Short-term Memory Component Research Articles

State monitoring and fault prediction of centrifugal compressors based on long short–term memory and principal component analysis (LSTM-PCA)

Centrifugal compressors are widely used in the petroleum and natural gas industry for gas compression, reinjection, and transportation. Early fault identification and fault evolution prediction for centrifugal compressors can improve equipment safety and reduce maintenance and operating costs. This article proposes a dynamic process monitoring method for centrifugal compressors based on long short-term memory (LSTM) and principal component analysis (PCA). This method constructs a sliding window for monitoring at each sampling point, which contains 100 data from the past and current time points, and uses LSTM to predict 30 future data points. At the same time, this method is also combined with the PCA threshold process monitoring method to construct a new LSTM-PCA monitoring algorithm. And the method was validated using centrifugal compressor process data. The results show that this method can effectively detect process anomalies, The improvements significantly reduced the false positive rate of detected anomalies, and can make multi-step advance predictions of system behavior after faults occur.

Predictability and Variation in Language Are Differentially Affected by Learning and Production.

General principles of human cognition can help to explain why languages are more likely to have certain characteristics than others: structures that are difficult to process or produce will tend to be lost over time. One aspect of cognition that is implicated in language use is working memory-the component of short-term memory used for temporary storage and manipulation of information. In this study, we consider the relationship between working memory and regularization of linguistic variation. Regularization is a well-documented process whereby languages become less variable (on some dimension) over time. This process has been argued to be driven by the behavior of individual language users, but the specific mechanism is not agreed upon. Here, we use an artificial language learning experiment to investigate whether limitations in working memory during either language learning or language production drive regularization behavior. We find that taxing working memory during production results in the loss of all types of variation, but the process by which random variation becomes more predictable is better explained by learning biases. A computational model offers a potential explanation for the production effect using a simple self-primingmechanism.

Charging prediction for new energy electric vehicles in the context of vehicle to grid using a hybrid ROCNN-BILSTM model

Abstract Vehicle to grid refers to the interaction between electric vehicles and the power grid through charging stations. It aims to guide owners of new energy vehicles to charge in an orderly and staggered manner, and even enabling power supply back to the grid. In the context of vehicle to grid, the charging behavior of new energy vehicles becomes different from the past due to uncertainties introduced by user plug-in/plug-out actions and weather conditions, which may disrupt owners’ future scheduling plans. In this article, we propose a charging prediction study based on the Reordering Convolutional Neural Network-Bidirectional Long Short-Term Memory (ROCNN-BILSTM) hybrid model specifically designed for the vehicle to grid context. The proposed model employs wavelet threshold denoising as a data preprocessing operation to remove unnecessary noise factors that could affect predictions. Subsequently, the 2-Dimensional Convolutional Neural Network (2D-CNN) component retains temporal features while extracting spatial features. Notably, the features are rearranged, combining highly correlated ones, to facilitate the extraction of high-level, abstract spatial features by the 2D-CNN. Finally, the Bidirectional Long Short-Term Memory (BILSTM) component utilizes a bidirectional structure to capture comprehensive dynamic information and assist in achieving the final charging prediction. Our proposed ROCNN-BILSTM eliminates uncertainty in the data, allowing deep learning models to better focus on important features. Additionally, our model emphasizes high-level spatiotemporal feature extraction, which helps achieve high-performance charging prediction. In the context of vehicle to grid, a real-world dataset of new energy vehicle charging data was used for multi-step prediction, different starting point predictions, and comparison with advanced models. The experimental results show that the proposed model outperforms CNN-LSTM and 2D-CNN models by up to 50.1% and 57.1% in terms of mean absolute error (MAE), and 45.8% and 51.5% in terms of mean squared error (MSE). The results validate the strong predictive performance of the hybrid model and provide robust support for the demands of the vehicle to grid market and new energy vehicle charging prediction technology. In future work, we will place greater emphasis on designing high-performance and interpretable models to explore the fundamental reasons behind different charging trends in new energy vehicles.

