Search

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]

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]

Inter prediction is a crucial part of hybrid video coding frameworks, utilized to exploit the temporal redundancy in video sequences and improve the coding performance. During inter prediction, a predicted block is typically derived from reference pictures using motion estimation and motion compensation. To improve the coding performance of inter prediction, a neural network based enhancement to inter prediction (NNIP) is proposed in this paper. NNIP is composed of three networks, namely residue estimation network, combination network, and deep refinement network. Specifically, first, a residue estimation network is designed to estimate the residue between current block and its predicted block using their available spatial neighbors. Second, the feature maps of the estimated residue and the predicted block are extracted and concatenated in a combination network. Finally, the concatenated feature maps are fed into a deep refinement network to generate a refined residue, which is added back to the predicted block to derive a more accurate predicted block. NNIP is integrated in HEVC to evaluate its efficiency. The experimental results demonstrate that NNIP can achieve 4.6%, 3.0%, and 2.7% BD-rate reduction on average under LDP, LDB, and RA configurations compared to HEVC. [ABSTRACT FROM AUTHOR]

Load forecasting is one of the most important and studied topics in modern power systems. Most of the existing research on day-ahead load forecasting try to build a good model to improve the forecasting accuracy. The forecasted load is then used as the input to generation scheduling with the ultimate goal of minimizing the cost of generation schedules. However, existing day-ahead load forecasting models do not consider this ultimate goal at the training/forecasting stage. This paper proposes a task-based day-ahead load forecasting model labeled as LfEdNet that combines two individual layers in one model, including a load forecasting layer based on deep neural network (Lf layer) and a day-ahead stochastic economic dispatch (SED) layer (Ed layer). The training of LfEdNet aims to minimize the cost of the day-ahead SED in the Ed layer by updating the parameters of the Lf layer. Sequential quadratic programming (SQP) is used to solve the day-ahead SED in the Ed layer. The test results demonstrate that the forecasted results produced by LfEdNet can lead to lower cost of day-ahead SED at the expense of slight reduction in forecasting accuracy. [ABSTRACT FROM AUTHOR]

For the emerging mobility-on-demand services, it is of great significance to predict passenger demands based on historical mobility trips towards better vehicle distribution. Prior works have focused on predicting next-step passenger demands at selected locations or hotspots. However, we argue that multi-step citywide passenger demands encapsulate both time-varying demand trends and global statuses, and hence are more beneficial to avoiding demand-service mismatching and developing effective vehicle distribution/scheduling strategies. Furthermore, we find that adaptations of single-step methods are unable to achieve robust prediction with high accuracy for further steps. In this paper, we propose an end-to-end deep neural network model to the prediction task. We employ an encoder-decoder framework based on convolutional and ConvLSTM units to identify complex features that capture spatiotemporal influence and pickup-dropoff interactions on citywide passenger demands. We introduce a multi-level attention model (global attention and temporal attention) to emphasize the effects of latent citywide mobility regularities and capture relevant temporal dependencies. We evaluate our proposed method using real-world mobility trips (taxis and bikes) and the experimental results show that our method achieves higher prediction accuracy than the state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Rear-end collision prediction has gained an increasing attention for safety improvement in smart cities. It is urgent to design efficient warning strategies for rear-end collisions which is one of the main causes of traffic accidents. The existing researches have been conducted to predict the collisions. Learning-based methods are proposed to solve this complicated issue, which the traditional methods are difficult to solve. However, due to some limitations in terms of the feature extraction and prediction performance, back-propagation learning methods are facing challenge. In this paper, we proposed a novel Rear-end Collision Prediction Mechanism with deep learning method (RCPM), in which a convolutional neural network model is established. In RCPM, the dataset is smoothed and expanded based on genetic theory to alleviate the class imbalance problem. The preprocessed dataset is divided into training and testing sets as the input to train our convolutional neural network model. The experimental results show that compared with the Honda, Berkeley and multi-layer perception neural-network-based algorithms, RCPM effectively improves performance to predict rear-end collisions. [ABSTRACT FROM AUTHOR]

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]

