Search

Abstract While the sensorimotor cortices are central neural substrates for motor control and learning, how the interaction between their subregions with visual cortices contributes to acquiring de novo visuomotor skills is poorly understood. We design a continuous visuomotor task in fMRI where participants control a cursor using their fingers while learning an arbitrary finger-to-cursor mapping. To investigate visuomotor interaction in the de novo motor task, we manipulate visual feedback of a cursor such that they learn to control using fingers under two alternating conditions: online cursor feedback is available or unavailable except when a target is reached. As a result, we find double dissociation of fMRI activity in subregions of the sensorimotor and visual cortices. Specifically, motor and late visual cortices are more active with online cursor feedback, and somatosensory and early visual cortices are more active without online cursor feedback. We also find a significant reduction in functional connectivity between somatosensory cortices and early visual cortices, which is highly correlated with performance improvement. These findings support the distinct interaction between subregions of sensorimotor cortices and visual cortices, while the connectivity analysis highlights the critical role of somatosensory cortices during de novo motor learning.

Abstract Corticostriatal regions play a pivotal role in visuomotor learning. However, less research has been done on how fMRI activity in their subregions is related to task performance, which is provided as visual feedback during motor learning. To address this, we conducted an fMRI experiment in which participants acquired a complex de novo motor skill using continuous or binary visual feedback related to performance. We found a highly selective response related to performance in the entire striatum in both conditions and a relatively higher response in the caudate nucleus for the binary feedback condition. However, the ventromedial prefrontal cortex (vmPFC) response was significant only for the continuous feedback condition. Furthermore, we also found functional distinction of the striatal subregions in random versus goal-directed motor control. These findings underscore the substantial effects of the visual feedback indicating performance on distinct corticostriatal responses, thereby elucidating its significance in reinforcement-based motor learning.

Abstract While zero-phase lag synchronization between multiple brain regions has been widely observed, relatively recent reports indicate that systematic phase delays between cortical regions reflect the direction of communications between cortical regions. For example, it has been suggested that a non-zero phase delay of electroencephalography (EEG) signals at the gamma frequency band between the bilateral parietal areas may reflect the direction of communication between these areas. We hypothesized that the direction of communication between distant brain areas might be modulated by multi-site transcranial alternating current stimulation (tACS) with specific phase delays other than 0° and 180°. In this study, a new noninvasive brain stimulation (NIBS) method called multi-site multi-phase tACS (msmp-tACS) was proposed. The efficacy of the proposed method was tested in a case study using a visuospatial working memory (VWM) paradigm in which the optimal stimulation conditions including amplitudes and phases of multiple scalp electrodes were determined using finite element analysis adopting phasor representation. msmp-tACS was applied over the bilateral intraparietal sulci (IPS) and showed that 80 Hz tACS with the phase for the right IPS leading that for the left IPS by 90° (= 3.125 ms) partialized VWM performance toward the right visual hemifield. The three stimulation conditions were synchronized, RL, and LR, which refers to stimulation condition with no phase lag, stimulation phase of right IPS (rIPS) leading left IPS (lIPS) by 90° and the stimulation of lIPS leading rIPS by 90°, respectively. The lateralization of VWM significantly shifted towards right visual hemifield under the RL condition compared to the synchronized and LR conditions. The shift in VWM was the result of the stimulation affecting both left and right visual hemifield trials to certain degrees, rather than significantly increasing or decreasing VWM capacity of a specific visual hemifield. Altered brain dynamics caused by msmp-tACS partialized VWM performance, likely due to modulation of effective connectivity between the rIPS and lIPS. Our results suggest that msmp-tACS is a promising NBS method that can effectively modulate cortical networks that cannot be readily modulated with conventional multi-site stimulation methods.

