Search

The thermal behavior of the motorized spindle is a key issue that restricts the accuracy and efficiency of machining centers. Spindle thermal error modeling and compensation methods usually predict thermal errors based on temperature sensors on the spindle. However, the undetectable temperature region inside the spindle is an important source of thermal deformation, which leads to the lack of sufficient robustness of existing thermal error models. To improve the robustness of real-time prediction of spindle thermal error, an enhanced modeling method based on a mixture of the mechanism model and machine learning is proposed. First, the thermal network of the spindle is developed to predict the transient temperature field by determining the parameters through finite element (FE) simulation and thermal behavior experiments. Then, a data-fusion approach of the predicted temperature field and the measured data was established to enhance thermal error models by providing more thermal characteristic information inside the spindle to the machine learning process. Cross-validation shows that this method is universal to various types of machine learning modeling methods, including GRU, LSTM, LSSVM, BP, and MLR. Compared with the traditional method of using only basic sensors, this proposed method greatly improves the accuracy and robustness at the same hardware cost. The root mean square error (RMSE) decreased by 17–59%, and the fluctuation range decreased by 38–60%. Compared with the traditional method of attaching additional temperature sensors on the spindle, it saves 85.7% of the sensor cost and reduces the average RMSE by −0.27–27% and the fluctuation range by 10–40%. Finally, a digital twin system with a physical-edge-cloud layer structure is established for real-time prediction of spindle thermal behavior based on the best cost–benefit configuration, the enhanced LSTM model with basic sensors. It improves the accuracy and robustness of thermal error prediction results with lower sensor cost and can monitor the temperature changes inside the spindle in real-time, providing the possibility for potential industrial applications of intelligent spindles. [ABSTRACT FROM AUTHOR]

Following triple La Niña events during 2020–2022, the future evolution of climate conditions over the tropical Pacific has been a focused interest in ENSO-related communities. Observations and modeling studies indicate that an El Niño event is occurring in 2023; however, large uncertainties remain in terms of its detailed evolution, and the factors affecting its resultant amplitude remain to be understood. Here, a novel deep learning-based Transformer model is adopted to make real-time predictions for the 2023–2024 climate conditions in the tropical Pacific. Several key fields vital to the El Niño and Southern Oscillation (ENSO) in the tropical Pacific are collectively and simultaneously utilized in model training and in making predictions; therefore, this purely data-driven model is configured in both training and predicting procedures such that the coupled ocean-atmosphere interactions are adequately represented. Also similar to dynamic models, the prediction procedure is executed in a rolling manner to allow ocean-atmosphere anomaly exchanges month by month; the related key fields during multi-month time intervals (TIs) prior to prediction target months are taken as input predictors, serving as initial conditions to precondition the future evolution more effectively. Real-time predictions indicate that the climate conditions in the tropical Pacific are surely to develop into an El Niño state in late 2023. Furthermore, sensitivity experiments are conducted to examine how prediction skills are affected by the input predictor specifications, including TIs during which information on initial conditions is retained for making predictions. A comparison with other dynamic coupled models is also made to demonstrate the prediction performance for the 2023–2024 El Niño event. [ABSTRACT FROM AUTHOR]

Roll force prediction plays a significant role in rolling schedule and optimization. For a specific steel grade, the roll force can be determined by the aid of several factors: rolling speed, initial thickness, ratio of thickness reduction, the starting temperature of the strip, and the friction coefficient in the contact region. Roll force prediction mathematical models are sometimes rare and inaccurate. This paper presents a new approach to predict roll separating force using semi-supervised support vector regression (SSSVR). The parameters affecting the sensitivity of the SSSVR were optimized using the genetic algorithm to maximize the r-squared accuracy score. The intelligent system is evaluated using two quality metrics: the root mean square error (RMSE) and the mean absolute error calculated between the measured force from the industrial rolling field and the predicted force using the proposed system from one side, and the measured force and the calculated force from another side. Obtained results show the improvement while using the intelligent predictive system. The reduction in RMSE was achieved by the proposed system by 66.9% and 32.1% for oval and round shape passes, respectively in comparison to the conventional calculation method. [ABSTRACT FROM AUTHOR]

