Location Prediction Research Articles

Displacement and strain data-driven damage detection in multi-component and heterogeneous composite structures

This work introduces the use of convolutional neural network (CNN) in combination with advanced structural theories for the damage detection of multi-component and composite structures. Well-established component-wise (CW) models based on the Carrera Unified Formulation (CUF) are developed first to demonstrate the effect of localized damages on the mechanical performance of thin-walled beams and laminates. Finite element Monte Carlo simulations of randomly damaged structures are then used to generate a large database of full-field displacement and strain images. These images are lately feed into a dedicated CNN for training purpose and for the prediction of location and intensity of structural damages occurring in unseen scenarios. The results demonstrate the validity of the present approach and suggest the importance of adopting opportune structural models to carry out localized damage detection by image-driven AI. Overall, the research provides good confidence for future investigation and experimental testing.

OHetTLAL: An Online Transfer Learning Method for Fingerprint-Based Indoor Positioning.

In an indoor positioning system (IPS), transfer learning (TL) methods are commonly used to predict the location of mobile devices under the assumption that all training instances of the target domain are given in advance. However, this assumption has been criticized for its shortcomings in dealing with the problem of signal distribution variations, especially in a dynamic indoor environment. The reasons are: collecting a sufficient number of training instances is costly, the training instances may arrive online, the feature spaces of the target and source domains may be different, and negative knowledge may be transferred in the case of a redundant source domain. In this work, we proposed an online heterogeneous transfer learning (OHetTLAL) algorithm for IPS-based RSS fingerprinting to improve the positioning performance in the target domain by fusing both source and target domain knowledge. The source domain was refined based on the target domain to avoid negative knowledge transfer. The co-occurrence measure of the feature spaces (Cmip) was used to derive the homogeneous new feature spaces, and the features with higher weight values were selected for training the classifier because they could positively affect the location prediction of the target. Thus, the objective function was minimized over the new feature spaces. Extensive experiments were conducted on two real-world scenarios of datasets, and the predictive power of the different modeling techniques were evaluated for predicting the location of a mobile device. The results have revealed that the proposed algorithm outperforms the state-of-the-art methods for fingerprint-based indoor positioning and is found robust to changing environments. Moreover, the proposed algorithm is not only resilient to fluctuating environments but also mitigates the model's overfitting problem.

Predictive Models for Length of Stay and Discharge Disposition in Elective Spine Surgery: Development, Validation, and Comparison to the ACS NSQIP Risk Calculator.

A retrospective study at a single academic institution. The purpose of this study is to utilize machine learning to predict hospital length of stay (LOS) and discharge disposition following adult elective spine surgery, and to compare performance metrics of machine learning models to the American College of Surgeon's National Surgical Quality Improvement Program's (ACS NSQIP) prediction calculator. A total of 3678 adult patients undergoing elective spine surgery between 2014 and 2019, acquired from the electronic health record. Patients were divided into three stratified cohorts: cervical degenerative, lumbar degenerative, and adult spinal deformity groups. Predictive variables included demographics, body mass index, surgical region, surgical invasiveness, surgical approach, and comorbidities. Regression, classification trees, and least absolute shrinkage and selection operator (LASSO) were used to build predictive models. Validation of the models was conducted on 16% of patients (N=587), using area under the receiver operator curve (AUROC), sensitivity, specificity, and correlation. Patient data were manually entered into the ACS NSQIP online risk calculator to compare performance. Outcome variables were discharge disposition (home vs. rehabilitation) and LOS (days). Of 3678 patients analyzed, 51.4% were male (n=1890) and 48.6% were female (n=1788). The average LOS was 3.66 days. In all, 78% were discharged home and 22% discharged to rehabilitation. Compared with NSQIP (Pearson R2 =0.16), the predictions of poisson regression ( R2 =0.29) and LASSO ( R2 =0.29) models were significantly more correlated with observed LOS ( P =0.025 and 0.004, respectively). Of the models generated to predict discharge location, logistic regression yielded an AUROC of 0.79, which was statistically equivalent to the AUROC of 0.75 for NSQIP ( P =0.135). The predictive models developed in this study can enable accurate preoperative estimation of LOS and risk of rehabilitation discharge for adult patients undergoing elective spine surgery. The demonstrated models exhibited better performance than NSQIP for prediction of LOS and equivalent performance to NSQIP for prediction of discharge location.