This paper proposes a method for forecasting surface solar irradiance (SSI), the most critical factor in solar photovoltaic (PV) power generation. The proposed method uses 16-channel data obtained by the GEO-KOMPSAT-2A (GK2A) satellite of South Korea as the main data for SSI forecasting. To determine feature variables related to SSI from the 16-channel data, the differences and ratios between the channels were utilized. Additionally, to consider the fundamental characteristics of SSI originating from the sun, solar geometry parameters, such as solar declination (SD), solar elevation angle (SEA), and extraterrestrial solar radiation (ESR), were used. Deep learning-based feature selection (Deep-FS) was employed to select appropriate feature variables that affect SSI from various feature variables extracted from the 16-channel data. Lastly, spatio-temporal deep learning models, such as convolutional neural network–long short-term memory (CNN-LSTM) and CNN–gated recurrent unit (CNN-GRU), which can simultaneously reflect temporal and spatial characteristics, were used to forecast SSI. Experiments conducted to verify the proposed method against conventional methods confirmed that the proposed method delivers superior SSI forecasting performance.

This study proposes a deep transfer learning method using a deep convolutional neural network pre-trained with ImageNet for transient stability assessment. The procedure of deep transfer learning, incorporating the role and considerations of the transient stability assessment system, is suggested. The transient assessment system learns the relationship between the severity of disturbances and transient stability in the power system through the proposed training method. The severity of a disturbance based on the physical causal relationship of the angle stability is proposed to train and implement the transient stability assessment model. The power system state variables obtained from the phasor measurement unit are converted into the feature map described by the severity of a disturbance, enabling the training of transient stability characteristics of the power system to the deep convolutional neural network. The training dataset is constructed using the time-domain simulation on IEEE 39 and IEEE 118-bus benchmark power system models configured in MATLAB/Simulink. As a result of deep transfer learning, which involves training with freezing some of the convolutional layers of the pre-trained deep network, the most suitable model for transient stability assessment is selected among the pre-trained deep networks. The effectiveness of the proposed method is compared with other approaches using the confusion matrix, and the robustness against noise interference is also investigated.

In recent years, battery fires have become more common owing to the increased use of lithium-ion batteries. Therefore, monitoring technology is required to detect battery anomalies because battery fires cause significant damage to systems. We used Mahalanobis distance (MD) and independent component analysis (ICA) to detect early battery faults in a real-world energy storage system (ESS). The fault types included historical data of battery overvoltage and humidity anomaly alarms generated by the system management program. These are typical preliminary symptoms of thermal runaway, the leading cause of lithium-ion battery fires. The alarms were generated by the system management program based on thresholds. If a fire occurs in an ESS, the humidity inside the ESS will increase very quickly, which means that threshold-based alarm generation methods can be risky. In addition, industrial datasets contain many outliers for various reasons, including measurement and communication errors in sensors. These outliers can lead to biased training results for models. Therefore, we used MD to remove outliers and performed fault detection based on ICA. The proposed method determines confidence limits based on statistics derived from normal samples with outliers removed, resulting in well-defined thresholds compared to existing fault detection methods. Moreover, it demonstrated the ability to detect faults earlier than the point at which alarms were generated by the system management program: 15 min earlier for battery overvoltage and 26 min earlier for humidity anomalies.

Machine tools are used in a wide range of applications, and they can manufacture workpieces flexibly. Furthermore, they require maintenance; the overall costs include maintenance costs, which constitute a significant portion, and the costs involved in ensuring product quality. Therefore, anomaly detection in tool conditions is required, because these tools are essential industrial elements. However, the data related to tool conditions present some challenges: data imbalances and deficiencies. Data imbalances and deficiencies can affect the performance of anomaly detection models. A model trained using data with imbalances and deficiencies may miscalculate that abnormal data are normal data, leasing to errors. To overcome these problems, the proposed method has been designed using the wavelet transform, color space conversion, color extraction, puzzle-based data augmentation, and double transfer learning. The proposed method generated image data from time-series data, effectively extracted features, and generated new image data using puzzle-based data augmentation. The color information was processed to highlight features, and the proposed puzzle-based data augmentation was applied during processing to increase the amount of data to improve the performance of the anomaly detection model. The experimental results showed that the proposed method can classify normal and abnormal data with greater accuracy. In particular, the accuracy of abnormal data classification increased from 25.00% to 91.67%. This demonstrates that the proposed method is effective and can overcome data imbalances and deficiencies.