Transcranial magnetic stimulation (TMS) is a device-based neuromodulation technique increasingly used to treat brain diseases. Electric field (E-field) modeling is an important technique in several TMS clinical applications, including the precision stimulation of brain targets with accurate stimulation density for the treatment of mental disorders and the localization of brain function areas for neurosurgical planning. Classical methods for E-field modeling usually take a long computation time. Fast algorithms are usually developed with significantly lower spatial resolutions that reduce the prediction accuracy and limit their usage in real-time or near real-time TMS applications. This review paper discusses several modern algorithms for real-time or near real-time TMS E-field modeling and their advantages and limitations. The reviewed methods include techniques such as basis representation techniques and deep neural-network-based methods. This paper also provides a review of software tools that can integrate E-field modeling with navigated TMS, including a recent software for real-time navigated E-field mapping based on deep neural-network models. [ABSTRACT FROM AUTHOR]

Background: Predicting patient mortality risk facilitates early intervention in intensive care unit (ICU) patients at greater risk of disease progression. This study applies machine learning methods to multidimensional clinical data to dynamically predict mortality risk in ICU patients. Methods: A total of 33,798 patients in the MIMIC-III database were collected. An integrated model NIMRF (Network Integrating Memory Module and Random Forest) based on multidimensional variables such as vital sign variables and laboratory variables was developed to predict the risk of death for ICU patients in four non overlapping time windows of 0–1 h, 1–3 h, 3–6 h, and 6–12 h. Mortality risk in four nonoverlapping time windows of 12 h was externally validated on data from 889 patients in the respiratory critical care unit of the Chinese PLA General Hospital and compared with LSTM, random forest and time-dependent cox regression model (survival analysis) methods. We also interpret the developed model to obtain important factors for predicting mortality risk across time windows. The code can be found in https://github.com/wyuexiao/NIMRF. Results: The NIMRF model developed in this study could predict the risk of death in four nonoverlapping time windows (0–1 h, 1–3 h, 3–6 h, 6–12 h) after any time point in ICU patients, and in internal data validation, it is suggested that the model is more accurate than LSTM, random forest prediction and time-dependent cox regression model (area under receiver operating characteristic curve, or AUC, 0–1 h: 0.8015 [95% CI 0.7725–0.8304] vs. 0.7144 [95%] CI 0.6824–0.7464] vs. 0.7606 [95% CI 0.7300–0.7913] vs 0.3867 [95% CI 0.3573–0.4161]; 1–3 h: 0.7100 [95% CI 0.6777–0.7423] vs. 0.6389 [95% CI 0.6055–0.6723] vs. 0.6992 [95% CI 0.6667–0.7318] vs 0.3854 [95% CI 0.3559–0.4150]; 3–6 h: 0.6760 [95% CI 0.6425–0.7097] vs. 0.5964 [95% CI 0.5622–0.6306] vs. 0.6760 [95% CI 0.6427–0.7099] vs 0.3967 [95% CI 0.3662–0.4271]; 6–12 h: 0.6380 [0.6031–0.6729] vs. 0.6032 [0.5705–0.6406] vs. 0.6055 [0.5682–0.6383] vs 0.4023 [95% CI 0.3709–0.4337]). External validation was performed on the data of patients in the respiratory critical care unit of the Chinese PLA General Hospital. Compared with LSTM, random forest and time-dependent cox regression model, the NIMRF model was still the best, with an AUC of 0.9366 [95% CI 0.9157–0.9575 for predicting death risk in 0–1 h]. The corresponding AUCs of LSTM, random forest and time-dependent cox regression model were 0.9263 [95% CI 0.9039–0.9486], 0.7437 [95% CI 0.7083–0.7791] and 0.2447 [95% CI 0.2202–0.2692], respectively. Interpretation of the model revealed that vital signs (systolic blood pressure, heart rate, diastolic blood pressure, respiratory rate, and body temperature) were highly correlated with events of death. Conclusion: Using the NIMRF model can integrate ICU multidimensional variable data, especially vital sign variable data, to accurately predict the death events of ICU patients. These predictions can assist clinicians in choosing more timely and precise treatment methods and interventions and, more importantly, can reduce invasive procedures and save medical costs. [ABSTRACT FROM AUTHOR]