BIAS-3D: Brain inspired attentional search model fashioned after what and where/how pathways for target search in 3D environment.

We propose a brain inspired attentional search model for target search in a 3D environment, which has two separate channels-one for the object classification, analogous to the "what" pathway in the human visual system, and the other for prediction of the next location of the camera, analogous to the "where" pathway. To evaluate the proposed model, we generated 3D Cluttered Cube datasets that consist of an image on one vertical face, and clutter or background images on the other faces. The camera goes around each cube on a circular orbit and determines the identity of the image pasted on the face. The images pasted on the cube faces were drawn from: MNIST handwriting digit, QuickDraw, and RGB MNIST handwriting digit datasets. The attentional input of three concentric cropped windows resembling the high-resolution central fovea and low-resolution periphery of the retina, flows through a Classifier Network and a Camera Motion Network. The Classifier Network classifies the current view into one of the target classes or the clutter. The Camera Motion Network predicts the camera's next position on the orbit (varying the azimuthal angle or "θ"). Here the camera performs one of three actions: move right, move left, or do not move. The Camera-Position Network adds the camera's current position (θ) into the higher features level of the Classifier Network and the Camera Motion Network. The Camera Motion Network is trained using Q-learning where the reward is 1 if the classifier network gives the correct classification, otherwise 0. Total loss is computed by adding the mean square loss of temporal difference and cross entropy loss. Then the model is trained end-to-end by backpropagating the total loss using Adam optimizer. Results on two grayscale image datasets and one RGB image dataset show that the proposed model is successfully able to discover the desired search pattern to find the target face on the cube, and also classify the target face accurately.

Stem and leaf rust-induced miRNAome in bread wheat near-isogenic lines and their comparative analysis.

Wheat rusts remain a major threat to global wheat production and food security. The R-gene-mediated resistance has been employed as an efficient approach to develop rust-resistant varieties. However, evolution of new fungal races and infection strategies put forward the urgency of unravelling novel molecular players, including non-coding RNAs for plant response. This study identified microRNAs associated with Sr36 and Lr45 disease resistance genes in response to stem and leaf rust, respectively. Here, small RNA sequencing was performed on susceptible and resistant wheat near-isogenic lines inoculated with stem and leaf rust pathotypes. microRNA mining in stem rust-inoculated cultivars revealed a total of distinct 26 known and 7 novel miRNAs, and leaf rust libraries culminated with 22 known and 4 novel miRNAs. The comparative analysis between two disease sets provides a better understanding of altered miRNA profiles associated with respective R-genes and infections. Temporal differential expression pattern of miRNAs pinpoints their role during the progress of infection. Differential expression pattern of miRNAs among various treatments as well as time-course expression of miRNAs revealed stem and leaf rust-responsive miRNAs and their possible role in balancing disease resistance/susceptibility. Disclosure of guide strand, passenger strand and a variant of novel-Tae-miR02 from different subgenome origins might serve as a potential link between stem and leaf rust defence mechanisms downstream to respective R-genes. The outcome from the analysis of microRNA dynamics among two rust diseases and further characterization of identified microRNAs can contribute to significant novel insights on wheat-rust interactions and rust management. KEY POINTS: • Identification and comparative analysis of stem and leaf rust-responsive miRNAs. • Chromosomal location and functional prediction of miRNAs. • Time-course expression analysis of pathogen-responsive miRNAs.

Simulation of a round supersonic combustor using wall-modeled large eddy simulation and partially-stirred reactor models

Simulations are presented for a generic, round supersonic combustor. Turbulence is modeled in the combustor using a wall-modeled large eddy simulation approach. Combustion is modeled using a small quasi-global mechanism and a more detailed skeletal mechanism. Both mechanisms are used in conjunction with two variations of the partially-stirred reactor model for sub-grid turbulence chemistry interactions. Sensitivity of the solutions to grid resolution is investigated. It is found that in order to achieve reasonable grid convergence in the mean wall pressure, the model constant that appears in the partially-stirred reactor model must be a function of both the chemistry mechanism and the grid resolution. Most of the combinations of mechanism and turbulent combustion model tested can be tuned in order to predict the location of the pre-combustion shock train and the peak mean pressure in the combustor. It is found that while the different models are able to reproduce the mean wall pressure, there are significant differences in the mean temperature and heat release rate fields. The sensitivity of the different combinations of mechanisms and partially-stirred reactor formulation is quantified and some combinations are found to be more prone to blowout. Two of the tuned models were tested across several fuel equivalence ratios with a single value of the partially-stirred reactor model constant. One model produced reasonable predictions of shock location and peak mean pressure for each equivalence ratio. The second model captured the global trends in the mean wall pressure, but was unable to quantitatively predict the shock location and peak mean pressure for all equivalence ratios tested.