Confronted with the climate crisis, the world is making tremendous efforts in energy transition, such as expanding renewable energy that does not emit carbon. The importance of virtual power plant (VPP) operation technology has emerged to secure grid flexibility in response to the expanding renewable energy implemented due to these efforts. Accordingly, VPPs, which include photovoltaics, wind turbines, heating, ventilation, and air conditioning (HVAC), load, and EV, have been constructed. HVAC, one of the component resources, is a system that controls and regulates temperature, humidity, and airflow. Since it responds sensitively to the building’s heat capacity and changes in the external environment, it requires continuous and stable control. In this paper, we used data-based modeling to implement the HVAC required for the optimal operation of VPP. Since accurately creating an equation-based HVAC model was difficult considering building information modeling and external environment variables, we used historical HVAC operation data to perform data-based modeling. The model was implemented using nonlinear regression and machine learning, such as a support vector machine and artificial neural network. Then, the data-based HVAC and the actual HVAC operation results were comparatively analyzed based on a case study, and the model’s goodness-of-fit was evaluated based on performance metrics. Model performance indicators confirmed that the ANN-based HVAC model was most similar to the actual HVAC system.

Particulate matter (PM) in the air can cause various health problems and diseases in humans. In particular, the smaller size of PM2.5 enable them to penetrate deep into the lungs, causing severe health impacts. Exposure to PM2.5 can result in respiratory, cardiovascular, and allergic diseases, and prolonged exposure has also been linked to an increased risk of cancer, including lung cancer. Therefore, forecasting the PM2.5 concentration in the surrounding is crucial for preventing these adverse health effects. This paper proposes a method for forecasting the PM2.5 concentration after 1 h using bidirectional long short-term memory (Bi-LSTM). The proposed method involves selecting input variables based on the feature importance calculated by random forest, classifying the data to assign weight variables to reduce bias, and forecasting the PM2.5 concentration using Bi-LSTM. To compare the performance of the proposed method, two case studies were conducted. First, a comparison of forecasting performance according to preprocessing. Second, forecasting performance between deep learning (long short-term memory, gated recurrent unit, and Bi-LSTM) and conventional machine learning models (multi-layer perceptron, support vector machine, decision tree, and random forest). In case study 1, The proposed method shows that the performance indices (RMSE: 3.98%p, MAE: 5.87%p, RRMSE: 3.96%p, and R2:0.72%p) are improved because weights are given according to the input variables before the forecasting is performed. In case study 2, we show that Bi-LSTM, which considers both directions (forward and backward), can effectively forecast when compared to conventional models (RMSE: 2.70, MAE: 0.84, RRMSE: 1.97, R2: 0.16). Therefore, it is shown that the proposed method can effectively forecast PM2.5 even if the data in the high-concentration section is insufficient.

In this manuscript, we are reporting for the first time one dimensional (1D) cerium hydrogen phosphate (Ce(HPO4)2.xH2O) electrode material for supercapacitor application. In short, a simple hydrothermal technique was employed to prepare Ce(HPO4)2.xH2O. The maximum surface area of 82 m2 g−1 was obtained from nitrogen sorption isotherm. SEM images revealed Ce(HPO4)2.xH2O exhibited a nanorod-like structure along with particles and clusters. The maximum specific capacitance of 114 F g−1 was achieved at 0.2 A g−1 current density for Ce(HPO4)/NF electrode material in a three-electrode configuration. Furthermore, the fabricated symmetric supercapacitor (SSC) based on Ce(HPO4)2.xH2O//Ce(HPO4)2.xH2O demonstrates reasonable specific energy (2.08 Wh kg−1), moderate specific power (499.88 W kg−1), and outstanding cyclic durability (retains 92.7% of its initial specific capacitance after 5000 GCD cycles).

In modern complex industrial processes, mode changes cause unplanned shutdowns, potentially shortening the lifespan of key equipment and incurring significant maintenance costs. To avoid this problem, a method that can detect the fault of equipment operating in various modes is required. Therefore, we propose a novel fault detection method that uses the k-nearest neighbor normalization-based weight local outlier factor (WLOF). The proposed method performs local normalization using neighbors to consider possible mode changes in the normal data and WLOF is used for fault detection. In contrast to statistical methods, such as principal component analysis (PCA) and independent component analysis (ICA), the local outlier factor (LOF) uses the density of neighbors. However, because LOF is significantly affected by the distance between its neighbors, the weight is multiplied proportionally to the distance between each neighbor to improve the fault detection performance of the LOF. The efficiency of the proposed method was evaluated using a multimode numerical case and a circulating fluidized bed boiler. The experimental results show that the proposed method outperforms conventional PCA, kernel PCA (KPCA), k-nearest neighbor (kNN), and LOF. In particular, the proposed method improved the detection accuracy by 20% compared with conventional methods. Therefore, the proposed method can be applied to a real process operating in multiple modes.

