Electricity Price Prediction Research Articles

A proposed method using neural network for electricity price prediction base on genetic algorithm

In predicting the electricity price, neural network is used as an applied model. To implement the neural network there are several parameters that we must determine such as learning rate and momentum, the problem is there is no standard guideline in determining the parameters to be used in this method, so that the experimental method is used. Therefore, we need method that can solve these problems, so parameters determination can be more optimal. The solution that can be applied is using genetic algorithm on neural network by optimizing the value of learning rate, momentum and training cycles parameter. The expected result is it can accelerate the process in getting the appropriate and optimal parameter values on neural network, so it can increase the accuracy of the prediction in predicting electricity price.

An Optimized Heterogeneous Structure LSTM Network for Electricity Price Forecasting

Electricity price is an important indicator of the market operation. Accurate prediction of electricity price will facilitate the maximization of economic benefits and reduction of risks to the power market. At the same time, because of the excellent performance of deep learning models, using long-short term memory neural network (LSTM) and other deep learning models to predict time series has gradually become a research hotspot. In this paper, an optimized heterogeneous structure LSTM model is proposed to solve the problems of the single network structure and hyperparameter selection existing in the current research on LSTM. The heterogeneous structure LSTM is constructed based on the decomposed and reconstructed electricity price data, and sequence model-based optimization (SMBO) is used to optimize its hyperparameters. In order to verify the proposed model, online hourly forecasting and day-ahead hourly forecasting are tested on the electricity markets of Pennsylvania-New Jersey-Maryland (PJM). The experimental results show that the performance of the proposed model is much better than that of the general LSTM model and traditional models in accuracy and stability, providing a new idea for the use of LSTM for time series prediction.

Short-term electricity price forecasting and classification in smart grids using optimized multikernel extreme learning machine

Short-term electricity price forecasting in deregulated electricity markets has been studied extensively in recent years but without significant reduction in price forecasting errors. Also demand-side management and short-term scheduling operations in smart grids do not require strictly very accurate forecast and can be executed with certain practical price thresholds. This paper, therefore, presents a multikernel extreme learning machine (MKELM) for both short-term electricity price forecasting and classification according to some prespecified price thresholds. The kernel ELM does not require the hidden layer mapping function to be known and produces robust prediction and classification in comparison with the conventional ELM using random weights between the input and hidden layers. Further in the MKELM formulation, the linear combination of the weighted kernels is optimized using vaporization precipitation-based water cycle algorithm (WCA) to produce significantly accurate electricity price prediction and classification. The combination of MKELM and WCA is named as WCA-MKELM in this work. To validate the effectiveness of the proposed approach, three electricity markets, namely PJM, Ontario and New South Wales, are considered for electricity price forecasting and classification producing fairly accurate results.

A Multi-Stage Intelligent Model for Electricity Price Prediction Based on the Beveridge–Nelson Disintegration Approach

Accurate electricity price prediction is key to the orderly operation of the electricity market. However, the uncertain, stochastic and fluctuant characteristics of electricity pricees make prediction difficult. With the aim of solving this issue, this investigation proposed a multi-stage intelligent model integrating the Beveridge–Nelson decomposition (B-N-D) model, the least square support vector machine (LSSVM), and a nature-inspired optimization model named the whale optimization algorithm (WOA). Firstly, the B-N-D model was utilized to decompose the hourly electricity price time series into determinacy component, periodic trend, and stochastic item. Secondly, the WOA–LSSVM model was proposed to forecast the future hourly data of three components respectively, of which the optimal parameters of LSSVM were determined by using WOA. Finally, the future hourly electricity price data were computed by multiplying the forecasted data of those terms. To verify the validity of the proposed electricity price prediction model in this paper, five comparison approaches based on the B-N-D approach were selected, which are auto-regressive integrated moving average (ARIMA), single LSSVM, LSSVM optimized by the fruit-fly optimization algorithm (FOA), LSSVM optimized by particle swarm optimization (PSO) models, and WOA–LSSVM without B-N-D. By comparatively analyzing the error criteria values of the above models through testing on the objective data of the Pennsylvania–New Jersey–Maryland (PJM) electricity market collected from 11 December 2017 to 18 December 2017, from 15 January 2018 to 22 January 2018, and from 1 February 2018 to 25 February 2018, we conclude that the constructed intelligent model in this paper can greatly enhance the prediction precision of electricity prices.