Mobility Sensing and V2X Communication for Emergency Services

The warning of traveling emergency response vehicles (ERVs) is performed through the help of both sound and lights when the ERV is approaching. This is not an ideal situation for, at least, two reasons: the information is received only when the ERV is very close to other vehicles; and the other vehicles cannot understand where exactly the ERV is, and which action is the best according to its location and direction. With the help of vehicular communications, this paper proposes a different approach for an ERV to be autonomously and preemptively detected. Based on the collected data through vehicle communication and sensors, primarily through Cooperative Awareness Messages (CAMs) and Radio Detection And Ranging (RADAR) data, a prediction of the ERV future location is performed in real-time, and warning messages are disseminated to vehicles just before the ERV arrival. Real-world evaluation tests, considering different warning dissemination scenarios, show that the warning messages lost during the process is significantly reduced from almost 80% down to 22%, especially when infrastructure-to-vehicle and vehicle-to-vehicle communications are used together. Moreover, 80% of the total warning messages are delivered in less than 100 ms.

TCSA-Net: A Temporal-Context-Based Self-Attention Network for Next Location Prediction

Next location prediction aims to find the location that the user will visit next. It plays a fundamental role for location-based applications. However, the heterogeneity and sparsity of the trajectory data pose great challenges to the task. Recently, RNN-based methods have shown promising performance in learining the spatio-temporal characteristics of the trajectory. While the effectiveness of location prediction has been improved, the computational efficiency and the long-term preferences still leave space for further research. The self-attention mechanism is viewed as a promising solution for parallel computation and exploiting sequential regularities from sparse data. But the huge memory cost and the neglect of temporal information make it infeasible to directly modeling human mobility regularities. In this paper, we propose a temporal-context-based self-attention network named TCSA-Net, which can simultaneously exploit long- and short-term mvoement preferences from sparse and long trajectories. In particular, we design a novel two-stage self-attention architecture that can learn long-term dependency under constrained memory budget. Further, we propose a multi-modal embedding layer to model two complementary temporal contexts and provide more abundant temporal and sequential information. Extensive experiments on two real-life datasets show that the TCSA-Net significantly outperforms the state-of-the-art methods in terms of standard evaluation metrics.

Pre-Training Time-Aware Location Embeddings from Spatial-Temporal Trajectories

With the increasing accumulation of spatial-temporal trajectory data, location-based data mining has recently been extensively studied. A fundamental research topic in this field is learning the embedding vectors of locations through self-supervised pre-training. Pre-trained embedding vectors can utilize the highly available unlabeled trajectory data, and benefit downstream tasks in multiple aspects. However, most existing methods ignore the temporal information hidden in the visited time of locations in trajectories. Considering that human activities are highly regulated by specific periods of a day, temporal information can reflect some intrinsic characteristics of locations, so it is necessary to fuse them into location embedding vectors. In this paper, we propose a Time-Aware Location Embedding (TALE) pre-training method based on the CBOW framework, which is able to incorporate temporal information into the learned embedding vectors of locations. A novel temporal tree structure is designed to extract temporal information during the calculation of Hierarchical Softmax. In order to verify the effectiveness of TALE, we apply the learned embedding vectors into three downstream location-based prediction tasks, i.e., location classification, location visitor flow prediction and user next location prediction. Experiments are conducted on four real-world user trajectory datasets, and the experimental results demonstrate that our TALE model can obviously help downstream tasks gain better performance.

Deep-Learning-Based Gas Leak Source Localization From Sparse Sensor Data

In this article, we address the problem of estimating the location of gas leak sources using sparse unreliable spatio-temporal chemical sensor data. We pose the task of estimating the underlying gas signal and predicting the source location as an inverse problem. For this purpose, we develop a novel deep-learning projection-based framework. We incorporate traditional projection-onto-convex-sets (POCS) iteration in the structure of the deep model to obtain a regularized solution that conforms to our prior knowledge of the spatio-temporal structure of the gas concentration distribution. We use discrete cosine transform (DCT) layers to model the smooth nature of the gas plume signal. In the DCT domain, we project the feature maps onto a low-pass region, whose boundary is determined during training using the backpropagation algorithm. This operation is equivalent to projecting onto a convex set. Furthermore, these projection operations are embedded in the nonlinear structure of a convolutional neural network. We address two different types of data: methane–propane leak from industrial plants and isopropyl alcohol (isopropanol) vapor leak in an indoor environment. Experimental results are presented. Our results show that we can obtain a smooth estimate of the underlying gas signal while obtaining a good source location prediction with high accuracy.