The Effectiveness of Training the Components of Compassion-Focused Therapy on Memory, Anxiety, and Sleep Quality in Patients with Obstructive Sleep Apnea

Background: Obstructive sleep apnea (OSA) is a common and debilitating sleep disorder with both mental and physical complications. Although medical treatments may improve OSA symptoms, they may not affect its mental sequelae. Objectives: The aim of this study was to evaluate the effect of training the components of compassion-focused therapy (CFT) on sleep quality, anxiety, and memory in patients with OSA. Methods: This experimental study was performed on 37 patients (18 to 65 years old) with confirmed OSA diagnosed in the past year with a minimum apnea index score of 5 and no history of medical or psychiatric diseases. The study was conducted between 2022 and 2023. Exclusion criteria were being absent from more than two training sessions, experiencing stressful events, and refusing to continue the study. The intervention group (n = 19) received the training component of CFT in 8 sessions (once a week), while the control group received no psychological intervention. The Pittsburg sleep quality index (PSQI), Rey Auditory-Verbal Learning Test (RAVLT), Rey-Osterrieth complex figure (ROCF), and Beck Anxiety Inventory (BAI), as well as a demographic information questionnaire, were filled out by the participants at the baseline and after the intervention. Results: There was no significant difference between the groups in terms of demographic variables. The training component of CFT was effective in reducing anxiety and improving sleep quality, as well as the auditory-verbal and visual-spatial components of short-term memory in patients with OSA (P < 0.05). Conclusions: The training component of CFT was efficient and beneficial in reducing anxiety and improving immediate memory and sleep quality in patients with OSA.

NIM: Generative Neural Networks for Automated Modeling and Generation of Simulation Inputs

Fitting stochastic input-process models to data and then sampling from them are key steps in a simulation study but highly challenging to non-experts. We present Neural Input Modeling (NIM), a Generative Neural Network (GNN) framework that exploits modern data-rich environments to automatically capture simulation input processes and then generate samples from them. The basic GNN that we develop, called NIM-VL , comprises (i) a variational autoencoder architecture that learns the probability distribution of the input data while avoiding overfitting and (ii) long short-term memory components that concisely capture statistical dependencies across time. We show how the basic GNN architecture can be modified to exploit known distributional properties—such as independent and identically distributed structure, nonnegativity, and multimodality—to increase accuracy and speed, as well as to handle multivariate processes, categorical-valued processes, and extrapolation beyond the training data for certain nonstationary processes. We also introduce an extension to NIM called Conditional Neural Input Modeling (CNIM), which can learn from training data obtained under various realizations of a (possibly time series valued) stochastic “condition,” such as temperature or inflation rate, and then generate sample paths given a value of the condition not seen in the training data. This enables users to simulate a system under a specific working condition by customizing a pre-trained model; CNIM also facilitates what-if analysis. Extensive experiments show the efficacy of our approach. NIM can thus help overcome one of the key barriers to simulation for non-experts.

Auditory Short-Term Memory Evaluation in Noise in Musicians.

Working memory, a short-term memory component, is a multicomponent system that manages attention and short-term memory in speech perception in challenging listening conditions. These challenging conditions cause listening effort that can be objectively evaluated by pupillometry. Studies show that auditory working memory is more developed in musicians for complex auditory tasks. This study aims to compare the listening effort and short-term memory in noise between musicians and nonmusicians. An experimental research design was adopted for the study. The study was conducted on 22 musicians and 20 nonmusicians between the ages of 20 and 45. Participants' effort analysis was measured with pupillometry; performance analysis was measured with short-term memory score by listening to the 15 word lists of Verbal Memory Processes Test. Participants are tested under three conditions: quiet, +15 signal-to-noise ratio (SNR), and +5 SNR. While nonmusicians showed significantly higher short-term memory score (STMS) than musicians in the quiet condition, musicians' STMS were significantly higher in both noise conditions (+15 SNR and +5 SNR). The nonmusician's percentage of pupil growth averages were higher than the musicians for three conditions. As a result, musicians had better memory performance in noise and less effort in the listening task according to lower pupil growth. This study objectively evaluated the differences between participants' listening efforts by pupillometry. It is also observed that the SNR and music training affect memory performance.