In this paper, a two-stage Bayesian learning-based approach is proposed to enable online dynamic security assessment (DSA) and preventive control in power systems. To develop a reliable data mining-based model for DSA, the Bayesian neural network is learned by Bayes by Backprop to detect the security status in real-time and to inform the system operators with the confidence of the prediction. For dynamic insecurity prevention, the security-constrained optimal power flow (SCOPF) model incorporating the Bayesian neural network based security constraints is proposed, which is solved by the derivation-free Bayesian optimization. Case studies on the IEEE 39-bus system show that the Bayesian neural network is able to detect the security status of the power system correctly and reliably and is able to avoid the over-fitting problem compared with the traditional point estimation neural network. Also, Bayesian optimization is able to solve the non-analytical SCOPF model efficiently to mitigate the dynamic insecurity by preventive generation rescheduling. [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]

Unsteady motion of the atmosphere incurs nonlinear and spatiotemporally coupled uncertainties in the wind power prediction (WPP) of multiple wind farms. This brings both opportunities and challenges to wind power probabilistic forecasting (WPPF) of a wind farm cluster or region, particularly when wind power is highly penetrated within the power system. This paper proposes an Improved Deep Mixture Density Network (IDMDN) for short-term WPPF of multiple wind farms and the entire region. In this respect, a deep multi-to-multi (m2m) mapping Neural Network model, which adopts the beta kernel as the mixture component to avoid the density leakage problem, is established to produce probabilistic forecasts in an end-to-end manner. A novel modified activation function and several general training procedures are then introduced to overcome the unstable behavior and NaN (Not a Number) loss issues of the beta kernel function. Verification of IDMDN is based on an open-source dataset collected from seven wind farms, and comparison results show that the proposed model improves the WPPF performance at both wind farm and regional levels. Furthermore, a laconic and accurate probabilistic expression of predicted power at each time step is produced by the proposed model. [ABSTRACT FROM AUTHOR]

This paper develops a hierarchical deep learning machine (HDLM) to efficiently achieve both quantitative and qualitative online transient stability prediction (TSP). For the sake of improving its online efficiency, multiple generators’ fault-on trajectories as well as the two closest data-points in pre-/post-fault stages are acquired by PMUs to form its raw inputs. An anti-noise graphical transient characterization technique is tactfully designed to transform multiplex trajectories into 2-D images, within which system-wide transients are concisely described. Then, following the divide-and-conquer philosophy, the HDLM trains a two-level convolutional neural network (CNN) based regression model. With stability margin regressions hierarchically refined, it manages to perform reliable and adaptive online TSP almost immediately after fault clearance. Test results on the IEEE 39-bus test system and the real-world Guangdong Power Grid in South China demonstrate the HDLM's superior performances on both stability status and stability margin predictions. [ABSTRACT FROM AUTHOR]

Domestic electric water heaters (DEWHs) can provide operational flexibility for load control due to their energy storage capacity. Load forecasting for aggregated DEWHs is important for providing information of baseline load and controlling electricity demand profile without negative impact to the normal end use. Advanced metering infrastructures nowadays provide more possibilities to further enhance forecasting with bottom-up method. This article proposes a bottom-up forecasting with Markov-based error reduction method to predict power consumption of aggregated DEWHs for multiple forecast horizons. DEWHs are randomly divided into small aggregations, whose power consumption is forecasted by independent forecast engines. In this paper, the engines are K-means and wavelet decomposition-based neural networks. After summing all forecasting of small aggregations up, a new Markov-based error reduction method is proposed to extract features in residuals and mitigate forecasting error accumulation introduced by the summation, providing opportunities to further improve forecasting accuracy for the total DEWH load. Differing from traditional Markov-based error reduction, two new compensation parameters (compensation coefficient, and compensation threshold) are proposed. They are determined by using particle swarm optimization algorithm. Experiments on real and simulated DEWH loads verified the effectiveness of the proposed forecasting method. The proposed method improved the forecast accuracy over selected benchmark algorithms by about 20% to 80%, according to four performance metrics: mean absolute error, mean absolute percentage error, root-mean-square error, normalized form RMSE. The aggregation effects on performance were also analyzed in theory and tested with simulated DEWHs, providing a good indication of the forecast dependence on the aggregation size. [ABSTRACT FROM AUTHOR]