Ship Motion-Based Prediction of Damage Locations Using Bidirectional Long Short-Term Memory

<i>The initial response to a marine accident can play a key role to minimize the accident. Therefore, various decision support systems have been developed using sensors, simulations, and active response equipment. In this study, we developed an algorithm to predict damage locations using ship motion data with bidirectional long short-term memory (BiLSTM), a type of recurrent neural network. To reflect the low frequency ship motion characteristics, 200 time-series data collected for 100 s were considered as input values. Heave, roll, and pitch were used as features for the prediction model. The F1-score of the BiLSTM model was 0.92; this was an improvement over the F1-score of 0.90 of a prior model. Furthermore, 53 of 75 locations of damage had an F1-score above 0.90. The model predicted the damage location with high accuracy, allowing for a quick initial response even if the ship did not have flood sensors. The model can be used as input data with high accuracy for a real-time progressive flooding simulator on board.</i>

EPM–DCNN: Earthquake Prediction Models Using Deep Convolutional Neural Networks

ABSTRACT Earthquakes usually cause severe injuries and loss of life, so researchers have developed various methods to predict them. However, the prediction accuracies of these methods are not satisfactory. Unlike most artificial intelligence earthquake prediction methods using earthquake catalogs or seismic wave data, this article proposes three earthquake prediction models based on deep convolutional neural network-based (EPM-DCNN) using 11 continuous earthquake precursory observation item data, including fluid, geomagnetic, and deformation disciplines. To enhance the accuracy of the location prediction of earthquakes, we propose a method to divide the research area into six prediction blocks based on the K-means++ clustering algorithm using the epicenter of historical earthquakes. Using earthquake precursory observation time-series data from 1 January 2015 to 31 December 2018, we construct approximately 34,000 samples by sliding a fixed window size. Each sample is subdivided into 13 categories by combining the magnitude label and prediction block label. The experimental results show that EPM–DCNN B proposed in this article has an accuracy of 99.0% and a recall of 99.8%, which demonstrates the effectiveness of EPM–DCNN for seismic prediction compared to several state-of-the-art baselines.

Paleo‐stratigraphic permeability anisotropy controls supergene mimetic martite goethite deposits

AbstractThe Hamersley Basin in Western Australia is one of the world's largest iron ore‐producing regions, hosting two types of ore in banded iron formations: the high‐grade martite‐microplaty haematite and the supergene martite‐goethite ores. With the high‐grade ores almost entirely mined in the last decade, the supergene ores have more recently become the dominant resource of interest. Consequently, understanding the genesis of these martite‐goethite deposits is a critical step for exploration. Yet, although various models exist, there is still no consensus on how these mineral resources formed, complicating the prediction of resource volume and location. Here, we show that the paleo‐stratigraphic permeability anisotropy (with higher permeability along strata than across) controls the supergene mimetic enrichment transport process and, subsequently, the mineralisation distribution. We introduce a flow model that implicitly represents strata with a potential function that orients the permeability tensor accurately. The numerical solver uses automatic mesh adaptivity to deliver robust solutions. By accurately reproducing the mineralisation patterns in specific deposits, we identify and quantify the paleo‐water table level and permeability anisotropy ratio as the two main controlling parameters for the mineralisation distribution. These insights provide new timing constraints for the mineralisation and the physical process of iron enrichment, suggesting much more potential mineralisation volume in the paleo‐reconstructed zones than previously anticipated. These flow models allow us to draw geological conclusions with few a priori assumptions required for the genetic model in which the transport component is dominant. The predictive power of this methodology will allow targeted drilling to narrow down the prospective areas and lower exploration costs. Furthermore, the methodology's generality applies to other commodities in sedimentary basins involving supergene processes and will improve our understanding of various genetic models.