Effective long short-term memory with differential evolution algorithm for electricity price prediction

Electric power, as an efficient and clean energy, has considerable importance in industries and human lives. Electricity price is becoming increasingly crucial for balancing electricity generation and consumption. In this study, long short-term memory (LSTM) with the differential evolution (DE) algorithm, denoted as DE–LSTM, is used for electricity price prediction. Several recent studies have adopted LSTM with considerable success in certain applications, such as text recognition and speech recognition. However, problems in the application of LSTM to solving nonlinear regression and time series problems have been encountered. DE, a novel evolutionary algorithm that effectively obtains optimal solutions, is designed to identify suitable hyperparameters for LSTM. Experiments are conducted to verify the performance of the DE–LSTM model under the electricity prices in New South Wales, Germany/Austria, and France. Results indicate that the proposed DE–LSTM model outperforms existing forecasting models in terms of forecasting accuracies.

An Hour Ahead Electricity Price Forecasting with Least Square Support Vector Machine and Bacterial Foraging Optimization Algorithm

<span lang="EN-US">Predicting electricity price has now become an important task in power system operation and planning. An hour-ahead forecast provides market participants with the pre-dispatch prices for the next hour. It is beneficial for an active bidding strategy where amount of bids can be reviewed or modified before delivery hours. However, only a few studies have been conducted in the field of hour-ahead forecasting. This is due to most power markets apply two-settlement market structure (day-ahead and real time) or standard market design rather than single-settlement system (real time). Therefore, a hybrid multi-optimization of Least Square Support Vector Machine (LSSVM) and Bacterial Foraging Optimization Algorithm (BFOA) was designed in this study to produce accurate electricity price forecasts with optimized LSSVM parameters and input features. So far, no works has been established on multistage feature and parameter optimization using LSSVM-BFOA for hour-ahead price forecast. The model was examined on the Ontario power market. A huge number of features were selected by five stages of optimization to avoid from missing any important features. The developed LSSVM-BFOA shows higher forecast accuracy with lower complexity than most of the existing models.</span>

An Electricity Price Forecasting Model by Hybrid Structured Deep Neural Networks

Electricity price is a key influencer in the electricity market. Electricity market trades by each participant are based on electricity price. The electricity price adjusted with the change in supply and demand relationship can reflect the real value of electricity in the transaction process. However, for the power generating party, bidding strategy determines the level of profit, and the accurate prediction of electricity price could make it possible to determine a more accurate bidding price. This cannot only reduce transaction risk, but also seize opportunities in the electricity market. In order to effectively estimate electricity price, this paper proposes an electricity price forecasting system based on the combination of 2 deep neural networks, the Convolutional Neural Network (CNN) and the Long Short Term Memory (LSTM). In order to compare the overall performance of each algorithm, the Mean Absolute Error (MAE) and Root-Mean-Square error (RMSE) evaluating measures were applied in the experiments of this paper. Experiment results show that compared with other traditional machine learning methods, the prediction performance of the estimating model proposed in this paper is proven to be the best. By combining the CNN and LSTM models, the feasibility and practicality of electricity price prediction is also confirmed in this paper.

Forecasting spot electricity prices: Deep learning approaches and empirical comparison of traditional algorithms

In this paper, a novel modeling framework for forecasting electricity prices is proposed. While many predictive models have been already proposed to perform this task, the area of deep learning algorithms remains yet unexplored. To fill this scientific gap, we propose four different deep learning models for predicting electricity prices and we show how they lead to improvements in predictive accuracy. In addition, we also consider that, despite the large number of proposed methods for predicting electricity prices, an extensive benchmark is still missing. To tackle that, we compare and analyze the accuracy of 27 common approaches for electricity price forecasting. Based on the benchmark results, we show how the proposed deep learning models outperform the state-of-the-art methods and obtain results that are statistically significant. Finally, using the same results, we also show that: (i) machine learning methods yield, in general, a better accuracy than statistical models; (ii) moving average terms do not improve the predictive accuracy; (iii) hybrid models do not outperform their simpler counterparts.