Artificial intelligence technologies and robot vision systems are core technologies in smart factories. Currently, there is scholarly interest in automatic data feature extraction in smart factories using deep learning networks. However, sufficient training data are required to train these networks. In addition, barely perceptible noise can affect classification accuracy. Therefore, to increase the amount of training data and achieve robustness against noise attacks, a data augmentation method implemented using the adaptive inverse peak signal-to-noise ratio was developed in this study to consider the influence of the color characteristics of the training images. This method was used to automatically determine the optimal perturbation range of the color perturbation method for generating images using weights based on the characteristics of the training images. The experimental results showed that the proposed method could generate new training images from original images, classify noisy images with greater accuracy, and generally improve the classification accuracy. This demonstrates that the proposed method is effective and robust to noise, even when the training data are deficient.

Episodic memory is thought to involve functional interactions of large-scale brain networks that dynamically reconfigure depending on task demands. Although the hippocampus and closely related structures have been implicated, little is known regarding how large-scale and distributed networks support different memory formation demands. We investigated patterns of interactions among distributed networks while human individuals formed item-context memories for two stimulus categories. Subjects studied object-scene and object-location associations in different fMRI sessions. Stimulus-responsive brain regions were organized based on their fMRI interconnectivity into networks and modules using probabilistic module-detection algorithms to maximize measurement of individual differences in modular structure. Although there was a great deal of consistency in the modular structure between object-scene and object-location memory formation, there were also significant differences. Interactions among functional modules predicted later memory accuracy, explaining substantial portions of variability in memory formation success. Increased interactivity of modules associated with internal thought and anti-correlation of these modules with those related to stimulus-evoked processing robustly predicted object-scene memory, whereas decreased interactivity of stimulus-evoked processing modules predicted object-location memory. Assessment of individual differences in network organization therefore allowed identification of distinct patterns of functional interactions that robustly predicted memory formation. This highlights large-scale brain network interactions for memory formation and indicates that although networks are largely robust to task demands, reconfiguration nonetheless occurs to support distinct memory formation demands.

[This corrects the article DOI: 10.1371/journal.pone.0210167.].

A weather radar is a frequently used device in remote sensing to identify meteorological phenomena using electromagnetic waves. It can observe atmospheric conditions in a wide area with a remarkably high spatiotemporal resolution, and its observation results are useful to meteorological research and services. Recent research on data analysis using radar data has concentrated on applying machine learning techniques to solve complicated problems, including quality control, quantitative precipitation estimation, and convective storm prediction. Convective storms, which consist of heavy rains and winds, are closely related to real-life and cause significant loss of life and property. This paper proposes a novel approach utilizing the given convective storms’ temporal properties based on machine learning models to predict future locations. The experimental results showed that the machine learning-based prediction models are capable of nowcasting future locations of convective storms with a slight difference.

Due to the recent development of information and communication technology (ICT), various studies using real-time data are now being conducted. The microgrid research field is also evolving to enable intelligent operation of energy management through digitalization. Problems occur when operating the actual microgrid, causing issues such as difficulty in decision making and system abnormalities. Using digital twin technology, which is one of the technologies representing the fourth industrial revolution, it is possible to overcome these problems by changing the microgrid configuration and operating algorithms of virtual space in various ways and testing them in real time. In this study, we proposed an energy storage system (ESS) operation scheduling model to be applied to virtual space when constructing a microgrid using digital twin technology. An ESS optimal charging/discharging scheduling was established to minimize electricity bills and was implemented using supervised learning techniques such as the decision tree, NARX, and MARS models instead of existing optimization techniques. NARX and decision trees are machine learning techniques. MARS is a nonparametric regression model, and its application has been increasing. Its performance was analyzed by deriving performance evaluation indicators for each model. Using the proposed model, it was found in a case study that the amount of electricity bill savings when operating the ESS is greater than that incurred in the actual ESS operation. The suitability of the model was evaluated by a comparative analysis with the optimization-based ESS charging/discharging scheduling pattern.