The optimization design of roller path is critical in conventional spinning as the roller path greatly influences the spinning status and forming quality. In this research, an innovative online intelligent method for roller path design was developed, which can capture the dynamic change of the spinning status under flexible roller path and greedily optimize the roller movement track progressively to achieve the design of whole roller path. In tandem with these, an online intelligent design system for roller path was developed with the aid of intelligent sensing, learning, optimization and execution. It enables the multi-functional of spinning condition monitoring, real-time prediction of spinning status, online dynamic processing optimization, and autonomous execution of the optimal processing. Through system implementation and verification by case studies, the results show that the intelligent processing optimization and self-adaptive control of the spinning process can be efficiently realized. The optimal roller path and matching spinning parameters (mandrel speed, feed ratio) can be efficiently obtained by only one simulation of the spinning process and no traditional trial-and-error is needed. Moreover, the optimized process can compromise the multi-objectives, including forming qualities (wall thickness reduction and flange fluctuation) and forming efficiency. The developed methodology can be generalized to handle other incremental forming processes. [ABSTRACT FROM AUTHOR]

Abstract The analysis of data streams offers a great opportunity for development of new methodologies and applications in the area of Intelligent Transportation Systems. In this paper, we propose two new incremental learning approaches for the travel time prediction problem for taxi GPS data streams in different scenarios and compare the same with three other existing methods. An extensive performance evaluation using four real life datasets indicate that when the training data size is small the Support Vector Regression method is the best choice considering both prediction accuracy and total computation time. However when the training data size is large to moderate then the Randomized K-Nearest Neighbor Regression with Spherical Distance (RKNNRSD) and the Incremental Polynomial Regression become the methods of choice. When continuous prediction of remaining travel time along the trajectory of a trip is considered we find that the RKNNRSD is the method of choice. A Real-time Speeding Alert System (RSAS) and a Driver Suspected Speeding Scorecard (DSSS) using the RKNNRSD method are proposed which have great potential for improving travel safety.

The analysis of data streams offers a great opportunity for development of new methodologies and applications in the area of Intelligent Transportation Systems. In this paper, we propose two new incremental learning approaches for the travel time prediction problem for taxi GPS data streams in different scenarios and compare the same with three other existing methods. An extensive performance evaluation using four real life datasets indicate that when the training data size is small the Support Vector Regression method is the best choice considering both prediction accuracy and total computation time. However when the training data size is large to moderate then the Randomized K-Nearest Neighbor Regression with Spherical Distance (RKNNRSD) and the Incremental Polynomial Regression become the methods of choice. When continuous prediction of remaining travel time along the trajectory of a trip is considered we find that the RKNNRSD is the method of choice. A Real-time Speeding Alert System (RSAS) and a Driver Suspected Speeding Scorecard (DSSS) using the RKNNRSD method are proposed which have great potential for improving travel safety. [ABSTRACT FROM AUTHOR]

The current research introduces a new success case toward monitoring the drilling fluid rheology with an automation system by implementing the machine learning application. For the first time, four developed models were developed to predict the fluid rheology of Micromax oil-based mud type by implementing the artificial neural network to overcome the conventional field lab tests for the mud rheology monitoring that take more time and have human errors. The developed models predict the mud rheological properties as the plastic and apparent viscosities, yield point, and flow behavior index to be in real-time for evaluating the mud functionality. Mud weight and Marsh funnel were utilized as the only inputs for the rheology prediction models as these inputs have easy high-frequency measurements during the drilling operation. Real field data set of 244 complete measurements are used to train the models and another unseen set of 138 points for testing the models' accuracy. The correlation coefficient (R) and average absolute percentage error were calculated between the predicted values and the actual measurements to determine the models performance. Through the models' development, the models perform a high accuracy level that has a coefficient of correlation higher than 0.96 for all models except the yield point model which shows 0.92, besides, an accepted average absolute percentage error that ranges from 4 to 9.3% for the training and testing phases. Furthermore, the study proposes model-based correlations for predicting the drilling fluid rheology for easy application in the rig site without the need for running the codes. [ABSTRACT FROM AUTHOR]

