.25, 2.31 to 8.54 and 3.95 to 12.7 million yuan, respectively. The relationship between construction costs and p m represented the relationship between economic benefit and system risk. The results indicated that USM was influenced by the rainfall capacity, construction costs would increase with the increase of precipitation, and the risks of constraint violation would decrease as the construction costs increase. When considering the four LID technologies during the planning stage of USM, the preferred selection order would be bioretention cells, permeable pavements, grass swales and green roofs. Therefore, the solutions under each p m level could provide the references for desired USM plans.

Uncertain Feature RCS Analyses based on Wavelet Decomposition

A novel uncertainty analysis approach for radar cross section (RCS) data is proposed in this paper. The wavelet decomposition method is applied to separate the origin data for uncertain feature extraction. Methods of distribution fitting and analysis are studied and data uncertainty distribution characteristics are also compared. Several benchmark cases from RCS simulation are taken to validate the proposed method. The impact of attitude uncertainty on the target RCS is also studied.

Optimal trading of plug-in electric vehicle aggregation agents in a market environment for sustainability

Ever since energy sustainability is an emergent concern, Plug-in Electric Vehicles (PEVs) significantly affect the approaching smart grids. Indeed, Demand Response (DR) brings a positive effect on the uncertainties of renewable energy sources, improving market efficiency and enhancing system reliability. This paper proposes a multi-stage stochastic model of a PEV aggregation agent to participate in day-ahead and intraday electricity markets. The stochastic model reflects several uncertainties such as the behaviour of PEV owners, electricity market prices, and activated quantity of reserve by the system operator. For this purpose, appropriate scenarios are utilized to realize the uncertain feature of the problem. Furthermore, in the proposed model, the PEV aggregation agents can update their bids/offers by taking part in the intraday market. To this end, these aggregation agents take part in Demand Response eXchange (DRX) markets designed in the intraday session by employing DR resources. The numerical results show that DR provides a perfect opportunity for PEV aggregation agents to increase the profit. In addition, the results reveal that the PEV aggregation agent not only can increase its profit by participating in the DRX market, but also can become an important player in the mentioned market.

Efficient clustering of uncertain data streams

Clustering uncertain data streams has recently become one of the most challenging tasks in data management because of the strict space and time requirements of processing tuples arriving at high speed and the difficulty that arises from handling uncertain data. The prior work on clustering data streams focuses on devising complicated synopsis data structures to summarize data streams into a small number of micro-clusters so that important statistics can be computed conveniently, such as Clustering Feature (CF) (Zhang et al. in Proceedings of ACM SIGMOD, pp 103---114, 1996) for deterministic data and Error-based Clustering Feature (ECF) (Aggarwal and Yu in Proceedings of ICDE, 2008) for uncertain data. However, ECF can only handle attribute-level uncertainty, while existential uncertainty, the other kind of uncertainty, has not been addressed yet. In this paper, we propose a novel data structure, Uncertain Feature (UF), to summarize data streams with both kinds of uncertainties: UF is space-efficient, has additive and subtractive properties, and can compute complicated statistics easily. Our first attempt aims at enhancing the previous streaming approaches to handle the sliding-window model by using UF instead of old synopses, inclusive of CluStream (Aggarwal et al. in Proceedings of VLDB, 2003) and UMicro (Aggarwal and Yu in Proceedings of ICDE, 2008). We show that such methods cannot achieve high efficiency. Our second attempt aims at devising a novel algorithm, cluUS , to handle the sliding-window model by using UF structure. Detailed analysis and thorough experimental reports on synthetic and real data sets confirm the advantages of our proposed method.

Sparse group LASSO based uncertain feature selection

Uncertain data management and mining is becoming a hot topic in recent years. However, little attention has been paid to uncertain feature selection so far. In this paper, we introduce the sparse group least absolution shrinkage and selection operator (LASSO) technique to construct a feature selection algorithm for uncertain data. Each uncertain feature is represented with a probability density function. We take each feature as a group of values. Through analysis of the current four sparse feature selection methods, LASSO, elastic net, group LASSO and sparse group LASSO, the sparse group LASSO is introduced to select feature selection from uncertain data. The proposed algorithm can select not only the features between groups, but also the sub-features in groups. As the trained weights of feature groups are sparse, the groups of features with weight zero are removed. Experiments on nine UCI datasets show that feature selection for uncertain data can reduce the number of features and sub-features at the same time. Moreover it can produce comparable accuracy with all features.