Deep learning is applied in various manufacturing domains. To train a deep learning network, we must collect a sufficient amount of training data. However, it is difficult to collect image datasets required to train the networks to perform object recognition, especially because target items that are to be classified are generally excluded from existing databases, and the manual collection of images poses certain limitations. Therefore, to overcome the data deficiency that is present in many domains including manufacturing, we propose a method of generating new training images via image pre-processing steps, background elimination, target extraction while maintaining the ratio of the object size in the original image, color perturbation considering the predefined similarity between the original and generated images, geometric transformations, and transfer learning. Specifically, to demonstrate color perturbation and geometric transformations, we compare and analyze the experiments of each color space and each geometric transformation. The experimental results show that the proposed method can effectively augment the original data, correctly classify similar items, and improve the image classification accuracy. In addition, it also demonstrates that the effective data augmentation method is crucial when the amount of training data is small.

Ground-based weather radar can observe a wide range with a high spatial and temporal resolution. They are beneficial to meteorological research and services by providing valuable information. Recent weather radar data related research has focused on applying machine learning and deep learning to solve complicated problems. It is a well-known fact that an adequate amount of data is a positively necessary condition in machine learning and deep learning. Generative adversarial networks (GANs) have received extensive attention for their remarkable data generation capacity, with a fascinating competitive structure having been proposed since. Consequently, a massive number of variants have been proposed; which model is adequate to solve the given problem is an inevitable concern. In this paper, we propose exploring the problem of radar image synthesis and evaluating different GANs with authentic radar observation results. The experimental results showed that the improved Wasserstein GAN is more capable of generating similar radar images while achieving higher structural similarity results.

Bayesian networks are probabilistic models used for prediction and decision making under uncertainty. The delivery quantity, the production quantity, and the inventory are changing according to various unexpected events. Then the prediction of a production inventory is required to cope with such irregular fluctuations. This paper considers a production adjustment method for an automobile parts production process by using a dynamic Bayesian network. All factors that may influence the production quantity, the delivery quantity, and the inventory quantity will be handled. This study also provides a production schedule algorithm that sequentially adjusts the production schedule in order to guarantee that all deadlines are met. Furthermore, an adjusting rule for the production quantities is provided in order to maintain guaranteed delivery.

A number of technologically advanced devices, such as radars and satellites, are used in an actual weather forecasting process. Among these devices, the radar is essential equipment in this process because it has a wide observation area and fine resolution in both the time and the space domains. However, the radar can also observe unwanted nonweather phenomena. Anomalous propagation echo is one of the representative nonprecipitation echoes generated by an abnormal refraction phenomenon of a radar beam. Abnormal refraction occurs when the temperature and the humidity change dramatically. In such a case, the radar recognizes either the ground or the sea surface as an atmospheric object. This false observation decreases the accuracy of both quantitative precipitation estimation and weather forecasting. Therefore, a system that can automatically recognize an anomalous propagation echo from the radar data needs to be developed. In this paper, we propose a classification method for separating anomalous propagation echoes from the rest of the weather data by using a combination of a support vector machine classifier and the synthetic minority oversampling technique, to solve the problem of imbalanced data. By using actual cases of anomalous propagation we have confirmed that the proposed method provides good classification results.

Recent computational and behavioral studies suggest that motor adaptation results from the update of multiple memories with different timescales. Here, we designed a model-based functional magnetic resonance imaging (fMRI) experiment in which subjects adapted to two opposing visuomotor rotations. A computational model of motor adaptation with multiple memories was fitted to the behavioral data to generate time-varying regressors of brain activity. We identified regional specificity to timescales: in particular, the activity in the inferior parietal region and in the anterior-medial cerebellum was associated with memories for intermediate and long timescales, respectively. A sparse singular value decomposition analysis of variability in specificities to timescales over the brain identified four components, two fast, one middle, and one slow, each associated with different brain networks. Finally, a multivariate decoding analysis showed that activity patterns in the anterior-medial cerebellum progressively represented the two rotations. Our results support the existence of brain regions associated with multiple timescales in adaptation and a role of the cerebellum in storing multiple internal models.