Metascheduling of HPC Jobs in Day-Ahead Electricity Markets

High performance grid computing is a key enabler of large scale collaborative computational science. With the promise of exascale computing, high performance grid systems are expected to incur electricity bills that grow super-linearly over time. In order to achieve cost effectiveness in these systems, it is essential for the scheduling algorithms to exploit electricity price variations, both in space and time, that are prevalent in the dynamic electricity price markets. In this paper, we present a metascheduling algorithm to optimize the placement of jobs in a compute grid which consumes electricity from the day-ahead wholesale market. We formulate the scheduling problem as a Minimum Cost Maximum Flow problem and leverage queue waiting time and electricity price predictions to accurately estimate the cost of job execution at a system. Using trace based simulation with real and synthetic workload traces, and real electricity price data sets, we demonstrate our approach on two currently operational grids, XSEDE and NorduGrid. Our experimental setup collectively constitute more than 433K processors spread across 58 compute systems in 17 geographically distributed locations. Experiments show that our approach simultaneously optimizes the total electricity cost and the average response time of the grid, without being unfair to users of the local batch systems.

On the Use of Functional Additive Models for Electricity Demand and Price Prediction

This paper presents an application of functional additive models in the context of electricity demand and price prediction. Data from the Spanish electricity market are used to obtain the pointwise predictions. Also prediction intervals, based on a bootstrap procedure, are computed. This approach is compared with the use of other functional regression methods applied to the same data set by Aneiros et al. (2016).

The price prediction for the energy market based on a new method

Regarding the complex behaviour of price signalling, its prediction is difficult, where an accurate forecasting can play an important role in electricity markets. In this paper, a feature selection based on mutual information is implemented for day ahead prediction of electricity prices, which are so valuable for determining the redundancy and relevancy of selected features. A combination of wavelet transform (WT) and a hybrid forecast method is presented based on a neural network (NN). Furthermore, an intelligent algorithm is considered for a prediction process to set the proposed forecast engine free parameters based NN. This optimisation process improved the accuracy of the proposed model. To demonstrate the validity of this model, the Pennsylvania-New Jersey-Maryland (PJM) electricity market is considered as a test case and compared with some of the most recent price forecast methods. These comparisons illustrate the effectiveness of the proposed strategy.

Combination of fuzzy based on a meta-heuristic algorithm to predict electricity price in an electricity markets

The price forecasting is one of the most important issues in electricity markets. For this purpose, an accurate prediction model is demanded for optimal operation as well as planning in power system. In this work, a novel approach composed of Wavelet Transform and Takagi–Sugeno (TS) fuzzy rule-based system is proposed for day-ahead price forecasting of electricity markets. In this method, the input of price data is clustered by TS fuzzy model. In the identification of the TS fuzzy model, a hyperplane prototype fuzzy clustering model is proposed, which obtain the rules. Furthermore, in this model, a new stochastic search algorithm is applied to optimize the clustering objective function. To implement the proposed forecast strategy, the price data is first decomposed by Wavelet Transform (WT). Then, each produced wavelet component is filtered by two stage feature selection based on mutual information. Afterward, the hybrid fuzzy neural network-based forecast engine is used to predict the future values of each price component. This model, is tested on real-world electricity markets of Ontario and New England through the comparison with other techniques. Obtained results demonstrate the validity of proposed technique.