Anomalous data can negatively impact energy forecasting by causing model parameters to be incorrectly estimated. This paper presents two approaches for the detection and imputation of anomalies in time series data. Autoregressive with exogenous inputs (ARX) and artificial neural network (ANN) models are used to extract the characteristics of time series. Anomalies are detected by performing hypothesis testing on the extrema of the residuals, and the anomalous data points are imputed using the ARX and ANN models. Because the anomalies affect the model coefficients, the data cleaning process is performed iteratively. The models are re-learned on “cleaner” data after an anomaly is imputed. The anomalous data are reimputed to each iteration using the updated ARX and ANN models. The ARX and ANN data cleaning models are evaluated on natural gas time series data. This paper demonstrates that the proposed approaches are able to identify and impute anomalous data points. Forecasting models learned on the unclean data and the cleaned data are tested on an uncleaned out-of-sample dataset. The forecasting model learned on the cleaned data outperforms the model learned on the unclean data with 1.67% improvement in the mean absolute percentage errors and a 32.8% improvement in the root mean squared error. Existing challenges include correctly identifying specific types of anomalies such as negative flows. [ABSTRACT FROM PUBLISHER]

Talent flow analysis is a process for analyzing and modeling the flows of employees into and out of targeted organizations, regions, or industries. A clear understanding of talent flows is critical for many applications, such as human resource planning, brain drain monitoring, and future workforce forecasting. However, existing studies on talent flow analysis are either qualitative or limited by coarse level quantitative modeling. To this end, in this paper, we provide a fine-grained data-driven approach to model the dynamics and evolving nature of talent flows by leveraging the rich information available in job transition networks. Specifically, we first investigate how to enrich the sparse talent flow data by exploiting the correlations between the stock price movement and the talent flows of public companies. Then, we formalize the talent flow modeling problem as to predict the increments of the edge weights in the dynamic job transition network. In this way, the problem is transformed into a multi-step time series forecasting problem. A deep sequence prediction model is developed based on the recurrent neural network model, which consumes multiple input sources derived from dynamic job transition networks. Finally, experimental results on real-world data show that the proposed model outperforms other benchmark models in terms of prediction accuracy. The results also indicate that the proposed model can provide reasonable performance even if the historical talent flow data are not completely available. [ABSTRACT FROM AUTHOR]

It is important to predict the hidden behavior of a complex system. In the existing models for predicting the hidden behavior, the hidden belief rule base (HBRB) is an effective model which can use qualitative knowledge and quantitative data. However, the frame of discernment (FoD) of HBRB which is composed of some states or propositions and the universal set including all states or propositions is not complete. The global ignorance and local ignorance cannot be considered at the same time, which may lead to the inaccurate forecasting results. To solve the problems, a new HBRB model named as PHBRB in which the hidden behavior is described on the FoD of the power set is proposed in this correspondence paper. Furthermore, by using the evidential reasoning rule as the inference tool of PHBRB, a new projection covariance matrix adaption evolution strategy is developed to optimize the parameters of PHBRB so that more accurate prediction results can be obtained. A case study of network security situation prediction is conducted to demonstrate the effectiveness of the newly proposed method. [ABSTRACT FROM AUTHOR]

The ever-increasing number of wind farms has brought both challenges and opportunities in the development of wind power forecasting techniques to take advantage of interdependencies between tens or hundreds of spatially distributed wind farms, e.g., over a region. In this paper, a sparsity-controlled vector autoregressive (SC-VAR) model is introduced to obtain sparse model structures in a spatio-temporal wind power forecasting framework by reformulating the original VAR model into a constrained mixed integer nonlinear programming (MINLP) problem. It allows controlling the sparsity of the coefficient matrices in direct manner. However this original SC-VAR is difficult to implement due to its complicated constraints and the lack of guidelines for setting its parameters. To reduce the complexity of this MINLP and to make it possible to incorporate prior expert knowledge to benefit model building and forecasting, the original SC-VAR is modified and a correlation-constrained SC-VAR (CCSC-VAR) is proposed based on spatial correlation information about wind farms. Our approach is evaluated based on a case study of very-short-term forecasting for 25 wind farms in Denmark. Comparison is performed with a set of traditional local methods and spatio-temporal methods. The results obtained show the proposed CCSC-VAR has better overall performance than both the original SC-VAR and other benchmark methods, taking into account all evaluation indicators, including sparsity-control ability, sparsity, accuracy, and efficiency. [ABSTRACT FROM AUTHOR]

Accurate prediction of cycle time (CT) plays an important role in the promises of a good delivery-time for semiconductor manufacturers. However, the CTs of wafer lots are difficult to forecast since they are influent by a lot factors (e.g., workload for every machine). The identification of key factors (called CT-related) can not only improve the prediction performance but also facilitate the control of CT, which is of particular significance. This paper focuses on gathering subtle variables into the candidate variables set along with the further analysis which is required in the correlation analysis between various candidate factors and CTs of wafer lots. The candidates set is first constructed to collect all factors that may affect wafer lots’ CT. Then, a regression-based model is proposed to select CT-related variables, which consists of three parts: 1) discretization; 2) adaptive logistic regression-based correlation analysis; and 3) CT-related factor selection. Subsequently, a parallel computation model is implemented to forecast the CTs of wafer lots. In the numerical experiments, 108 CT-related factors stood out from 774 candidates and replaced six global factors (used as reference) to predict CTs of wafer lots. The results indicated that the proposed approach had higher accuracy than linear regression and back propagation network in CT forecasting in large scale data analysis. [ABSTRACT FROM PUBLISHER]