An energy efficient dead reckoning localization for mobile Underwater Acoustic Sensor Networks

A wide range of marine applications require accurate positioning information of sensor nodes in Underwater Acoustic Sensor Networks (UASNs). In large-scale UASNs, locating the accurate position of the Mobile Sensor Node (MSN) is difficult without Global Positioning System (GPS) signal. Though the prediction of active MSN location is possible by the received information, predicting the location of a dead MSN without consuming sensor node energy is a challenging task for many decades, which has not gained any practical solution. In literature, the underwater sensor nodes are localized by sending and receiving the beacons from the source to a Base Station (BS) by consuming sensors energy. In this article, an Energy-free Heuristic Neural Network (HNN) Localization approach is proposed using Deep Learning (DL) technique to locate the dead MSN in a large-scale UASN. The proposed HNN leverages the Long- and Short-Term Memory (LSTM) technique to train huge data sets, and introduces Particle Swarm Optimization (PSO) with Repeated Iterative Technique (RFT) to reduce prediction errors during dead reckoning localization. The HNN localization achieves minimum localization error and high accuracy compared to the existing DL approaches. The localization is carried out at the Bay of Bengal, the Indian Ocean, and the Arctic Ocean. The simulation results of HNN model provide a good localization accuracy of 99% at the Bay of Bengal, 98% accuracy at the Indian Ocean, and 97% accuracy at the Arctic Ocean respectively. This model also achieves less time synchronization and low communication overhead without consuming the energy of MSN.

Https://www.ijmtst.com/vol8issue09.html

Many applications in intelligent transportation systems are demanding an accurate web application-based location prediction. In this study, we satisfy this demand by designing an automated mobile user location prediction system based on the well-known traditional Auto-Regressive Integrated Moving Average (ARIMA). To increase the proposed model accuracy, make it dynamic, and reduce its execution time, the traditional ARIMA model has been modified extensively by using different combinations of design options of the model. To perform user location prediction, the proposed model depends the previous recorded user locations to predict the user future locations. To make the proposed model dynamic, it is designed to regenerate all its parameters periodically. To deal with such dynamic environment, only a specified window of the historical data is used. To reduce the regeneration of the model execution time, the model selection process is enhanced and several model selection approaches are proposed. The proposed model and the different design options are evaluated using a realistic user location dataset trace that are recorded using a WIFI embedded, as well as, using traces from a previous study called the Kaggle Dataset. To deal with any imperfection in the data used in generating the model in this study. The results show that the proposed framework can generate ARIMA models that can predict the future user locations of a user accurately and with a reasonable execution time. The results also show that the proposed model can predict the user’s location for several future steps with an acceptable accuracy.

Predicting rupture locations of ascending aortic aneurysms using CT-based finite element models

Accurate rupture risk assessment of ascending aortic aneurysms is important for reducing aneurysm-related mortality. More recently, computational models have been shown to better predict rupture risk than diameter-based measurements. However, it remains unclear whether finite element (FE) models of the ascending aorta can predict rupture location, and over what timeframe those predictions are reliable. The aim of this study was to evaluate FE models of the ascending aorta generated from computed tomography (CT) scans in predicting rupture location. Pre- and post-rupture CT scans were obtained of 12 patients who underwent emergency surgical repair for ascending aorta rupture with varying time intervals between scans (20 days – 6 years). A rigid iterative closest point (ICP) registration was used to overlay post-rupture aortic geometries with pre-rupture FE models and directly compare predicted regions of high equivalent strain with actual rupture. The FE model predicted the rupture location in the 5 patients with the shortest time intervals between the pre- and post-rupture scans (20 days – 2 years, 3 months). However, rupture location was not predicted in the 4/5 patients with greater than 3 years between scans. Achieving a physiological equivalent strain distribution in the FE model was highly dependent on the resolution of the pre-rupture scan and whether contrast agent was present. The results suggest there may be a time interval beyond which FE predictions of rupture location may not be reliable. The results in this study may help clinical validation of FE models of ascending aortic aneurysms predicting rupture risk.