Recently, space applications have become complicated for post-flight analysis when a sensor fails. Various scholars have researched many types of research based on fault data prediction in aircraft applications, utilizing numerous deep learning techniques. Using these deep learning techniques, the sensor data is re-created with the help of other related sensor data and will help other post-flight analyses. Once the launch vehicle is lifted off, there is no possibility of solving a problem in sensors. Sometimes that particular sensor is very crucial for the onboard decisions. There have to adapt real-time sensor prediction techniques. So, this paper focused on designing an effective prediction technique for fault sensor data with the aid of its corresponding sensor data. The fault sensor data prediction process is performed in two stages such as real-time and offline. Here we apply three deep learning algorithms to predict the fault sensor data, and the three algorithms, LSTM, GRU and CNN, are applied for real-time and offline data prediction. Furthermore, the experimental setup helps for predicting accurate real-time fault sensor data. The results obtained through experimental and simulation analysis are closely matched for a failed sensor, which is very helpful for analyzing and validating the launch vehicle performance. [ABSTRACT FROM AUTHOR]

Horizontal wells have become the predominant well type for developing complex oil and gas reservoirs, yielding favorable development outcomes. Accurately predicting the rate of penetration (ROP) is of great significance for cost reduction and efficiency improvement in the drilling process of horizontal wells. Owing to the continuous demand for directional drilling and the intricate trajectory, it is prone to induce complex mechanical behaviors, such as pipe buckling and wellbore stability. It is difficult to accurately predict ROP of horizontal wells. To solve this problem, a novel data-driven ROP prediction model for horizontal wells is proposed by combining a self-attention-based Gated recurrent unit (GRU) network and Back-propagation (BP) Neural Network. At first, the selection of the model's processed input characteristics, in addition to the azimuth and inclination of the well trajectory, is guided by physical laws and statistical analyses. Then, temporal features, such as weight on bit and rotary torque, are input into GRU network, and non-temporal features, such as bottom hole assembly (BHA) and formation layers, are characterized and dimensionally reduced through unique hot encoding and embedding layers before being input into the BP neural network. Next, to accomplish real-time prediction, the model is dynamically updated utilizing the sliding window approach and data flow after training with various well section (vertical, build-up and horizontal well) data. Finally, the model's performance is comprehensively analyzed using field drilling datasets after optimizing model hyperparameters through the orthogonal experiment method. Compared with the BP and LSTM models, the mean absolute percentage error is separately reduced 9.09% and 5.99%. By introducing an incremental update mechanism, the prediction mean absolute percent error (MAPE) of the dynamic update model is reduced by 5.42% compared to the offline model. Results indicate that the model demonstrates strong superiority due to the self-attention-based dual-input network structure, segmented modeling, and dynamic update mechanism. The model exhibits high accuracy and robustness, offering valuable theoretical support for the optimal and efficient drilling of horizontal wells. [ABSTRACT FROM AUTHOR]

The rock porosity is considered a key petrophysical property for the rock due to its great impact on the hydrocarbon reserve estimation and petroleum economics. The conventional methods for determining the rock porosity either from the logging tools, lab measurements for the cored samples, or using empirical correlations from other parameters are costly, time-consuming, or did not provide the required level of accuracy. The new horizon for implementing machine learning techniques as a new approach for predicting the rock porosity overcomes all of the above drawbacks. Therefore, the objective of this research is to develop a new model based on an artificial neural network (ANN) for predicting rock porosity from only drilling parameters that include weight on bit, torque, standpipe pressure, drill string rotation speed, rate of penetration, and pump rate. The study used two data sets for building the model (3767 data points) and the second one for validating the developed ANN model (1676 data points). ANN model was built and optimized with deep sensitivity analysis for the ANN model parameters to achieve strong prediction results. ANN model showed a correlation coefficient (R) between the predicted and actual porosity values of 0.97 and 0.92 with average absolute percentage errors (AAPE) of 6.2 and 9.3% for training and testing, respectively. The model validation enhanced the high prediction performance as ANN achieved R of 0.95 and AAPE of 8.5%. The study provides new contributions as predicting the rock porosity for complex lithology formations (sandstone, shale, and carbonate), developing an ANN porosity model with a high level of accuracy, and a newly developed ANN-based equation for estimating the porosity from only the surface drilling data. [ABSTRACT FROM AUTHOR]