Electricity Consumption Forecasting System Using Deep Learning

Energy modeling in Smart Buildings (SB) and planning and operating the generation of power based on the information extracted are key components of the Smart Grid’s (SG’s) energy management system. In the world, buildings use a significant amount of energy that contributes to energy efficiency programs. Additionally, excessive utilization power generation appliance also including air conditioners and heater, breathing, and climate control (HVAC) units, improper microclimate control, and inappropriate start-up and ordering of power equipment waste a lot of heat pumps. The utility can mitigate energy generation costs when it anticipates electrical loads and schedules generation resources in accordance with the demand. To estimate electricity usage at varying tiers of utility grid systems, a range of techniques have already been used. The goal of this study will be to create a hybrid deep learning model can predict resource utilization in infrastructures. Model building and data cleaning are the two stages of the proposed framework. Data cleaning involves pre-processing techniques and adding additional lag values to raw data. This hybrid deep learning (DL) approach is made up of a series of completely connected layered and linear Long Short-Term Memory (LSTM) sections layered over bi-directional Long Short-Term Memory (LSTM) components and is based also on collected information. You could incorporate the dependence structure of electricity usage on regressors and increase computation efficiency, training time, as well as computational complexity by using the results obtained.

Lateralization of short-term memory in the frontal cortex.

Despite essentially symmetric structures in mammalian brains, the left and right hemispheres do not contribute equally to certain cognitive functions. How both hemispheres interact to cause this asymmetry remains unclear. Here, we study this question in the anterior lateral motor cortex (ALM) of mice performing five versions of a tactile-based decision-making task with a short-term memory (STM) component. Unilateral inhibition of ALM produces variable behavioral deficits across tasks, with the left, right, or both ALMs playing critical roles in STM. Neural activity and its encoding capability are similar across hemispheres, despite that only one hemisphere dominates in behavior. Inhibition of the dominant ALM disrupts encoding capability in the non-dominant ALM, but not vice versa. Variable behavioral deficits are predicted by the influence on contralateral activity across sessions, mice, and tasks. Together, these results reveal that the left and right ALM interact asymmetrically, leading to their differential contributions to STM.

A multi-output deep learning model based on Bayesian optimization for sequential train delays prediction

ABSTRACT Accurate train arrival delay predictions can provide timely information for passengers and train dispatchers. Previous work mainly focused on predicting the delay of a single train, which is not enough to assist dispatchers, because making more comprehensive decisions considering more trains needs more future delay information of a group of trains. Therefore, this paper proposes a Bayesian optimization-based multi-output deep learning model, which includes a fully connected neural network (FCNN) and two long–short-term memory (LSTM) components, to predict the arrival delays of multiple trains simultaneously. The proposed model is calibrated and validated with the train operation data from the Wuhan–Guangzhou (W-G) high-speed railway. The test results show that the mean absolute error and the root mean square error of the proposed model is 1.379 and 2.021 min, over the four subsequent trains. Moreover, the proposed model outperforms the standard train delay prediction benchmark model.

Gait Synergy Analysis and Modeling on Amputees and Stroke Patients for Lower Limb Assistive Devices.

The concept of synergy has drawn attention and been applied to lower limb assistive devices such as exoskeletons and prostheses for improving human–machine interaction. A better understanding of the influence of gait kinematics on synergies and a better synergy-modeling method are important for device design and improvement. To this end, gait data from healthy, amputee, and stroke subjects were collected. First, continuous relative phase (CRP) was used to quantify their synergies and explore the influence of kinematics. Second, long short-term memory (LSTM) and principal component analysis (PCA) were adopted to model interlimb synergy and intralimb synergy, respectively. The results indicate that the limited hip and knee range of motions (RoMs) in stroke patients and amputees significantly influence their synergies in different ways. In interlimb synergy modeling, LSTM (RMSE: 0.798° (hip) and 1.963° (knee)) has lower errors than PCA (RMSE: 5.050° (hip) and 10.353° (knee)), which is frequently used in the literature. Further, in intralimb synergy modeling, LSTM (RMSE: 3.894°) enables better synergy modeling than PCA (RMSE: 10.312°). In conclusion, stroke patients and amputees perform different compensatory mechanisms to adapt to new interlimb and intralimb synergies different from healthy people. LSTM has better synergy modeling and shows a promise for generating trajectories in line with the wearer’s motion for lower limb assistive devices.

