Search

The paper proposes a novel block-wise prediction paradigm based on Convolutional Neural Networks (CNNs) for lossless video coding. A deep neural network model which follows a multi-resolution design is employed for block-wise prediction. Several contributions are proposed to improve neural network training. A first contribution proposes a novel loss function formulation for an efficient network training based on a new approach for patch selection. Another contribution consists in replacing all HEVC-based angular intra-prediction modes with a CNN-based intra-prediction method, where each angular prediction mode is complemented by a CNN-based prediction mode using a specifically trained model. Another contribution consists in an efficient adaptation of the CNN-based intra-prediction residual for lossless video coding. Experimental results on standard test sequences show that the proposed coding system outperforms the HEVC standard with an average bitrate improvement of around 5%. To our knowledge, the paper is the first to replace all the traditional HEVC-based angular intra-prediction modes with an intra-prediction method based on modern Machine Learning techniques for lossless video coding applications. [ABSTRACT FROM AUTHOR]

Image retargeting techniques adjust images into different sizes and have attracted much attention recently. Objective quality assessment (OQA) of image retargeting results is often desired to automatically select the best results. Existing OQA methods train a model using some benchmarks (e.g., RetargetMe), in which subjective scores evaluated by users are provided. Observing that it is challenging even for human subjects to give consistent scores for retargeting results of different source images (diff-source-results), in this paper we propose a learning-based OQA method that trains a General Regression Neural Network (GRNN) model based on relative scores—which preserve the ranking—of retargeting results of the same source image (same-source-results). In particular, we develop a novel training scheme with provable convergence that learns a common base scalar for same-source-results. With this source specific offset, our computed scores not only preserve the ranking of subjective scores for same-source-results, but also provide a reference to compare the diff-source-results. We train and evaluate our GRNN model using human preference data collected in RetargetMe. We further introduce a subjective benchmark to evaluate the generalizability of different OQA methods. Experimental results demonstrate that our method outperforms ten representative OQA methods in ranking prediction and has better generalizability to different datasets. [ABSTRACT FROM AUTHOR]

In this paper, a new ensemble forecasting model for short-term load forecasting (STLF) is proposed based on extreme learning machine (ELM). Four important improvements are used to support the ELM for increased forecasting performance. First, a novel wavelet-based ensemble scheme is carried out to generate the individual ELM-based forecasters. Second, a hybrid learning algorithm blending ELM and the Levenberg–Marquardt method is proposed to improve the learning accuracy of neural networks. Third, a feature selection method based on the conditional mutual information is developed to select a compact set of input variables for the forecasting model. Fourth, to realize an accurate ensemble forecast, partial least squares regression is utilized as a combining approach to aggregate the individual forecasts. Numerical testing shows that proposed method can obtain better forecasting results in comparison with other standard and state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In time-varying fading channels, channel coefficients are estimated using pilot symbols that are transmitted every coherence interval. For channels with high Doppler spread, the rapid channel variations over time will require considerable bandwidth for pilot transmission, leading to poor throughput. In this paper, we propose a novel receiver architecture using deep recurrent neural networks (RNNs) that learns the channel variations and thereby reduces the number of pilot symbols required for channel estimation. Specifically, we design and train an RNN to learn the correlation in the time-varying channel and predict the channel coefficients into the future with good accuracy over a wide range of Dopplers and signal-to-noise ratios (SNR). The proposed training methodology enables accurate channel prediction through the use of techniques such as teacher-force training, early-stop, and reduction of learning rate on plateau. Also, the robustness of prediction for different Dopplers and SNRs is achieved by adapting the number of predictions into the future based on the Doppler and SNR. Numerical results show that good bit error performance is achieved by the proposed receiver in time-varying fading channels. We also propose a data decision driven receiver architecture using RNNs that further reduces the pilot overhead while maintaining good bit error performance. [ABSTRACT FROM AUTHOR]