Comparing Variational and Empirical Mode Decomposition in Forecasting Day-Ahead Energy Prices

Recently, variational mode decomposition (VMD) has been proposed as an advanced multiresolution technique for signal processing. This study presents a VMD-based generalized regression neural network ensemble learning model to predict California electricity and Brent crude oil prices. Its performance is compared to that of the empirical mode decomposition (EMD) based generalized regression neural network (GRNN) ensemble model. Particle swarm optimization is used to optimize each GRNN initial weight within the ensemble system. Experimental results showed that the VMD-based ensemble outperformed EMD-based ensemble forecasting system in terms of mean absolute error, mean absolute percentage error, and root mean-squared error. It also outperformed the conventional auto-regressive moving average model used for comparison purpose. As a result, the VMD-based GRNN ensemble forecasting paradigm could be a promising methodology for California electricity and Brent crude oil price prediction.

An Adaptive Functional Autoregressive Forecast Model to Predict Electricity Price Curves

We propose an adaptive functional autoregressive (AFAR) forecast model to predict electricity price curves. With time-varying operators, the AFAR model can be safely used in both stationary and nonstationary situations. A closed-form maximum likelihood (ML) estimator is derived under stationarity. The result is further extended for nonstationarity, where the time-dependent operators are adaptively estimated under local homogeneity. We provide theoretical results of the ML estimator and the adaptive estimator. Simulation study illustrates nice finite sample performance of the AFAR modeling. The AFAR model also exhibits a superior accuracy in the forecast exercise of the California electricity daily price curves compared to several alternatives.

Alternative approaches to the prediction of electricity prices

PurposeThis paper aims to present the results of the comparison between different approaches to the prediction of electricity prices. It is well-known that the properties of the data generation process may prefer some modeling methods over the others. The data having an origin in social or market processes are characterized by unexpectedly wide realization space resulting in the existence of the long tails in the probabilistic density function. These data may not be easy in time series prediction using standard approaches based on the normal distribution assumptions. The electricity prices on the deregulated market fall into this category.Design/methodology/approachThe paper presents alternative approaches, i.e. memory-based prediction and fractal approach compared with established nonlinear method of neural networks. The appropriate interpretation of results is supported with the statistical data analysis and data conditioning. These algorithms have been applied to the problem of the energy price prediction on the deregulated electricity market with data from Polish and Austrian energy stock exchanges.FindingsThe first outcome of the analysis is that there are several situations in the task of time series prediction, when standard modeling approach based on the assumption that each change is independent of the last following random Gaussian bell pattern may not be a true. In this paper, such a case was considered: price data from energy markets. Electricity prices data are biased by the human nature. It is shown that more relevant for data properties was Cauchy probabilistic distribution. Results have shown that alternative approaches may be used and prediction for both data memory-based approach resulted in the best performance.Research limitations/implications“Personalization” of the model is crucial aspect in the whole methodology. All available knowledge should be used on the forecasted phenomenon and incorporate it into the model. In case of the memory-based modeling, it is a specific design of the history searching routine that uses the understanding of the process features. Importance should shift toward methodology structure design and algorithm customization and then to parameter estimation. Such modeling approach may be more descriptive for the user enabling understanding of the process and further iterative improvement in a continuous striving for perfection.Practical implicationsMemory-based modeling can be practically applied. These models have large potential that is worth to be exploited. One disadvantage of this modeling approach is large calculation effort connected with a need of constant evaluation of large data sets. It was shown that a graphics processing unit (GPU) approach through parallel calculation on the graphical cards can improve it dramatically.Social implicationsThe modeling of the electricity prices has big impact of the daily operation of the electricity traders and distributors. From one side, appropriate modeling can improve performance mitigating risks associated with the process. Thus, the end users should receive higher quality of services ultimately with lower prices and minimized risk of the energy loss incidents.Originality/valueThe use of the alternative approaches, such as memory-based reasoning or fractals, is very rare in the field of the electricity price forecasting. Thus, it gives a new impact for further research enabling development of better solutions incorporating all available process knowledge and customized hybrid algorithms.