With rapid increase in wind power penetration into the power grid, wind power forecasting is becoming increasingly important to power system operators and electricity market participants. The majority of the wind forecasting tools available in the literature provide deterministic prediction, but given the variability and uncertainty of wind, such predictions limit the use of the existing tools for decision-making under uncertain conditions. As a result, probabilistic forecasting, which provides information on uncertainty associated with wind power forecasting, is gaining increased attention. This paper presents a novel hybrid intelligent algorithm for deterministic wind power forecasting that utilizes a combination of wavelet transform (WT) and fuzzy ARTMAP (FA) network, which is optimized by using firefly (FF) optimization algorithm. In addition, support vector machine (SVM) classifier is used to minimize the wind power forecast error obtained from WT+FA+FF. The paper also presents a probabilistic wind power forecasting algorithm using quantile regression method. It uses the wind power forecast results obtained from the proposed hybrid deterministic WT+FA+FF+SVM model to evaluate the probabilistic forecasting performance. The performance of the proposed forecasting model is assessed utilizing wind power data from the Cedar Creek wind farm in Colorado. [ABSTRACT FROM PUBLISHER]

A wet road is slippery so vehicles often slow down their speed to increase the safety margin, thus usually reducing the average speed. This reduction in average speed may produce a chain reaction that shifts, extends, or amplifies a slowdown on downstream road segments. Conventional average-speed forecasting approaches are unable to respond to sudden chain reactions because these approaches do not consider the effect of weather factors and upstream road segments. Since accurate forecast of average speed can improve gas consumption, carbon dioxide emissions, and travel time, this paper proposes a short-term average-speed forecast and adjustment (ASFA) approach based on a study of prediction bias correlation among adjacent road segments and on weather factors, such as temperature, humidity, and rainfall. First, this approach applies an artificial neural network to predict the average speed of a road segment. Then, vehicles can monitor current traffic to calculate average speed via vehicular ad hoc networks (VANETs). Finally, the vehicles adjust the predicted average speed according to the observed average speed. Real traffic measurements and weather data are used for the evaluation of the proposed scheme and Civic Boulevard in Taipei City is selected as the prediction target. The results show that the proposed ASFA improves accuracy by 57.4% when compared with a hybrid approach on an urban street during rush hour. This paper estimates and simulates a case study by aggregating traffic data in 186 ms on Shi-Min Boulevard via the VANETs during rush hour. [ABSTRACT FROM PUBLISHER]

In this paper, a novel hybrid architecture is proposed for developing a prognosis and health monitoring methodology for nonlinear systems through integration of model-based and computationally intelligent-based techniques. In our proposed framework, the well-known particle filters (PFs) method is utilized to estimate the states as well as the health parameters of the system. Simultaneously, the system observations are predicted through an observation forecasting scheme that is developed based on neural networks (NNs) paradigms. The objective is to construct observation profiles that are to be used in future time horizons. Our proposed online training that is utilized for observation forecasting enables the NNs models to track nonergodic changes in the profiles that are present due to presence of hidden damage affecting the system health parameters. The forecasted observations are then utilized in the PFs to predict the evolution of the system states as well as the health parameters (which are considered to be time-varying due to effects of degradation and damage) into future time horizons. Our proposed hybrid architecture enables one to select health signatures for determining the remaining useful life of the system or its components not only based on the system observations but also by taking into account the system health parameters that are not physically measurable. Our proposed hybrid health monitoring methodology is constructed and developed by invoking a special framework where implementation of the observation forecasting scheme is not dependent on the structure of the utilized NNs model. In other words, changing the network structure will not significantly affect the prediction accuracy associated with the entire health prediction scheme. To verify and validate the above results and as a case study, our proposed hybrid approach is applied to predict the health condition of a gas turbine engine when it is affected by and subjected to fouling and erosion degradation and fault damages. [ABSTRACT FROM PUBLISHER]

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]