Wind power ramp events (WPREs) have received increasing attention in recent years as they have the potential to impact the reliability of power grid operations. In this paper, a novel WPRE forecasting method is proposed which is able to estimate the probability distributions of three important properties of the WPREs. To do so, a neural network (NN) is first proposed to model the wind power generation (WPG) as a stochastic process so that a number of scenarios of the future WPG can be generated (or predicted). Each possible scenario of the future WPG generated in this manner contains the ramping information, and the distributions of the designated WPRE properties can be stochastically derived based on the possible scenarios. Actual wind power data from a wind power plant in the Bonneville Power Administration (BPA) were selected for testing the proposed ramp forecasting method. Results showed that the proposed method effectively forecasted the probability of ramp events. [ABSTRACT FROM PUBLISHER]

Wind speed forecasting is challenging due to its intermittent nature. The wind speed time series (TS) has nonlinear and nonstationary characteristics and not normally distributed, which make it difficult to be predicted by statistical or computational intelligent methods. Empirical mode decomposition (EMD) and its improved versions are powerful tools to decompose a complex TS into a collection of simpler ones. The improved versions discussed in this paper include ensemble EMD (EEMD), complementary EEMD (CEEMD), and complete EEMD with adaptive noise (CEEMDAN). The EMD and its improved versions are hybridized with two computational intelligence-based predictors: support vector regression (SVR) and artificial neural network (ANN). The EMD-based hybrid forecasting methods are evaluated with 12 wind speed TS. The performances of the hybrid methods are compared and discussed. It shows that EMD and its improved versions enhance the performance of SVR significantly but marginally on ANN, and among the EMD-based hybrid methods, the proposed CEEMDAN-SVR is the best method. Possible future works are also recommended for wind speed forecasting. [ABSTRACT FROM PUBLISHER]

The paper proposes a hybrid model of an artificial neural network (ANN) and Fourier series model based on the least squares method (FLSM) for monthly forecasting of tidal current magnitude and direction. The proposed hybrid model is highly accurate and outperforms either of the ANN or the FLSM applied alone. This study was done using data collected from the Bay of Fundy, NS, Canada, in 2008. [ABSTRACT FROM PUBLISHER]

Due to continual day-to-day increase in electricity demand, and hazardous and critical threats of fossil fuels to the environment, researchers are scrutinizing over substitute energy sources. Solar radiation intensity prediction is essential for conducting various research work in the emerging field of Renewable Energy Sources (RESs). This paper has presented development of monthly averaged solar radiation intensity prediction model by employing Artificial Neural Network (ANN) algorithm. Various meteorological parameters have been considered over period of 2 years to execute forecasting for Chandigarh, India. Different normalisation techniques such as min-max, decimal and z-score have been utilised to normalise database. Structure and parameter learning of ANNs has been carried out. Comparative analysis has been done to select optimal architecture based on different performance evaluation measures such as mean square error (MSE), mean absolute percentage error (MAPE), mean absolute error (MAE), and correlation coefficient (R-value) and training time. The network topology with least forecasting errors, higher R-value has been found to be optimum and further simulated for predicting monthly averaged solar radiation intensity for Chandigarh region. [ABSTRACT FROM AUTHOR]

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

In this paper, a generalized formulation for intelligent energy management of a microgrid is proposed using artificial intelligence techniques jointly with linear-programming-based multiobjective optimization. The proposed multiobjective intelligent energy management aims to minimize the operation cost and the environmental impact of a microgrid, taking into account its preoperational variables as future availability of renewable energies and load demand (LD). An artificial neural network ensemble is developed to predict 24-h-ahead photovoltaic generation and 1-h-ahead wind power generation and LD. The proposed machine learning is characterized by enhanced learning model and generalization capability. The efficiency of the microgrid operation strongly depends on the battery scheduling process, which cannot be achieved through conventional optimization formulation. In this paper, a fuzzy logic expert system is used for battery scheduling. The proposed approach can handle uncertainties regarding to the fuzzy environment of the overall microgrid operation and the uncertainty related to the forecasted parameters. The results show considerable minimization on operation cost and emission level compared to literature microgrid energy management approaches based on opportunity charging and Heuristic Flowchart (HF) battery management. [ABSTRACT FROM PUBLISHER]