Method of Collaborative Filtering Based on Uncertain User Interests Cluster

Recommender systems have been proven to be valuable means for Web online users to cope with the information overload and have become one of the most powerful and popular tools in electronic commerce. The suggestions provided are aimed at supporting their users in various decision-making processes, such as what items to buy, what music to listen, or what news to read. In the paper, we introduce the uncertain interests of users because computer logs take down the data that have uncertain features. Consequently, user interest data recorded in computer logs is uncertain data. First of all the definition of uncertain interests are put forward. Then, some method using a clustering algorithm can solve the uncertain feature. In the algorithm, we compute the between-class entropy of any two clusters and get the stable classes. Secondly, because calculation input is uncertainty, the results of clustering algorithm are latent non-determinacy. The trustworthy degree of uncertain interests is defined that it can measure the rationality of clustering algorithm results. Thirdly, improvement of collaborative filtering is presented as an advanced CF algorithm based on trustworthy degree of uncertain interests. At last, we provide some evaluation of the algorithms and propose the more improve ideas in the future.

Analysis of LWR model with fundamental diagram subject to uncertainties

The LWR model is of interest since it is simple and can successfully reproduce some essential features of traffic flow, such as the formation and propagation of traffic disturbances. In this article, we investigate the LWR model from an uncertainty perspective. We attempt to analyse how reliable the LWR model prediction will be if the fundamental diagram (FD) in use is not accurately specified. To fulfil this end, we postulate a flux function (equivalently a FD) driven by a random free flow speed, which accommodates the uncertain feature observed in the speed–density data. We provide essential mathematical properties and solution schemes of the LWR model with the probabilistic FD. In case studies, the approach to evaluate the uncertainty of traffic disturbance propagation with this model is presented. We find that if FD in a LWR model cannot be perfectly specified, the uncertainty associated with the location of a traffic disturbance would increase over time. In contrast, the magnitude of the traffic disturbance can still be accurately predicted.

Evaluation model and algorithm of product disassembly process with stochastic feature

Disassembly planning is considered as the optimization of disassembly sequences with the target of the shortest disassembly time, the lowest disassembly cost, and the minimum disassembly energy consumption. However, obsolete products suffer from the influence of a variety of uncertainties, the disassembly process of products has the strong uncertain feature. Traditionally, to account for this uncertainty, each removal operation or removal task is assumed to be an activity or event with certain probability, and the determination of the optimal path of a disassembly process is merely a probabilistic planning problem based on this assumption. In this article, based on the established stochastic disassembly network graph, combined with different disassembly decision-making criterion, typical stochastic models for disassembly time analysis are developed. In addition, a two-phase approach is proposed to solve the typical stochastic models. Initially, according to different removal probability density functions, disassembly probability density functions of feasible disassembly paths are determined by a time-domain method or frequency-domain method, and additionally, after the disassembly probability density functions have been obtained, the quantitative evaluation of a product disassembly process and stochastic optimization of feasible disassembly paths are realized by a numerical solution method. Finally, a numerical example is illustrated to test the proposed concepts and the effectiveness of the proposed approach.

A Random-Finite-Set Approach to Bayesian SLAM

This paper proposes an integrated Bayesian frame work for feature-based simultaneous localization and map building (SLAM) in the general case of uncertain feature number and data association. By modeling the measurements and feature map as random finite sets (RFSs), a formulation of the feature-based SLAM problem is presented that jointly estimates the number and location of the features, as well as the vehicle trajectory. More concisely, the joint posterior distribution of the set-valued map and vehicle trajectory is propagated forward in time as measurements arrive, thereby incorporating both data association and feature management into a single recursion. Furthermore, the Bayes optimality of the proposed approach is established. A first-order solution, which is coined as the probability hypothesis density (PHD) SLAM filter, is derived, which jointly propagates the posterior PHD of the map and the posterior distribution of the vehicle trajectory. A Rao-Blackwellized (RB) implementation of the PHD-SLAM filter is proposed based on the Gaussian-mixture PHD filter (for the map) and a particle filter (for the vehicle trajectory). Simulated and experimental results demonstrate the merits of the proposed approach, particularly in situations of high clutter and data association ambiguity.