A deep learning method for predicting microvoid growth in heterogeneous polycrystals

In heterogeneous polycrystals, microvoid growth presents inherent randomness and dispersion, which generally follows a statistical law. This statistical characteristic intrinsically arises from randomly distributed grain-orientations around microvoids. It remains a huge challenge to explicitly depict the inherent correlation between microvoid growth and grain-orientation distribution by conventional deterministic damage models. In recent years, deep learning has gained in popularity in materials science and has been demonstrated to exhibit excellent data-mining abilities. To our best knowledge, deep learning has not been applied to investigate the statistical damage-evolution issues hitherto. In this work, a novel microvoid growth model based on deep neural network is creatively designed, incorporating both convolutional and long short-term memory components. The former extracts the spatial grain-orientation information, and the latter captures the causal effect of strain history on the microvoid growth. Moreover, to train and test the deep learning-based model, a microvoid-growth database is generated through a large number of crystal plasticity-based finite element simulations, incorporating randomly-oriented grains and different void locations. All the sample data (i.e., the grain-orientation distributions, microvoid locations and microvoid-growth curves) are processed by specific methods (e.g., the pixel-based method) to be amenable for the training process. Our results show that this novel model well captures the statistical characteristic of the microvoid growth in heterogeneous polycrystals. It is expected that the deep learning-based method can provide a new way to predict the microvoid growth at the grain-level.

Extraction of binary black hole gravitational wave signals from detector data using deep learning

Accurate extractions of the detected gravitational wave (GW) signal waveforms are essential to validate a detection and to probe the astrophysics behind the sources producing the GWs. This however could be difficult in realistic scenarios where the signals detected by existing GW detectors could be contaminated with non-stationary and non-Gaussian noise. While the performance of existing waveform extraction methods are optimal, they are not fast enough for online application, which is important for multi-messenger astronomy. In this paper, we demonstrate that a deep learning architecture consisting of Convolutional Neural Network and bidirectional Long Short-Term Memory components can be used to extract binary black hole (BBH) GW waveforms from realistic noise in a few milli-seconds. We have tested our network systematically on injected GW signals, with component masses uniformly distributed in the range of 10 to 80 solar masses, on Gaussian noise and LIGO detector noise. We find that our model can extract GW waveforms with overlaps of more than 0.95 with pure Numerical Relativity templates for signals with signal-to-noise ratio (SNR) greater than six, and is also robust against interfering glitches. We then apply our model to all ten detected BBH events from the first (O1) and second (O2) observation runs, obtaining greater than 0.97 overlaps for all ten extracted BBH waveforms with the corresponding pure templates. We discuss the implication of our result and its future applications to GW localization and mass estimation.

Deep convolutional recurrent model for region recommendation with spatial and temporal contexts

Spatiotemporal-aware region recommendation satisfies a user by providing an region of POIs (point-of-interests) that he/she may prefer. This recommendation is typically performed by analyzing the region mobility patterns of the user with some spatial and temporal contexts. This kind of recommendation can help, for example, a businessman to enjoy his urban life, or a tourist to travel in an unfamiliar area. In this study, we propose a deep-learning framework to model region-level mobility patterns of users, where personal and global user preferences across regions as well as spatiotemporal dependencies are comprehensively incorporated. To be specific, we model user preferences through a pyramidal Convolutional Long Short-Term Memory (ConvLSTM) component, and induce the dynamic region attributes through a recurrent component. By fusing two components to recommend next time region, our framework can tackle three complex challenges: (1) Modeling users’ distinctive spatio-temporal preferences over regions; (2) tracing diverse region mobility patterns of users over time; and (3) capturing the intrinsic correlations between regions. Extensive experiments on real-world datasets validate the effectiveness of the novel approach.