Predicting the millimeter wave (mmWave) beams and blockages using sub-6 GHz channels has the potential of enabling mobility and reliability in scalable mmWave systems. Prior work has focused on extracting spatial channel characteristics at the sub-6 GHz band and then use them to reduce the mmWave beam training overhead. This approach still requires beam refinement at mmWave and does not normally account for the different dielectric properties at the different bands. In this paper, we first prove that under certain conditions, there exist mapping functions that can predict the optimal mmWave beam and blockage status directly from the sub-6 GHz channel. These mapping functions, however, are hard to characterize analytically which motivates exploiting deep neural network models to learn them. For that, we prove that a large enough neural network can predict mmWave beams and blockages with success probabilities that can be made arbitrarily close to one. Then, we develop a deep learning model and empirically evaluate its beam/blockage prediction performance using a publicly available dataset. The results show that the proposed solution can predict the mmWave blockages with more than 90% success probability and can predict the optimal mmWave beams to approach the upper bounds while requiring no beam training overhead. [ABSTRACT FROM AUTHOR]

Recurrent neural networks are prime candidates for learning evolutions in multi‐dimensional time series data. The performance of such a network is judged by the loss function, which is aggregated into a scalar value that decreases during training. Observing only this number hides the variation that occurs within the typically large training and testing data sets. Understanding these variations is of highest importance to adjust network hyper‐parameters, such as the number of neurons, number of layers or to adjust the training set to include more representative examples. In this paper, we design a comprehensive and interactive system that allows users to study the output of recurrent neural networks on both the complete training data and testing data. We follow a coarse‐to‐fine strategy, providing overviews of annual, monthly and daily patterns in the time series and directly support a comparison of different hyper‐parameter settings. We applied our method to a recurrent convolutional neural network that was trained and tested on 25 years of climate data to forecast meteorological attributes, such as temperature, pressure and wind velocity. We further visualize the quality of the forecasting models, when applied to various locations on the Earth and we examine the combination of several forecasting models. Recurrent neural networks are prime candidates for learning evolutions in multi‐dimensional time series data. We describe a comprehensive and interactive system to visually analyse and compare time series predictions that were generated by convolutional and recurrent neural networks in the context of weather forecasting. [ABSTRACT FROM AUTHOR]

Very short-term load forecasting predicts the loads 1 h into the future in 5-min steps in a moving window manner based on real-time data collected. Effective forecasting is important in area generation control and resource dispatch. It is however difficult in view of the noisy data collection process and complicated load features. This paper presents a method of wavelet neural networks with data pre-filtering. The key idea is to use a spike filtering technique to detect spikes in load data and correct them. Wavelet decomposition is then used to decompose the filtered loads into multiple components at different frequencies, separate neural networks are applied to capture the features of individual components, and results of neural networks are then combined to form the final forecasts. To perform moving forecasts, 12 dedicated wavelet neural networks are used based on test results. Numerical testing demonstrates the effects of data pre-filtering and the accuracy of wavelet neural networks based on a data set from ISO New England. [ABSTRACT FROM PUBLISHER]

Following the growing share of wind energy in electric power systems, several wind power forecasting techniques have been reported in the literature in recent years. In this paper, a wind power forecasting strategy composed of a feature selection component and a forecasting engine is proposed. The feature selection component applies an irrelevancy filter and a redundancy filter to the set of candidate inputs. The forecasting engine includes a new enhanced particle swarm optimization component and a hybrid neural network. The proposed wind power forecasting strategy is applied to real-life data from wind power producers in Alberta, Canada and Oklahoma, U.S. The presented numerical results demonstrate the efficiency of the proposed strategy, compared to some other existing wind power forecasting methods. [ABSTRACT FROM PUBLISHER]