The dynamic Young's modulus (Edyn) is an important factor for evaluating the static Young's modulus which is an important parameter influences hydraulic fracturing design, drilling performance, in situ stresses estimation, and wellbore stability evaluation. Currently, the prediction of the Edyn requires the knowledge of the shear and compressional velocities and bulk density, which may not always be available. For the first time, an equation for evaluating Edyn in real-time from the weight on bit, the rate of penetration, torque, standpipe pressure, drill pipe rotation speed, and the drilling fluid flow rate is introduced. This new correlation was developed from the artificial neural networks (ANN) which were trained on 2054 data points from Well-A. The developed correlation was tested and validated on 871 and 2912 data points from Well-B and Well-C, respectively. The results showed that the ANN model evaluated Edyn with average absolute percentage error (AAPE) of 3.09%, for the training data. The developed equation estimated the Edyn with AAPE of only 3.38% and 3.73%, for the testing and validation data, respectively. [ABSTRACT FROM AUTHOR]

This paper describes the walking motion prediction of a user and prediction-based joint trajectory generation of a robotic transfemoral prosthetic leg. To this end, two artificial neural networks are proposed to predict the user's walking intention and percent of gait cycle (PGC) based on ground foot contact, kinematics, and electromyography (EMG) information. Through the sensor module designed for our study, EMG signals were measured at the tibialis anterior (TA) and gastrocnemius (GAS) of the sound leg, and inertial information was measured at the shin of the sound leg and the thigh of the affected leg. Notably, the time sequence feature data suggested based on the characteristics of the phase portrait of each joint motion showed better learning performance than when using the current time feature data. Walking and standing states were accurately predicted through an artificial neural network (ANN) for walking state classification, and the intention to start and stop walking could be predicted within approximately 83 ms. For generating gait motion of the robotic prosthetic leg, PGC while walking was also accurately predicted through an artificial neural network. Consequently, we developed an algorithm for the generation of gait motion using two artificial neural networks, and its performance was verified through dynamic simulations. [ABSTRACT FROM AUTHOR]

The COronaVIrus Disease 2019 (COVID-19) pandemic is unfortunately highly transmissible across the people. In order to detect and track the suspected COVID-19 infected people and consequently limit the pandemic spread, this paper entails a framework integrating the machine learning (ML), cloud, fog, and Internet of Things (IoT) technologies to propose a novel smart COVID-19 disease monitoring and prognosis system. The proposal leverages the IoT devices that collect streaming data from both medical (e.g., X-ray machine, lung ultrasound machine, etc.) and non-medical (e.g., bracelet, smartwatch, etc.) devices. Moreover, the proposed hybrid fog-cloud framework provides two kinds of federated ML as a service (federated MLaaS); (i) the distributed batch MLaaS that is implemented on the cloud environment for a long-term decision-making, and (ii) the distributed stream MLaaS, which is installed into a hybrid fog-cloud environment for a short-term decision-making. The stream MLaaS uses a shared federated prediction model stored into the cloud, whereas the real-time symptom data processing and COVID-19 prediction are done into the fog. The federated ML models are determined after evaluating a set of both batch and stream ML algorithms from the Python's libraries. The evaluation considers both the quantitative (i.e., performance in terms of accuracy, precision, root mean squared error, and F1 score) and qualitative (i.e., quality of service in terms of server latency, response time, and network latency) metrics to assess these algorithms. This evaluation shows that the stream ML algorithms have the potential to be integrated into the COVID-19 prognosis allowing the early predictions of the suspected COVID-19 cases. [ABSTRACT FROM AUTHOR]

We propose a CFD algorithm for real-time prediction of urban flow and dispersion based on large-eddy simulation (LES). To efficiently handle complex urban building geometry, we implement a modified immersed boundary method (IBM), which can be applied to bluff building boundary on a staggered grid. For an introduction of proper inflow condition, we apply a synthetic-eddy method to the periodic domain, which is necessary for direct solving of Poisson equation for pressure. All these implementations are conducted on GPU system for a significant reduction of calculation time for real-time prediction. For validation of our algorithm, we test our model in the prediction of flow and dispersion in urban area in Seoul against wind-tunnel experiment result and other simulation using fire dynamics simulator (FDS). Our model simulation results of flow and dispersion show good agreement with them. Simulation time is reasonably short to warrant real-time prediction of flow and dispersion. [ABSTRACT FROM AUTHOR]