Forecasting models for vertically bundled electricity market prices — A review and future trend

An electricity price forecasting is offers either for developing bidding strategies or for making investment decisions economically for use in power producers and consumers in the current deregulated electricity markets. During the past decade several new techniques are being used for electricity price forecasting to predict the future prices. In this paper conscientious activity is made to review the various electricity price forecasting models. Traditionally, hard and soft computing models could be used to predict electricity prices using single and hybrid models. The hard computing models such as time series AR, ARIMA, and GARCH and soft computing models such as ANN, FL, GA, SVM, and ELM used in electricity price forecasting models in various deregulated power system markets. In addition, it also suggests the alternate solution and future trend for numerical prediction or point-forecasting.

Rational and self-adaptive evolutionary extreme learning machine for electricity price forecast

Electricity price forecast is of great importance to electricity market participants. Given the sophisticated time-series of electricity price, various approaches of extreme learning machine (ELM) have been identified as effective prediction approaches. However, in high dimensional space, evolutionary extreme learning machine (E-ELM) is time-consuming and difficult to converge to optimal region when just relying on stochastic searching approaches. In the meanwhile, due to the complicated functional relationship, objective function of E-ELM seems difficult also to be mined directly for some useful mathematical information to guide the optimum exploring. This paper proposes a new differential evolution (DE) like algorithm to enhance E-ELM for more accurate and reliable prediction of electricity price. An approximation model for producing DE-like trail vector is the key mechanism, which can use simpler mathematical mapping to replace the original yet complicated functional relationship within a small region. Thus, the evolutionary procedure frequently dealt with some rational searching directions can make the E-ELM more robust and faster than supported only by the stochastic methods. Experimental results show that the new method can improve the performance of E-ELM more efficiently.

How does market power affect the impact of large scale wind investment in 'energy only' wholesale electricity markets?

In the short run, it is well known that increasing wind penetration is likely to reduce spot market electricity prices due to the merit order effect. The long run effect is less clear because there will be a change in new capacity investment in response to the wind penetration. In this paper we examine the interaction between capacity investment, wind penetration and market power by first using a least-cost generation expansion model to simulate capacity investment with increasing amounts of wind generation, and then using a computer agent-based model to predict electricity prices in the presence of market power. We find the degree to which firms are able to exercise market power depends critically on the ratio of capacity to peak demand. For our preferred long run generation scenario we show market power increases for some periods as wind penetration increases however the merit order counteracts this with the results that prices overall remain flat. Returns to peakers increase significantly as wind penetration increases. The market power in turn leads to inefficient dispatch which is exacerbated with large amounts of wind generation.

Prediction of the Italian electricity price for smart grid applications

In this paper we address the problem of one day-ahead hourly electricity price forecast for smart grid applications. To this aim, we investigate the application of a number of predictive models for time-series, including methods based on empirical strategies frequently adopted in the smart grid community, Kalman Filters and Echo State Networks (ESNs). The considered methods have been suitably modified to address the electricity price forecast problem. Strategies based on daily re-adaptation of models’ parameters are taken into consideration as well. The predictive performance achieved by the considered models is assessed, and the methods are compared among each other on recent real data from the Italian electricity market. As a result of the comparison over three years data, ESN methods appear to provide the most accurate price predictions, which could imply significant economic savings in many smart grid activities, such as switching on power plants to support power generation from renewable sources, electric vehicle recharging or usage of household appliances.

Control del sobreajuste en redes neuronales tipo cascada correlación aplicado a la predicción de precios de contratos de electricidad

Prediction of electricity prices is considered a difficult task due to the number and complexity of factors that influence their performance, and their relationships. Neural networks cascade correlation - CASCOR allows to do a constructive learning and it captures better the characteristics of the data; however, it has a high tendency to overfitting. To control overfitting in some areas regularization techniques are used. However, in the literature there are no studies that: i) use regularization techniques to control overfitting in CASCOR networks, ii) use CASCOR networks in predicting of electrical series iii) compare the performance with tra­ditional neural networks or statistical models. The aim of this paper is to model and predict the behavior of the price series of electricity contracts in Colombia, using CASCOR networks and controlling the overfitting by regularization techniques