Extended linear regression model for vessel trajectory prediction with a-priori AIS information

ABSTRACT As maritime activities increase globally, there is a greater dependency on technology in monitoring, control, and surveillance of vessel activity. One of the most prominent systems for monitoring vessel activity is the Automatic Identification System (AIS). An increase in both vessels fitted with AIS transponders and satellite and terrestrial AIS receivers has resulted in a significant increase in AIS messages received globally. This resultant rich spatial and temporal data source related to vessel activity provides analysts with the ability to perform enhanced vessel movement analytics, of which a pertinent example is the improvement of vessel location predictions. In this paper, we propose a novel strategy for predicting future locations of vessels making use of historic AIS data. The proposed method uses a Linear Regression Model (LRM) and utilizes historic AIS movement data in the form of a-priori generated spatial maps of the course over ground (LRMAC). The LRMAC is an accurate low complexity first-order method that is easy to implement operationally and shows promising results in areas where there is a consistency in the directionality of historic vessel movement. In areas where the historic directionality of vessel movement is diverse, such as areas close to harbors and ports, the LRMAC defaults to the LRM. The proposed LRMAC method is compared to the Single-Point Neighbor Search (SPNS), which is also a first-order method and has a similar level of computational complexity, and for the use case of predicting tanker and cargo vessel trajectories up to 8 hours into the future, the LRMAC showed improved results both in terms of prediction accuracy and execution time.

Construction of Virtual Interaction Location Prediction Model Based on Distance Cognition

Due to the difference in distance cognition between virtual and real symmetric space, it is difficult for users to accurately interact with the target in the Digital Twin system. In order to study the cross-effects of interaction task, target size and target location on the accuracy of egocentric peripersonal distance cognition, a 2 × 5 × 9 × 5 asymmetric experiment was designed and carried out. There were two kinds of interaction tasks, five kinds of interaction target widths and nine kinds of spatial locations set to estimate the five egocentric peripersonal distances. Based on the experimental data, with interaction task, target width and the actual spatial location as independent variables and virtual interaction location as a dependent variable, the mapping model between the actual physical location and virtual interaction location of different interaction targets was constructed and evaluated by multiple linear regression method. The results showed that the prediction model constructed by stepwise regression method was simple and less computationally intensive, but it had better stability and prediction ability. The correlation coefficients R2 on xp, yp and zp were 0.994, 0.999 and 0.998, RMSE values were 2.583 cm, 1.0774 cm and 1.3155 cm, rRMSE values were 26.57%, 12.60% and 1.15%, respectively. The research of relevant experiments and the construction of models are helpful to solve the layout optimization problem of virtual interactive space in the Digital Twin system.

A comparative study on fatigue indicator parameters for near‐α titanium alloys

AbstractNucleation of in‐service cracks leads to detrimental consequences for structural components of near‐α titanium alloys subjected to fatigue loads. Experimental observations show that the fatigue initiation facets usually form in certain crystallographic orientation ranges of “hard” primary α grains which differ between pure and dwell fatigue loads. In this manuscript, a comparative study has been performed using several fatigue indicator parameters (FIPs) to assess their ability to predict the location of fatigue crack nucleation in near‐α titanium alloy microstructures. All selected FIPs are implemented within the same polycrystalline plasticity finite element modeling framework to facilitate one‐to‐one comparisons. Comparison on predictability of critical initiation locations and their crystallographic orientations is studied for incorporated FIPs under pure and dwell fatigue. The critical locations predicted by some FIPs were found to be close to each other, and consistent with the crystallographic orientation ranges from fractography measurements, in addition to the range transition from pure to dwell fatigue loads. Critical locations from slip driven FIPs are obtained to be several grains away from that of the former ones and are inclined to capture orientations of slip traces from experiments.

A correction for Reynolds‐averaged‐Navier–Stokes turbulence model under the effect of shock waves in hypersonic flows

AbstractStudies on the unphysical increase of turbulent quantities for RANS simulation induced by shock waves in hypersonic flows are carried out. Numerical experiments on the hypersonic flow over a blunt body reveal that the phenomenon of unphysical increase of turbulent quantities across the detached shock wave is induced by the strain‐rate‐based production terms of the k‐ and k‐ SST turbulence models, which leads to the over‐prediction of aerothermal prediction. While this phenomenon does not occur for Spalart–Allmaras (S–A) turbulence model because of its vorticity‐based production term. In order to eliminate this unphysical phenomenon, and to maintain the accuracy of the original models for boundary layer and separation flows, a new correction method for the k‐ and k‐ SST models is proposed: by comparing the orders of magnitude between the strain‐rate‐based and vorticity‐based production terms, the vorticity‐based production term is used near the shock waves, while the original strain‐rate‐based production term is still used in other regions. Finally, the correction method is applied to turbulence and transition flows over blunt bodies, and the numerical results show that the correction method effectively eliminates the unphysical increase of turbulent quantities across shock waves and improves the accuracy of aerothermal and transition onset location prediction.