Real-time prediction of spatial raster time series, such as those derived from satellite remote sensing imagery, is important for making emergency decisions on various geo-spatial processes/events. However, because of the scalability issue and large training time requirement, the neural network (NN)-based models often fail to perform real-time prediction, in spite of their tremendous potential. In this paper, we propose ContRast, a variant of recurrent NN-based context-aware raster time series prediction model that attempts to resolve these issues by: (1) eliminating the need for offline adjustment of network structure by employing self-evolving autonomous learning of recurrent neural network, (2) saving training time by adopting single-pass parameter learning mechanism, and (3) reducing redundant learning by skipping sub-regional data associated with similar spatio-temporal context and reusing already learned parameters to predict for the same. Experimental evaluations with respect to predicting normalized difference vegetation index (NDVI)-raster derived from MODIS Terra satellite remote sensing imagery show that ContRast is highly effective for real-time prediction of spatial raster time series, and it significantly outperforms the existing models. In addition, the theoretical analyses of model complexity and computational cost further justify our empirical observations. [ABSTRACT FROM AUTHOR]

A new framework of data assessment and prioritization for real-time prediction of spatial information is presented. The real-time prediction of spatial information is promising for next-generation mobile networks. Recent developments in machine learning technology have enabled prediction of spatial information, which will be quite useful for smart mobility services including navigation, driving assistance, and self-driving. Other key enablers for forming spatial information are image sensors in mobile devices like smartphones and tablets and in vehicles such as cars and drones and real-time cognitive computing like automatic number/license plate recognition systems and object recognition systems. However, since image data collected by mobile devices and vehicles need to be delivered to the server in real time to extract input data for real-time prediction, the uplink transmission speed of mobile networks is a major impediment. This paper proposes a framework of data assessment and prioritization that reduces the uplink traffic volume while maintaining the prediction accuracy of spatial information. In our framework, machine learning is used to estimate the importance of each data element and to predict spatial information under the limitation of available data. A numerical evaluation using an actual vehicle mobility dataset demonstrated the validity of the proposed framework. Two extension schemes in our framework, which use the ensemble of importance scores obtained from multiple feature selection methods, are also presented to improve its robustness against various machine learning and feature selection methods. We discuss the performance of those schemes through numerical evaluation. [ABSTRACT FROM AUTHOR]

A new system for computing the multimodel ensemble predictions of water levels, known as HydroProg, is applied in the process of real-time modelling and forecasting riverflow in the upper Nysa Kłodzka basin (SW Poland). The HydroProg system automatically produces early warnings against high flows on a basis of prognoses from external hydrologic models which are run in a cloud-like fashion using the real-time hydrometeorological observations. The predictions offered by the models serve as ensemble members and become inputs to HydroProg which does data pre- and post-processing as well as computes multimodel ensemble predictions. The HydroProg system is used in this paper to verify the operational hypothesis that a two-model ensemble hydrologic prediction, solely based on two prognoses computed from the multi- and univariate autoregressive statistical models, reveals better skills than the individual members. The analysis is conducted for all lead times and selected phases of high flow development, including a rising limb of hydrograph but excluding a peak flow. The investigation is carried out using the first HydroProg prototype that works experimentally for the upper Nysa Kłodzka basin (SW Poland). The implementation offers 3-h predictions of water levels, based on the regularly (15 mins) calibrated and updated individual models (the emphasis is put on two models, but other models are also presented for reference) which become inputs to the multimodel ensemble solution (mainly the two-model approach, with the six-model solution presented for comparison). The most significant high flows that occurred in the study area between 01/09/2013 and 31/08/2015 have been investigated. It has been found that for slightly rising water level the two-model ensemble is recommended which, for a subsequently moderately or rapidly rising limb of hydrograph, should be replaced by the vector autoregressive model. [ABSTRACT FROM AUTHOR]