Modeling Global Spatial–Temporal Graph Attention Network for Traffic Prediction

Accurate and efficient traffic prediction is the key to the realization of intelligent transportation system (ITS), which helps to alleviate traffic congestion and reduce traffic accidents. Due to the complex dynamic spatial-temporal dependence between traffic networks, traffic prediction is extremely challenging. In previous studies, convolution neural network (CNN) and graph convolution network (GCN) were used to model spatial correlation. However, the non-Euclidean correlation of road network reduces the effect of convolution operator modeling. In addition, only considering the traffic interaction around the concerned points simplifies the influence of traffic network. In order to address the above problems, this article proposes an end-to-end global spatial-temporal graph attention network (GST-GAT), which uses the “global interaction + node query” to model the dynamic spatial-temporal correlation of traffic. In the encoder, the long short-term memory (LSTM) component flexibly transforms the traffic dynamic spatial-temporal graph into feedforward differentiable features. Global traffic interaction is proposed to summarize traffic network context changes and integrate all node features at each moment through a forward calculation. Then, each node computes the influence of traffic global interaction on a single node in parallel, and the spatial-temporal interaction information is adaptive fused by gating fusion mechanism. Finally, the end-to-end network structure is used to train the rich mixed feature coding to generate the traffic prediction status of each node. Experiments on public transportation data sets show that GST-GAT performs better than previous work in terms of accuracy and inference speed.

A Deep Learning Model for Automated Classification of Intraoperative Continuous EMG.

Intraoperative neurophysiological monitoring (IONM) is the use of electrophysiological methods during certain high-risk surgeries to assess the functional integrity of nerves in real time and alert the surgeon to prevent damage. However, the efficiency of IONM in current practice is limited by latency of verbal communications, inter-rater variability, and the subjective manner in which electrophysiological signals are described. In an attempt to address these shortcomings, we investigate automated classification of free-running electromyogram (EMG) waveforms during IONM. We propose a hybrid model with a convolutional neural network (CNN) component and a long short-term memory (LSTM) component to better capture complicated EMG patterns under conditions of both electrical noise and movement artifacts. Moreover, a preprocessing pipeline based on data normalization is used to handle classification of data from multiple subjects. To investigate model robustness, we also analyze models under different methods for processing of artifacts. Compared with several benchmark modeling methods, CNN-LSTM performs best in classification, achieving accuracy of 89.54% and sensitivity of 94.23% in cross-patient evaluation. The CNN-LSTM model shows promise for automated classification of continuous EMG in IONM. This technique has potential to improve surgical safety by reducing cognitive load and inter-rater variability.

Modeling train operation as sequences: A study of delay prediction with operation and weather data

This paper presents a carefully designed train delay prediction model, called FCLL-Net, which combines a fully-connected neural network (FCNN) and two long short-term memory (LSTM) components, to capture operational interactions. The performance of FCLL-Net is tested using data from two high speed railway lines in China. The results show that FCLL-Net has significantly improved prediction performance, over 9.4% on both lines, in terms of the selected absolute and relative metrics compared to the commonly used state-of-the-art models. Additionally, the sensitivity analysis demonstrates that interactions of train operations and weather-related features are of great significance to consider in delay prediction models.

Bubbles in turbulent flows: Data-driven, kinematic models with history terms

We present data driven kinematic models for the motion of bubbles in high-Re turbulent fluid flows based on recurrent neural networks with long-short term memory enhancements. The models extend empirical relations, such as Maxey-Riley (MR) and its variants, whose applicability is limited when either the bubble size is large or the flow is very complex. The recurrent neural networks are trained on the trajectories of bubbles obtained by Direct Numerical Simulations (DNS) of the Navier Stokes equations for a two-component incompressible flow model. Long short term memory components exploit the time history of the flow field that the bubbles have encountered along their trajectories and the networks are further augmented by imposing rotational invariance to their structure. We first train and validate the formulated model using DNS data for a turbulent Taylor-Green vortex. Then we examine the model predictive capabilities and its generalization to Reynolds numbers that are different from those of the training data on benchmark problems, including a steady (Hill’s spherical vortex) and an unsteady (Gaussian vortex ring) flow field. We find that the predictions of the developed model are significantly improved compared with those obtained by the MR equation. Our results indicate that data-driven models with history terms are well suited in capturing the trajectories of bubbles in turbulent flows.