Short-term load forecasts are needed for the efficient management of power systems. Although weather-based modeling is common, univariate models can be useful when the lead time of interest is less than one day. A class of univariate methods that has performed well with intraday data is exponential smoothing. This paper considers five recently developed exponentially weighted methods that have not previously been used for load forecasting. These methods include several exponential smoothing formulations, as well as methods using discount weighted regression, cubic splines, and singular value decomposition (SVD). In addition, this paper presents a new SVD-based exponential smoothing formulation. Using British and French half-hourly load data, these methods are compared for point forecasting up to one day ahead. Although the new SVD-based approach showed some potential, the best performing method was a previously developed exponential smoothing method. A second empirical study showed the better of the univariate methods outperforming a weather-based method up to about five hours ahead, with a combination of these methods producing the best results overall. [ABSTRACT FROM PUBLISHER]

Short-term load forecasting at the distribution level predicts the load of substations, feeders, transformers, and possibly customers from half an hour to one week ahead. Effective forecasting is important for the planning and operation of distribution systems. The problem, however, is difficult in view of complicated load features, the large number of distribution-level nodes, and possible switching operations. In this paper, a new forecasting approach within the hierarchical structure is presented to solve these difficulties. Load of the root node at any user-defined subtree is first forecast by a wavelet neural network with appropriate inputs. Child nodes categorized as “regular” and “irregular” based on load pattern similarities are then forecast separately. Load of a regular child node is simply forecast as the proportion from the parent node load forecast while the load of an irregular child node is forecast by an individual neural network model. Switching operation detection and follow-up adjustments are also performed to capture abnormal changes and improve the forecasting accuracy. This new approach captures load characteristics of nodes at different levels, takes advantage of pattern similarities between a parent node and its child nodes, detects abnormalities, and provides high quality forecasts as demonstrated by two practical datasets. [ABSTRACT FROM PUBLISHER]

This paper proposes a novel stratification-based wind power forecasting method and develops a hybrid forecasting model at different stratifications using charged system search algorithm. The proposed model applies the concept of segmentation from the theory of optimal stratification to forecast short-term wind power outputs. Additionally, the proposed method elucidates different weighting values of each individual model at different segmentation blocks. Based on the forecasting results, the proposed stratification-based hybrid model outperforms traditional stand-alone models and unstratified hybrid models in terms of forecasting accuracy, which verifies the proposed forecasting model for accurate wind power forecasting. [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]

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]

In a deregulated power market, the real-time wholesale market price of electricity varies dramatically within a single day due to the availability of the resources. Moreover, the price of electricity can be different from one location to the other at the same time period due to the location of the available resources and transmission constraints. This is the so-called locational marginal price (LMP). Since wind power is noncontrollable and partially unpredictable, it is difficult to schedule its output to exploit LMP variations. While energy storage system (ESS) may accommodate wind farm output, it requires significant initial financial commitment. Accurately forecasted wind power and LMP information can reduce the required capacity and make it financially feasible for the ESS to perform desired functions. In this paper, artificial neural network (ANN) technique is employed to forecast the day-ahead wind power and LMP, and a hybrid ESS consisting of two storage facilities is developed. The primary ESS is utilized for the optimizing wind-storage system production schedule with day-ahead forecasting data, while the secondary ESS is applied to address the forecasting errors during real-time operation. With this hybrid ESS design, financial benefits are achieved for the wind farm. [ABSTRACT FROM AUTHOR]

On-road sensor systems installed on freeways are used to capture traffic flow data for short-term traffic flow predictors for traffic management, to reduce traffic congestion and improve vehicular mobility. This paper intends to tackle the impractical time-invariant assumptions which underlie the methods currently used to develop short-term traffic flow predictors: 1) the characteristics of current data captured by on-road sensors are assumed to be time invariant with respect to those of the historical data, which is used to developed short-term traffic flow predictors; and 2) the configuration of the on-road sensor systems is assumed to be time-invariant. In fact, both assumptions are impractical in the real world, as the current traffic flow characteristics can be very different from the historical ones, and also the on-road sensor systems are time varying in nature due to damaged sensors or component wear. Therefore, misleading forecasting results are likely to be produced when short-term traffic flow predictors are designed using these two time-invariant assumptions. To tackle these time-invariant assumptions, an intelligent particle swarm optimization (IPSO) algorithm is proposed to develop short-term traffic flow predictors by integrating the mechanisms of PSO, neural network and fuzzy inference system, to adapt to the time-varying traffic flow characteristics and the time-varying configurations of the on-road sensor systems. The proposed IPSO was applied to forecast traffic flow conditions on a section of freeway in Western Australia, whose traffic flow information can be captured on-line by the on-road sensor system. These results clearly demonstrate the effectiveness of using the proposed IPSO for real-time traffic flow forecasting based on traffic flow data captured by on-road sensor systems. [ABSTRACT FROM PUBLISHER]