Tropical cyclone (TC) 'Hudhud' (October 2014) over the Bay of Bengal was the first TC to strike the Port city, Vishakhapatnam, with intensity of 100 knots (180 km h) since 1891. The advanced research weather research and forecasting (ARW) model with 9-km grid spacing is coupled offline with dynamical storm surge model and is used for real-time prediction of Hudhud. In the current study, the successful story of prediction of Hudhud is substantiated. The ARW model was able to predict the genesis in advance of 3 days with reasonable accuracy in position and strength. The mean initial vortex position and intensity errors are 46 km and 6 knots, respectively. The model predictions showed that Hudhud would intensify to a very severe cyclonic storm (max. sustained wind speed ~97 knots from all the experiments) in advance of ~4 days which is consistent with the observation. The mean absolute intensity forecast error is significantly less (18 knots) up to 96-h forecast. The model could reproduce the upper-level warming of ~7.5 °C at ~50 km radial distance similar to that of satellite observation. The ARW model showed appreciable performance in track prediction (with errors ≤150 km up to 96-h forecast). On the landfall day, the rainfall patterns are realistic and consistent with IMD rainfall observation. The dynamical storm surge model, forced with the outputs of the ARW model, was able to predict realistic surge ranging between 1.4-2.2 m in 24- to 96-h forecast lengths. This study highlights the importance and need of high-resolution numerical models for operational TC and associated surge forecast over the region. [ABSTRACT FROM AUTHOR]

In this paper, we propose a real-time cooling load forecasting system in order to overcome the problems of the conventional methods. The proposed system is a new load forecasting model that hierarchically combines Support Vector Data Descriptions (SVDDs). The proposed system selects an optimal attribute subset by our cooling load forecasting system that enables real-time load data generation and collection. The system is composed of two layers: The first layer predicts the time slots in three representative forms: morning, midday and afternoon. The second layer performs specialized prediction of each individual time slot. Since the proposed system enables both coarse-and fine-grained forecasting, it can efficient cooling load management. Moreover, even when a new time slot emerges, it can be easily adapted for incremental updating and scaling. The performance of the proposed system is validated via experiments which confirm that the recall and precision measures of the method are satisfactory. [ABSTRACT FROM AUTHOR]

Real-time predictions for the JAL severe cyclone formed in November 2010 over Bay of Bengal using a high-resolution Weather Research and Forecasting (WRF ARW) mesoscale model are presented. The predictions are evaluated with different initial conditions and assimilation of observations. The model is configured with two-way interactive nested domains and with fine resolution of 9 km for the region covering the Bay of Bengal. Simulations are performed with NCEP GFS 0.5° analysis and forecasts for initial/boundary conditions. To examine the impact of initial conditions on the forecasts, eleven real-time numerical experiments are conducted with model integration starting at 00, 06, 12, 18 UTC 4 Nov, 5 Nov and 00, 06, 12 UTC 6 Nov and all ending at 00 UTC 8 Nov. Results indicated that experiments starting prior to 18 UTC 04 Nov produced faster moving cyclones with higher intensity relative to the IMD estimates. The experiments with initial time at 18 UTC 04 Nov, 00 UTC 05 Nov and with integration length of 78 h and 72 h produced best prediction comparable with IMD estimates of the cyclone track and intensity parameters. To study the impact of observational assimilation on the model predictions FDDA, grid nudging is performed separately using (1) land-based automated weather stations (FDDAAWS), (2) MODIS temperature and humidity profiles (FDDAMODIS), and (3) ASCAT and OCEANSAT wind vectors (FDDAASCAT). These experiments reduced the pre-deepening period of the storm by 12 h and produced an early intensification. While the assimilation of AWS data has shown meagre impact on intensity, the assimilation of scatterometer winds produced an intermittent drop in intensity in the peak stage. The experiments FDDAMODIS and FDDAQSCAT produced minimum error in track and intensity estimates for a 90-h prediction of the storm. [ABSTRACT FROM AUTHOR]