Traffic hotspots, a severe form of network congestion, can be caused unexpectedly in a network-on-chip (NoC) due to the immanent spatio-temporal unevenness of application traffic. Hotspots reduce the NoC’s effective throughput, where in the worst-case scenario, network traffic flows can be frozen indefinitely. To alleviate this problematic phenomenon several adaptive routing algorithms employ online load-balancing schemes, aiming to reduce the possibility of hotspots arising. Since most are not explicitly hotspot-agnostic, they cannot completely prevent hotspot formation(s) as their reactive capability to hotspots is merely passive. This paper presents a pro-active Hotspot-Preventive Routing Algorithm (HPRA) which uses the advance knowledge gained from network-embedded artificial neural network-based (ANN) hotspot predictors to guide packet routing in mitigating any unforeseen near-future hotspot occurrences. First, these ANN-based predictors are trained offline and during multicore operation they gather online statistical data to predict about-to-be-formed hotspots, promptly informing HPRA to take appropriate hotspot-preventive action(s). Next, in a holistic approach, additional ANN training is performed with data acquired after HPRA interferes, so as to further improve hotspot prediction accuracy; hence, the ANN mechanism does not only predict hotspots, but is also aware of changes that HPRA imposes upon the interconnect infrastructure. Evaluation results, including utilizing real application traffic traces gathered from parallelized workload executions onto a chip multiprocessor architecture, show that HPRA can improve network throughput up to $81$ percent when compared with prior-art. Hardware synthesis results affirm the HPRA mechanism’s moderate overhead requisites. [ABSTRACT FROM AUTHOR]

This paper proposes a novel neural network (NN) training method that employs the hybrid exponential smoothing method and the Levenberg–Marquardt (LM) algorithm, which aims to improve the generalization capabilities of previously used methods for training NNs for short-term traffic flow forecasting. The approach uses exponential smoothing to preprocess traffic flow data by removing the lumpiness from collected traffic flow data, before employing a variant of the LM algorithm to train the NN weights of an NN model. This approach aids NN training, as the preprocessed traffic flow data are more smooth and continuous than the original unprocessed traffic flow data. The proposed method was evaluated by forecasting short-term traffic flow conditions on the Mitchell freeway in Western Australia. With regard to the generalization capabilities for short-term traffic flow forecasting, the NN models developed using the proposed approach outperform those that are developed based on the alternative tested algorithms, which are particularly designed either for short-term traffic flow forecasting or for enhancing generalization capabilities of NNs. [ABSTRACT FROM PUBLISHER]

This letter proposes a transferrable model-agnostic meta-learning (T-MAML) approach for short-term load forecasting for single households. The proposed approach enables multiple households to collaboratively train a generic artificial neural network (ANN) model. The generic ANN model is then further trained at each target household node for the STLF purpose. The proposed T-MAML based STLF approach is featured by: (1) significant reduction of computation and communication costs on the household side; and (2) superior STLF performance, especially when there is limited load data for training in a target household. Experiments based on a real Australian residential dataset are conducted to validate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

The abnormal events, such as the unprecedented COVID-19 pandemic, can significantly change the load behaviors, leading to huge challenges for traditional short-term forecasting methods. This article proposes a robust deep Gaussian processes (DGP)-based probabilistic load forecasting method using a limited number of data. Since the proposed method only requires a limited number of training samples for load forecasting, it allows us to deal with extreme scenarios that cause short-term load behavior changes. In particular, the load forecasting at the beginning of abnormal event is cast as a regression problem with limited training samples and solved by double stochastic variational inference DGP. The mobility data are also utilized to deal with the uncertainties and pattern changes and enhance the flexibility of the forecasting model. The proposed method can quantify the uncertainties of load forecasting outcomes, which would be essential under uncertain inputs. Extensive comparison results with other state-of-the-art point and probabilistic forecasting methods show that our proposed approach can achieve high forecasting accuracies with only a limited number of data while maintaining the excellent performance of capturing the forecasting uncertainties. [ABSTRACT FROM AUTHOR]