Numerous methods have been proposed for forecasting load for normal days. Modeling of anomalous load, however, has often been ignored in the research literature. Occurring on special days, such as public holidays, anomalous load conditions pose considerable modeling challenges due to their infrequent occurrence and significant deviation from normal load. To overcome these limitations, we adopt a rule-based approach, which allows incorporation of prior expert knowledge of load profiles into the statistical model. We use triple seasonal Holt-Winters-Taylor (HWT) exponential smoothing, triple seasonal autoregressive moving average (ARMA), artificial neural networks (ANNs), and triple seasonal intraweek singular value decomposition (SVD) based exponential smoothing. These methods have been shown to be competitive for modeling load for normal days. The methodological contribution of this paper is to demonstrate how these methods can be adapted to model load for special days, when used in conjunction with a rule-based approach. The proposed rule-based method is able to model normal and anomalous load in a unified framework. Using nine years of half-hourly load for Great Britain, we evaluate point forecasts, for lead times from one half-hour up to a day ahead. A combination of two rule-based methods generated the most accurate forecasts. [ABSTRACT FROM PUBLISHER]

On-road sensors provide proactive traffic control centers with current traffic flow conditions in order to forecast the future conditions. However, the number of on-road sensors is usually huge, and not all traffic flow conditions captured by these sensors are useful for predicting future traffic flow conditions. The inclusion of all captured traffic flow conditions is an ineffective means of predicting future traffic flow. Therefore, the selection of appropriate on-road sensors, which are significantly correlated to future traffic flow, is essential, although the trial and error method is generally used for the selection. In this paper, the Taguchi method, which is a robust and systematic optimization approach for designing reliable and high-quality models, is proposed for determinations of appropriate on-road sensors, in order to capture useful traffic flow conditions for forecasting. The effectiveness of the Taguchi method is demonstrated by developing a traffic flow predictor based on the architecture of fuzzy neural networks which can perform well on traffic flow forecasting. The case study was conducted based on traffic flow data captured by on-road sensors located on a Western Australia freeway. The advantages of using the Taguchi method can be indicated: 1) Traffic flow predictors with high accuracy can be designed, and 2) development time of traffic flow predictors is reasonable. [ABSTRACT FROM PUBLISHER]

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]

Artificial neural networks (ANNs) have been widely applied in electricity price forecasts due to their nonlinear modeling capabilities. However, it is well known that in general, traditional training methods for ANNs such as back-propagation (BP) approach are normally slow and it could be trapped into local optima. In this paper, a fast electricity market price forecast method is proposed based on a recently emerged learning method for single hidden layer feed-forward neural networks, the extreme learning machine (ELM), to overcome these drawbacks. The new approach also has improved price intervals forecast accuracy by incorporating bootstrapping method for uncertainty estimations. Case studies based on chaos time series and Australian National Electricity Market price series show that the proposed method can effectively capture the nonlinearity from the highly volatile price data series with much less computation time compared with other methods. The results show the great potential of this proposed approach for online accurate price forecasting for the spot market prices analysis. [ABSTRACT FROM PUBLISHER]

Multinodal load forecasting deals with the loads of several interest nodes in an electrical network system, which is also known as bus load forecasting. To perform this demand, a technique that is precise, reliable, and has short-time processing is necessary. This paper uses two methodologies for short-term multinodal load forecasting. The first individually forecasts the local loads and the second forecasts the global load and individually forecasts the load participation factors to estimate the local loads. For the forecasts, a modified general regression neural network and a procedure to automatically reduce the number of inputs of the artificial neural networks are proposed. To design the forecasters, the previous study of the local loads was not necessary, thus reducing the complexity of the multinodal load forecasting. Tests were carried out by using a New Zealand distribution subsystem and the results obtained were found to be compatible with those available in the specialized literature. [ABSTRACT FROM AUTHOR]