Elevation angle is one of the most significant parameters of land mobile satellite channels, subject to rapid variations in the case of Low Earth Orbit (LEO) satellite systems. In this paper a novel trace-based framework is proposed and analyzed which is capable of predicting elevation angle as a function of time during satellite visibility window. Trace-time based modeling makes the framework practical for real-time evaluation of elevation angle and its alteration incurred by communication links in LEO satellite systems. The proposed method is particularly suitable for development of communication channel models and services in mobile LEO satellite networks where path variability is of great importance. [ABSTRACT FROM AUTHOR]

Copyright of Arabian Journal of Geosciences is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

It is of great social and scientific importance and also very difficult to make reliable prediction for dust weather frequency (DWF) in North China. In this paper, the correlation between spring DWF in Beijing and Tianjin observation stations, taken as examples in North China, and seasonally averaged surface air temperature, precipitation, Arctic Oscillation, Antarctic Oscillation, South Oscillation, near surface meridional wind and Eurasian westerly index is respectively calculated so as to construct a prediction model for spring DWF in North China by using these climatic factors. Two prediction models, i.e. model-I and model-II, are then set up respectively based on observed climate data and the 32-year (1970–2001) extra-seasonal hindcast experiment data as reproduced by the nine-level Atmospheric General Circulation Model developed at the Institute of Atmospheric Physics (IAP9L-AGCM). It is indicated that the correlation coefficient between the observed and predicted DWF reaches 0.933 in the model-I, suggesting a high prediction skill one season ahead. The corresponding value is high up to 0.948 for the subsequent model-II, which involves synchronous spring climate data reproduced by the IAP9L-AGCM relative to the model-I. The model-II can not only make more precise prediction but also can bring forward the lead time of real-time prediction from the model-I’s one season to half year. At last, the real-time predictability of the two models is evaluated. It follows that both the models display high prediction skill for both the interannual variation and linear trend of spring DWF in North China, and each is also featured by different advantages. As for the model-II, the prediction skill is much higher than that of original approach by use of the IAP9L-AGCM alone. Therefore, the prediction idea put forward here should be popularized in other regions in China where dust weather occurs frequently. [ABSTRACT FROM AUTHOR]

Real-time rapid prediction of variations of the Earth's rotational rate is of great scientific and practical importance. However, due to the complicated time-variable characteristics of variations of the Earth's rotational rate (i.e., length of day, LOD), it is usually difficult to obtain satisfactory predictions by conventional linear time series analysis methods. This study employs the nonlinear artificial neural networks (ANN) to predict the LOD variations. The topology of the ANN model is determined by minimizing the root mean square errors (RMSE) of the predictions. Considering the close relationships between the LOD variations and the atmospheric circulation movement, the operational prediction series of axial atmospheric angular momentum (AAM) is incorporated into the ANN model as an additional input in the real-time rapid prediction of LOD variations with 1-5 days ahead. The results show that the LOD prediction is significantly improved after introducing the operational prediction series of AAM into the ANN model. [ABSTRACT FROM AUTHOR]

Acid rain and photochemical smog are two regionalair-quality issues that have received considerableattention in the last two decades due to their harmfuleffects. Health impacts of particulate matter (PM) inthe atmosphere is another issue of concern. Sulphur dioxide emission controls were introducedin both Canada and the U.S.A. to reduceacid-deposition-related damage. While these emissionreductions have already resulted in reduced sulphatedeposition, based on results from modelling studiesmuch of southeastern Canada is still expected toexperience damaging levels of acid deposition evenafter all currently legislated emission controls arefully implemented. Moreover, there has not been acorresponding reduction in the acidity ofprecipitation. This may be attributable to aconcurrent reduction in base-cation concentration inprecipitation. Models were also developed to understand theformation, transport and diffusion of troposphericozone. The models have been used to provide policyguidance for emission control options to reduceground-level ozone to acceptable limits. In thesummer of 1997 a Canadian pilot project was initiatedto provide real-time forecasts of ground-level ozonein the southeastern part of the province of NewBrunswick in eastern Canada. With the emergence of fine Particulate Matter(PM2.5) as a health concern, efforts are underwayin Canada to develop a “unified'' regional air-qualitymodel that will address the combined impacts ofvarious pollutants in the atmosphere. In this effortthe atmosphere is viewed as a single entity where theimpacts of multiple pollutants are considered at thesame time. [ABSTRACT FROM AUTHOR]