This letter investigates a Branching Dueling Q-Network (BDQ) based online operation strategy for a microgrid with distributed battery energy storage systems (BESSs) operating under uncertainties. The developed deep reinforcement learning (DRL) based microgrid online optimization strategy can achieve a linear increase in the number of neural network outputs with the number of distributed BESSs, which overcomes the curse of dimensionality caused by the charge and discharge decisions of multiple BESSs. Numerical simulations validate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

A motion cueing algorithm (MCA) is employed to transform the linear and angular motion signals generated from a motion simulator without violating the physical and dynamical boundaries of the motion platform. In this respect, the accurate prediction of the motion scenarios is essential to enhance the efficiency of the MCA using prepositioning or time-varying reference model predictive control. While a recent approach that utilizes a feedforward neural network (NN) to forecast the motion scenarios is useful, the feedforward NN model has only forward dynamics relating to the signals without any feedback loop. In this article, a time-delay feedforward NN, a recurrent NN, and a nonlinear autoregressive (NAR) models with three different training procedures, i.e., Levenberg–Marquardt, Bayesian regularization, and scaled conjugate gradient, are exploited to predict the motion scenarios. As the NAR model employs the historical signals as the inputs, it can predict the motion scenarios with higher accuracy rates. Based on the series of empirical evaluations, NAR trained with Levenberg–Marquardt is able to outperform the other two counterparts in producing more accurate predictions of the motion signals. The NAR method has a lower computational load as compared with that of the recurrent NN, facilitating its real-time application. In addition to the MCA, the NAR method can be employed in other areas, including autonomous vehicles and motion sickness studies. It can also be easily implemented for air, sea, and/or land vehicle simulators for training purposes in virtual reality environments. [ABSTRACT FROM AUTHOR]

The method of estimation and forecasting in intelligent decision support systems is developed. The essence of the proposed method is the ability to analyze the current state of the object under analysis and the possibility of shortterm forecasting of the object state. The possibility of objective and complete analysis is achieved through the use of improved fuzzy temporal models of the object state, an improved procedure for forecasting the object state and an improved procedure for training evolving artificial neural networks. The concepts of a fuzzy cognitive model, in contrast to the known fuzzy cognitive models, are connected by subsets of fuzzy influence degrees, arranged in chronological order, taking into account the time lags of the corresponding components of the multidimensional time series. This method is based on fuzzy temporal models and evolving artificial neural networks. The peculiarity of this method is the ability to take into account the type of a priori uncertainty about the state of the analyzed object (full awareness of the object state, partial awareness of the object state and complete uncertainty about the object state). The ability to clarify information about the state of the monitored object is achieved through the use of an advanced training procedure. It consists in training the synaptic weights of the artificial neural network, the type and parameters of the membership function, as well as the architecture of individual elements and the architecture of the artificial neural network as a whole. The object state forecasting procedure allows conducting multidimensional analysis, consideration and indirect influence of all components of a multidimensional time series with different time shifts relative to each other under uncertainty. [ABSTRACT FROM AUTHOR]

The predispatch price forecast plays a key element in the electricity market. However, such a forecast usually depends on the traditional offline batch-learning technologies, which cannot respond in time to the unexpected changes in the local power system environment. Further, the predispatch local price forecast is often affected by the dynamic price changes from the neighboring regions. This article proposes a novel online learning forecast approach to overcome the above issues to provide a better predispatch price forecast by using the online sequential extreme learning machine (OS-ELM) algorithm. The article proposes a novel data structure in the form of a 2-D orthogonal list and two corresponding OS-ELM modules. One module provides the rolling day-ahead price prediction and prediction intervals using the day-by-day online training update, while the other provides the rolling 30-min prediction using the 2-h-by-2-h online training update. The proposed approach can continuously perceive any unexpected events and any price fluctuations from the neighboring regions in the nonlinear patterns. The proposed approach is validated using simulation studies based on the data from the Australian electricity market, and the simulation results show that the proposed approach can help in improving the forecast accuracy, especially when unexpected changes occur both locally and in the neighboring area. [ABSTRACT FROM AUTHOR]