In a deregulated environment, system operators are required to procure certain ancillary services, which, among others, may include compensation for active-power losses. This compensation usually involves long-term energy purchases and additional short-term energy purchases to cover the daily fluctuations. The short-term energy purchases require an accurate and quick short-term forecasting method that has to be efficiently applicable in day-ahead markets. This paper presents a novel short-term active-power-loss forecast method using power-flow analysis for the forecasted day. Specifically, this includes short-term load and generation forecasts as well as network-topology forecasts, which are used for the power-flow calculations and the resulting active-power loss calculations. To minimize the forecast errors, a fuzzy-weight grouping of the different short-term load and generation forecast results is proposed. An additional step for input-data pre-processing is presented, where the fuzzy clustering considers the patterns for training the forecasting models. The proposed approach was verified by using real data for the ENTSO-E interconnection and tested for the Slovenian power system. The forecasting results demonstrate the improved accuracy of the proposed approach. [ABSTRACT FROM PUBLISHER]

A novel hybrid approach, combining wavelet transform, particle swarm optimization, and adaptive-network-based fuzzy inference system, is proposed in this paper for short-term electricity prices forecasting in a competitive market. Results from a case study based on the electricity market of mainland Spain are presented. A thorough comparison is carried out, taking into account the results of previous publications. Finally, conclusions are duly drawn. [ABSTRACT FROM AUTHOR]

The ultimate goal of an electric utility is to create maximum profit while maintaining reliability and security of the power supply. The operation and control of power system is sensitive to system demand. Therefore, improvements in load-forecasting accuracy will lead to cost savings and enhance system security. Due to Taiwan's distinct climate characteristics, it is difficult to obtain satisfactory load-forecasting results by treating the whole island as a single region. In addition, weather factors, such as temperature, relative humidity, and the Heat Index (HI) (a human-perceived equivalent temperature) may also affect load-consumption patterns. This paper proposes a multiregion short-term load-forecasting methodology, taking into account the HI to improve load-forecasting accuracy in Taiwan Power Company's (Taipower's) system. The results show that adopting the HI as a parameter can effectively improve the accuracy if the temperature of the region under investigation is above 27 ^\circ\C (80 ^ \circ\F). By considering both the load/weather diversity and the HI, further improvements to the load forecasting for the Taipower system during summer can be achieved. [ABSTRACT FROM AUTHOR]

During the last decade, neural networks have emerged as one of the most powerful and accurate nonlinear models for load forecasting. However, using neural networks requires users to have in-depth knowledge to determine the model structure and parameters, which limits their wide application. To overcome this weakness, this paper proposes an integrated approach which combines a self-organizing fuzzy neural network (SOFNN) learning method with a bilevel optimization method. SOFNNs can automatically determine both the model structure and parameters, while the bilevel optimization method automatically selects the best pre-training parameters to ensure that the best fuzzy neural networks be identified. Therefore, the proposed approach is able to automatically identify the best fuzzy neural network for a given forecasting task and is much easier to use in practice. The proposed approach is tested on real-load data from the Southern Power Network of Hebei Province, China, and on the EUNITE competition data. Results show the proposed approach improves existing load forecasting models. [ABSTRACT FROM AUTHOR]

This paper investigates whether spike-timing-dependent plasticity (STDP) can minimize the effect of mismatch within the context of a depth-from-motion algorithm. To improve noise rejection, this algorithm contains a spike prediction element, whose performance is degraded by analog very large scale integration (VLSI) mismatch. The error between the actual spike arrival time and the prediction is used as the input to an STDP circuit, to improve future predictions. Before STDP adaptation, the error reflects the degree of mismatch within the prediction circuitry. After STDP adaptation, the error indicates to what extent the adaptive circuitry can minimize the effect of transistor mismatch. The circuitry is tested with static and varying prediction times and chip results are presented. The effect of noisy spikes is also investigated. Under all conditions the STDP adaptation is shown to improve performance. [ABSTRACT FROM AUTHOR]

Despite its great importance, there has been no general consensus on how to model the trends in time-series data. Compared to traditional approaches, neural networks (NNs) have shown some promise in time-series forecasting. This paper investigates how to best model trend time series using NNs. Four different strategies (raw data, raw data with time index, detrending, and dif- ferencing) are used to model various trend patterns (linear, non- linear, deterministic, stochastic, and breaking trend). We find that with NNs differencing often gives meritorious results regardless of the underlying data generating processes (DGPs). This finding is also confirmed by the real gross national product (GNP) series. [ABSTRACT FROM AUTHOR]

