Hirofumi Toyama

Notice of Violation of IEEE Publication Principles A Hybrid ARIMA and Neural Network Model for Short-Term Price Forecasting in Deregulated Market

In the framework of competitive electricity markets, power producers and consumers need accurate price forecasting tools. Price forecasts embody crucial information for producers and consumers when planning bidding strategies in order to maximize their benefits and utilities, respectively. The choice of the forecasting model becomes the important influence factor on how to improve price forecasting accuracy. This paper provides a hybrid methodology that combines both autoregressive integrated moving average (ARIMA) and artificial neural network (ANN) models for predicting short-term electricity prices. This method is examined by using the data of Australian national electricity market, New South Wales, in the year 2006. Comparison of forecasting performance with the proposed ARIMA, ANN, and hybrid models are presented. Empirical results indicate that a hybrid ARIMA-ANN model can improve the price forecasting accuracy.

Thermal unit commitment strategy with solar and wind energy systems using genetic algorithm operated particle swarm optimization

This paper presents a methodology for solving unit commitment problem for thermal units integrated with wind and solar energy systems. The renewable energy sources are included in this model due to their low electricity cost and positive effect on environment. The unit commitment problem is solved by a genetic algorithm operated improved binary particle swarm optimization (PSO) algorithm. Unlike trivial PSO, this algorithm runs the refinement process of the solutions within multiple populations. Some genetic algorithm operators such as crossover, elitism, mutation are applied within the higher potential solutions to generate new solutions for next population. The PSO includes a new variable for updating velocity in accordance with population best with particle best and global best. The algorithm performs effectively in various sized thermal power system with equivalent solar and wind energy system and is able to produce high quality (minimized production cost) solutions. The simulation results show the effectiveness of this algorithm by comparing the outcome with several established methods.

Combination of artificial neural network and ARIMA time series models for short term price forecasting in deregulated market

In the framework of competitive electricity markets, power producers and consumers need accurate price forecasting tools. Price forecasts embody crucial information for producers and consumers when planning bidding strategies in order to maximize their benefits and utilities, respectively. The choice of the forecasting model becomes the important influence factor how to improve price forecasting accuracy. In this paper provides a combination methodology that combines both ARIMA and ANN models for predicting short term electricity prices. This method is examined by using the data of Australian National Electricity Market (NEM), New South Wales regional in year 2006. Comparison of forecasting performance with the proposed ARIMA and ARIMA-ANN models are presented. Empirical results indicate that an ARIMA-ANN model can improve the price forecasting accuracy.

A new method for next-day price forecasting for PJM electricity market

In the framework of the competitive electricity markets, electricity price forecasting is important for market participants in a deregulated electricity market. Rather than forecasting the value, market participants are sometimes more interested interval of the peak electricity price forecasting. Forecasting the peak price is essential for estimating the uncertainty involved in the price and thus is highly useful for making generation bidding strategies and investment decisions. The choice of the forecasting model becomes the important influence factor how to improve price forecasting accuracy. This paper proposes new approach to reduce the prediction error at occurrence time of the peak electricity price, and aims to enhance the accuracy of the next day electricity price forecasting. In the proposed method, the weekly variation data is used for input factors of the ANN at occurrence time of the peak electricity price in order to catch the price variation. Moreover, learning data for the ANN is selected by rough sets theory at occurrence time of the peak electricity price. This method is examined by using the data of the PJM electricity market. From the simulation results, it is observed that the proposed method provides a more accurate and effective forecasting, which helpful for suitable bidding strategy and risk management tool for market participants in a deregulated electricity market.

Next-day electricity price forecasting on deregulated power market

This paper proposes the approach to reduce the prediction error at occurrence time of peak electricity price, and aims to enhance the accuracy of next day electricity price forecasting. In the proposed method, the weekly variation data is used for input factors of the NN at occurrence time of peak electricity price in order to catch the price variation. Moreover, learning data for the neural network (NN) is selected by rough sets theory at occurrence time of peak electricity price. This method is examined by using the data of PJM electricity market.

Thermal generation scheduling strategy using binary clustered particle swarm optimization algorithm

This paper presents a multi-population binary clustered particle swarm optimization (BCPSO) algorithm to solve short term thermal generation scheduling problem. The potential solution schedules are distributed among several clusters based on their corresponding fitness values. Each cluster contains a cluster best schedule. Each solution of a particular population then flies through to its cluster space towards the cluster best as well as personal best solution instead of only global best (unlike trivial PSO). Therefore, this algorithm provides a way to explore larger search space and thus reduces the probability of local trapping. An intelligent mutation operator is applied on the solutions of the highest fitted cluster for a particular population. This mutation rate is adjusted dynamically based on the population number to reduce solution disturbance. Simulation result is provided to show the effectiveness of BCPSO.

Next day price forecasting for electricity market

This paper proposes new approach to reduce the prediction error at occurrence time of the peak price, and aims to enhance the accuracy of the next day price forecasting. In the proposed method, the weekly variation data is used for input factors of the ANN at occurrence time of the peak price in order to catch the price variation. Moreover, learning data for the ANN is selected by rough sets theory at occurrence time of the peak price. From the simulation results, it is observed that the proposed method provides a more accurate and effective forecasting, which helpful for suitable bidding strategy and risk management tool for market participants in a deregulated electricity market.

Next-day peak electricity price forecasting using NN based on rough sets theory

This paper proposes an approach for next-day peak electricity price forecasting using neural networks (NN), based on rough sets. In the proposed method, input factors of the NN are selected by using correlation analysis. Moreover, learning data used for training of the NN, is selected by rough sets. The proposed method for creating learning data based on temperature fluctuation is used for generation of new learning data. The proposed method is examined by using the data of PJM electricity market. From the simulation results, it is observed that the proposed method is useful for next-day peak electricity price forecasting.

Generation scheduling methodology for thermal units with wind energy system considering unexpected load deviation

This paper presents a methodology of short term generation scheduling (unit commitment) for thermal units integrated with wind energy system considering the unexpected deviation on load demand. The deviation in load demand occurs mainly due to variation in temperature which in turns yields error in load forecasting. Since the usual unit commitment (UC) scheduling as well as economic power dispatch procedures are based on predicted load demand, the sudden deviation results non optimal solution and hence increases the thermal unit fuel cost. This method tracks down the load deviation at a particular hour and using a sophisticated load forecasting technique (based on neural network) re-predicts the load demand for the hours to come. This way a relatively accurate load forecasting is achieved and the learning process of neural network is improved which will eventually reduce the fuel cost. Meanwhile the fuel cost is further minimized by the inclusion of wind energy system with the base thermal unit system. A genetic algorithm (GA) is used to solve the UC problem with some useful problem specific operators. Simulation results show the effectiveness of this proposed method considering various cases temperature deviations.