One of the most groundbreaking structures in deep learning is ResNet, which utilizes skip connections to make neural networks (NNs) significantly deeper. While there are many successful applications of skip connections in computer vision (CV) and natural language processing (NLP), the applications of skip connections in electric load forecasting are still quite limited. Moreover, as compared to the deep NNs used in CV and NLP, most NNs used in load forecasting are relatively shallow NNs, whose performance is partially restricted by the relatively shallow structures. To improve probabilistic forecasting by deepening NN, we investigate the applicability of the skip connections in probabilistic load forecasting NN by proposing new structures and studying the philosophy behind the forecasting improvements brought by skip connections. Case studies show that by deepening the NNs, the proposed structures can generate more accurate probabilistic load forecasts than state-of-the-art methods, which implies that the ideas of skip connections have significant value in probabilistic load forecasting. It is also validated in observations and mathematical proofs that the proposed structures improve the probabilistic forecasting by alleviating the gradient vanishing and exploding problems in deep NNs. [ABSTRACT FROM AUTHOR]

The stochastic nature of energy consumption patterns in a port varies daily and weekly. The port must manage their load to save electricity costs, plan future infrastructure development and consolidate the power utility's optimal power generation efforts. This requires accurate Short-Term Load Forecast (STLF) for quality, quantity and efficiency. Recently Artificial Neural Networks (ANNs) have been proposed for short-term load forecasting (STLF). Here we aim to create a more exact method of short-term load forecasting using the nonlinear, autoregressive, multi-variable exogenous input (NARX) ANNs. We propose a new network architecture: an open-loop training neural network with real load and weather information, followed by a closed-loop network to produce a prediction with the predicted load as its feedback data. The proposed method utilizes its own performance to improve precision, essentially implementing a load feedback loop that is less dependent on external data. In our proposed system, mean absolute percentage errors were reached in the forecast range of 1%: 30% better on the average error by feedforward ANNs. Our test used actual reticulation network load data from Transnet Port Terminal in East London and climate statistics to predict port terminal load for a week in advance. [ABSTRACT FROM AUTHOR]

This letter presents a novel technique for the forecast of the ground horizontal irradiance (GHI) from satellite-based images. To enhance the forecast accuracy, spatial information in addition to temporal information has been considered. This produced an increase in the computational load of the forecast process. Dimensionality reduction techniques based on nonlinear principal component analysis (PCA) are used to project the original data set into low-dimension feature space. A multilayer feedforward neural network classifier is used to model the signal through a training operation involving past history of the considered spatiotemporal signal. Experiments have been carried out on two different data sets. Comparisons with classical forecasting techniques demonstrate that the introduction of the spatial information permits to obtain better short-term forecast measurements for all types of sky conditions. Moreover, further analysis demonstrates that, compared with linear PCA, the nonlinear PCA is more appropriate for dimensionality reduction of spatiotemporal GHI data set. [ABSTRACT FROM PUBLISHER]

Rapidly evolving businesses generate massive amounts of time-stamped data sequences and cause a demand for both univariate and multivariate time series forecasting. For such data, traditional predictive models based on autoregression are often not sufficient to capture complex nonlinear relationships between multidimensional features and the time series outputs. In order to exploit these relationships for improved time series forecasting while also better dealing with a wider variety of prediction scenarios, a forecasting system requires a flexible and generic architecture to accommodate and tune various individual predictors as well as combination methods. In reply to this challenge, an architecture for combined, multilevel time series prediction is proposed, which is suitable for many different universal regressors and combination methods. The key strength of this architecture is its ability to build a diversified ensemble of individual predictors that form an input to a multilevel selection and fusion process before the final optimized output is obtained. Excellent generalization ability is achieved due to the highly boosted complementarity of individual models further enforced through cross-validation-linked training on exclusive data subsets and ensemble output postprocessing. In a sample configuration with basic neural network predictors and a mean combiner, the proposed system has been evaluated in different scenarios and showed a clear prediction performance gain. [ABSTRACT FROM AUTHOR]