Forecasting of time series that have seasonal and other variations remains an important problem for forecasters. This paper presents a neural network (NN) approach to forecasting quarterly time series. With a large data set of 756 quarterly time series from the M3 forecasting competition, we conduct a comprehensive investigation of the effectiveness of several data preprocessing and modeling approaches. We consider two data preprocessing methods and 48 NN models with different possible combinations of lagged observations, seasonal dummy variables, trigonometric variables, and time index as inputs to the NN. Both parametric and nonparametric statistical analyses are performed to identify the best models under different circumstances and categorize similar models. Results indicate that simpler models, in general, outperform more complex models. In addition, data preprocessing especially with deseasonalization and detrending is very helpful in improving NN performance. Practical guidelines are also provided. [ABSTRACT FROM AUTHOR]

Short-run forecasting of electricity prices has become necessary for power generation unit schedule, since it is the basis of every profit maximization strategy. In this article a new and very easy method to compute accurate forecasts for electricity prices using mixed models is proposed. The main idea is to develop an efficient tool for one-step-ahead forecasting in the future, combining several prediction methods for which forecasting performance has been checked and compared for a span of several years. Also as a novelty, the 24 hourly time series has been modelled separately, instead of the complete time series of the prices. This allows one to take advantage of the homogeneity of these 24 time series. The purpose of this paper is to select the model that leads to smaller prediction errors and to obtain the appropriate length of time to use for forecasting. These results have been obtained by means of a computational experiment. A mixed model which combines the advantages of the two new models discussed is proposed. Some numerical results for the Spanish market are shown, but this new methodology can be applied to other electricity markets as well. [ABSTRACT FROM AUTHOR]

Medium-term electric energy demand forecasting is an essential tool for power system planning and operation, mainly in those countries whose power systems operate in a deregulated environment. This paper proposes a novel approach to monthly electric energy demand time series forecasting, in which it is split into two new series: the trend and the fluctuation around it. Then two neural networks are trained to forecast them separately. These predictions are added up to obtain an overall forecasting. Several methods have been tested to find out which of them provides the best performance in the trend extraction. The proposed technique has been applied to the Spanish peninsular monthly electric consumption. The results obtained are better than those reached when only one neural network was used to forecast the original consumption series and also than those obtained with the ARIMA method. [ABSTRACT FROM AUTHOR]

This paper aims to develop a load forecasting method for short-term load forecasting, based on an adaptive two-stage hybrid network with self-organized map (SOM) and support vector machine (SVM). In the first stage, a SOM network is applied to cluster the input data set into several subsets in an unsupervised manner. Then, groups of 24 SVMs for the next day's load profile are used to fit the training data of each subset in the second stage in a supervised way. The proposed structure is robust with different data types and can deal well with the nonstationarity of load series. In particular, our method has the ability to adapt to different models automatically for the regular days and anomalous days at the same time. With the trained network, we can straight-forwardly predict the next-day hourly electricity load. To confirm the effectiveness, the proposed model has been trained and tested on the data of the historical energy load from New York Independent System Operator. [ABSTRACT FROM AUTHOR]

The objective of this paper is to present a new methodology for midterm energy forecasting. The proposed model is an adaptive artificial neural network (ANN), which properly transforms the input variables to differences or relative differences, in order to predict energy values not included in the training set. The ANN parameters, such as the finally used input variables, the number of neurons, initial values, and time periods of momentum term and training rate, are simultaneously selected by an optimization process. Another characteristic of the model is the use of a minimal training set of patterns. Results from an extensive analysis conducted by the developed method for the Greek power system and for different categories of customers are compared to those obtained from the application of standard regression methods. [ABSTRACT FROM AUTHOR]

This paper presents an approach for short-term load forecast problem, based on hybrid correction method. Conventional artificial neural network based short-term load forecasting techniques have limitations especially when weather changes are seasonal. Hence, we propose a load correction method by using fuzzy logic approach in which a fuzzy logic, based on similar days, corrects the neural network output to obtain the next day forecasted load. An Euclidean norm with weighted factors is used for the selection of similar days. The load correction method for the generation of new similar days is also proposed. The neural network has an advantage of dealing with the nonlinear parts of the forecasted load curves, whereas, the fuzzy rules are constructed based on the expert knowledge. Therefore, by combining these two methods, the test results show that the proposed forecasting method could provide a considerable improvement of the forecasting accuracy especially as it shows how to reduce neural network forecast error over the test period by 23 % through the application of a fuzzy logic correction. The suitability of the proposed approach is illustrated through an application to actual load data of the Okinawa Electric Power Company in Japan. [ABSTRACT FROM AUTHOR]