HABNet: Machine Learning, Remote Sensing-Based Detection of Harmful Algal Blooms

This article describes the application of machine learning techniques to develop state-of-the-art detection and prediction system for spatiotemporal events found within remote sensing data; specifically, harmful algal bloom (HAB) events. We propose HAB detection system based on a ground truth historical record of HAB events, a novel spatiotemporal datacube representation of each event (from MODIS and GEBCO bathymetry data), and a variety of machine learning architectures utilizing the state-of-the-art spatial and temporal analysis methods based on convolutional neural networks, long short-term memory components together with random forest, and support vector machine classification methods. This work has focused specifically on the case study of the detection of Karenia brevis algae ( K. brevis ) HAB events within the coastal waters of Florida (over 2850 events from 2003 to 2018; an order of magnitude larger than any previous machine learning detection study into HAB events). The development of multimodal spatiotemporal datacube data structures and associated novel machine learning methods give a unique architecture for the automatic detection of environmental events. Specifically, when applied to the detection of HAB events, it gives a maximum detection accuracy of 91% and a Kappa coefficient of 0.81 for the Florida data considered. A HAB forecast system was also developed where a temporal subset of each datacube was used to predict the presence of a HAB in the future. This system was not significantly less accurate than the detection system being able to predict with 86% accuracy up to 8 d in the future.

Recurrent Compressed Convolutional Networks for Short Video Event Detection

Short videos are popular information carriers on the Internet, and detecting events from them can well benefit widespread applications, e.g., video browsing, management, retrieval and recommendation. Existing video analysis methods always require decoding all frames of videos in advance, which is very costly in time and computation power. These short videos are often untrimmed, noisy and even incomplete, adding much difficulty to event analysis. Unlike previous works focusing on actions, we target short video event detection and propose Recurrent Compressed Convolutional Networks (RCCN) for discovering the underlying event patterns within short videos possibly including a large proportion of non-event videos. Instead of using the whole videos, RCCN performs representation learning at much lower cost within the compressed domain where the encoded motion information reflecting the spatial relations among frames can be easily obtained to capture dynamic tendency of event videos. This alleviates the information incompleteness problem that frequently emerges in user-generated short videos. In particular, RCCN leverages convolutional networks as the backbone and the Long Short-Term Memory components to model the variable-range temporal dependency among untrimmed video frames. RCCN not only learns the common representation shared by the short videos of the same event, but also obtains the discriminative ability to detect dissimilar videos. We benchmark the model performance on a set of short videos generated from publicly available event detection database YLIMED, and compare RCCN with several baselines and state-of-the-art alternatives. Empirical studies have verified the preferable performance of RCCN.

RPM-Net

We introduce RPM-Net, a deep learning-based approach which simultaneously infers movable parts and hallucinates their motions from a single, un-segmented, and possibly partial, 3D point cloud shape. RPM-Net is a novel Recurrent Neural Network (RNN), composed of an encoder-decoder pair with interleaved Long Short-Term Memory (LSTM) components, which together predict a temporal sequence of pointwise displacements for the input point cloud. At the same time, the displacements allow the network to learn movable parts, resulting in a motion-based shape segmentation. Recursive applications of RPM-Net on the obtained parts can predict finer-level part motions, resulting in a hierarchical object segmentation. Furthermore, we develop a separate network to estimate part mobilities, e.g., per-part motion parameters, from the segmented motion sequence. Both networks learn deep predictive models from a training set that exemplifies a variety of mobilities for diverse objects. We show results of simultaneous motion and part predictions from synthetic and real scans of 3D objects exhibiting a variety of part mobilities, possibly involving multiple movable parts.