Electric load forecasting is indispensable for the effective planning and operation of power systems. Various decisions related to power systems depend on the future behavior of loads. In this paper, we propose a new input selection procedure, which combines the group method of data handling (GMDH) and bootstrap method for support vector regression based hourly load forecasting. To construct the GMDH network, a learning dataset is divided into training and test datasets by bootstrapping. After constructing GMDH networks several times, the inputs that appeared frequently in the input layers of the completed networks were selected as the significant inputs. Filter methods based on linear correlation and mutual information (MI) were employed as comparison methods, and the performance of hybrids of the filter methods and the proposed method were also confirmed. In total, five input selection methods were compared. To verify the performance of the proposed method, hourly load data from South Korea was used and the results of one-hour, one-day and one-week-ahead forecasts were investigated. The experimental results demonstrated that the proposed method has higher prediction accuracy compared with the filter methods. Among the five methods, a hybrid of an MI-based filter with the proposed method shows best prediction performance.

Accurate detection and isolation of possible faults are indispensable for operating complex industrial processes more safely, effectively, and economically. In this paper, we propose a fault isolation method for steam boilers in thermal power plants via classification and regression tree (CART)-based variable ranking. In the proposed method, binary classification trees are constructed by applying the CART algorithm to a training dataset which is composed of normal and faulty samples for classifier learning then, to perform faulty variable isolation, variable importance values for each input variable are extracted from the constructed trees. The importance values for non-faulty variables are not influenced by faulty variables, because the values are extracted from the trees with decision boundaries only in the original input space; the proposed method does not suffer from smearing effect. Furthermore, the proposed method, based on the nonparametric CART classifier, can be applicable to nonlinear processes. To confirm the effectiveness, the proposed and comparison methods are applied to two benchmark problems and 250 MW drum-type steam boiler. Experimental results show that the proposed method isolates faulty variables more clearly without the smearing effect than the comparison methods.

Learning a motor task with temporally spaced presentations or with other tasks intermixed between presentations reduces performance during training, but can enhance retention post training. These two effects are known as the spacing and contextual interference effect, respectively. Here, we aimed at testing a unifying hypothesis of the spacing and contextual interference effects in visuomotor adaptation, according to which forgetting between trials due to either spaced presentations or interference by another task will promote between-trial forgetting, which will depress performance during acquisition, but will promote retention. We first performed an experiment with three visuomotor adaptation conditions: a short inter-trial-interval (ITI) condition (SHORT-ITI); a long ITI condition (LONG-ITI); and an alternating condition with two alternated opposite tasks (ALT), with the same single-task ITI as in LONG-ITI. In the SHORT-ITI condition, there was fastest increase in performance during training and largest immediate forgetting in the retention tests. In contrast, in the ALT condition, there was slowest increase in performance during training and little immediate forgetting in the retention tests. Compared to these two conditions, in the LONG-ITI, we found intermediate increase in performance during training and intermediate immediate forgetting. To account for these results, we fitted to the data six possible adaptation models with one or two time scales, and with interference in the fast, or in the slow, or in both time scales. Model comparison confirmed that two time scales and some degree of interferences in either time scale are needed to account for our experimental results. In summary, our results suggest that retention following adaptation is modulated by the degree of between-trial forgetting, which is due to time-based decay in single adaptation task and interferences in multiple adaptation tasks.

Several types of non-precipitation echoes appear in radar images and disrupt the weather forecasting process. An anomalous propagation echo is an unwanted observation result similar to a precipitation echo. It occurs through radar-beam ducting because of the temperature, humidity distribution, and other complicated atmospheric conditions. Anomalous propagation echoes should be removed because they make weather forecasting difficult. In this paper, we suggest an ensemble classification method based on an artificial neural network and a clustering-based subset-selection method. This method allows us to implement an efficient classification method when a feature space has complicated distributions. By separating the input data into atomic and non-atomic clusters, each derived cluster will receive its own base classifier. In the experiments, we compared our method with a standalone artificial neural network classifier. The suggested ensemble classifier showed 84.14% performance, which was about 2% higher than that of the k-means clustering-based ensemble classifier and about 4% higher than the standalone artificial neural network classifier.