Search results

This paper aims to Solar photovoltaic (PV) power can significantly impact the power system because of its intermittent nature. Hence, an accurate solar PV power forecasting model is required for appropriate power system planning.

In this paper, a long short-term memory (LSTM)-based double deep Q-learning (DDQL) neural network (NN) is proposed for forecasting solar PV power indirectly over the long-term horizon. The past solar irradiance, temperature and wind speed are used for forecasting the solar PV power for a place using the proposed forecasting model.

The LSTM-based DDQL NN reduces over- and underestimation and avoids gradient vanishing. Thus, the proposed model improves the forecasting accuracy of solar PV power using deep learning techniques (DLTs). In addition, the proposed model requires less training time and forecasts solar PV power with improved stability.

The proposed model is trained and validated for several places with different climatic patterns and seasons. The proposed model is also tested for a place with a temperate climatic pattern by constructing an experimental solar PV system. The training, validation and testing results have confirmed the practicality of the proposed solar PV power forecasting model using LSTM-based DDQL NN.

This study aims to short-therm forecasting of power generation output for this purpose, an adaptive neuro-fuzzy inference system (ANFIS) is designed to forecast the output power of power plant based on climate factors considering wind speed and wind direction simultaneously.

Several methods and algorithms have been proposed for systems forecasting in various fields. One of the strongest methods for modeling complex systems is neuro-fuzzy that refers to combinations of artificial neural network and fuzzy logic. When the system becomes more complex, the conventional algorithms may fail for network training. In this paper, an integrated approach, including ANFIS and metaheuristic algorithms, is used for increasing forecast accuracy.

Power generation in power plants is dependent on various factors, especially climate factors. Operating power plant in Iran is very much influenced because of climate variation, including from tropical to subpolar, and severely varying temperature, humidity and air pressure for each region and each season. On the other hands, when wind speed and wind direction are used simultaneously, the training process does not converge, and the forecasting process is unreliable. The real case study is mentioned to show the ability of the proposed approach to remove the limitations.

First, ANFIS is applied for forecasting based on climate factors, including wind speed and wind direction, that have rarely been used simultaneously in previous studies. Second, the well-known and more widely used metaheuristic algorithms are applied to improve the learning process for forecasting output power and compare the results.

The objective of the paper is to explore the out-of-sample forecasting connections in income growth across the globe.

An autoregressive distributed lag (ARDL) framework is employed and the forecasting performance is analyzed across several horizons using different forecast combination techniques.

Results show that the foreign country's income provides superior forecasts beyond what is provided by the country's own past income movements. Superior forecasting power is particularly held by Belgium, Korea, New Zealand, the UK and the US, while these countries' income is rather difficult to predict by global counterparts. Contrary to conventional wisdom, improved forecasts of income can be obtained even for longer horizons using our approach. Results also show that the forecast combination techniques yield higher forecasting gains relative to individual model forecasts, both in magnitude and the number of countries.

The forecasting paths of income movement across the globe reveal that predictive power greatly differs across countries, regions and forecast horizons. The countries that are difficult to predict in the short run are often seen to be predictable by global income movements in the long run.

Even while it is difficult to predict the income movements at an individual country level, combining information from the income growth of several countries is likely to provide superior forecasting gains. And these gains are higher for long-horizon forecasts as compared to the short-horizon forecast.

In evaluating the forward-looking social implications of economic policy changes, the policymakers should also consider the possible global forecasting connections revealed in the study.

Employing an ARDL model to explore global income forecasting connections across several forecast horizons using different forecast combination techniques.

Solar power forecasting will have a significant impact on the future of large-scale renewable energy plants. Predicting photovoltaic power generation depends heavily on climate conditions, which fluctuate over time. In this research, we propose a hybrid model that combines machine-learning methods with Theta statistical method for more accurate prediction of future solar power generation from renewable energy plants. The machine learning models include long short-term memory (LSTM), gate recurrent unit (GRU), AutoEncoder LSTM (Auto-LSTM) and a newly proposed Auto-GRU. To enhance the accuracy of the proposed Machine learning and Statistical Hybrid Model (MLSHM), we employ two diversity techniques, i.e. structural diversity and data diversity. To combine the prediction of the ensemble members in the proposed MLSHM, we exploit four combining methods: simple averaging approach, weighted averaging using linear approach and using non-linear approach, and combination through variance using inverse approach. The proposed MLSHM scheme was validated on two real-time series datasets, that sre Shagaya in Kuwait and Cocoa in the USA. The experiments show that the proposed MLSHM, using all the combination methods, achieved higher accuracy compared to the prediction of the traditional individual models. Results demonstrate that a hybrid model combining machine-learning methods with statistical method outperformed a hybrid model that only combines machine-learning models without statistical method.

The purpose of this paper is to improve the accuracy of short time load forecasting to ensure the economical and safe operation of power systems. The traditional neural network applied in time series like load forecasting, easily plunges into local optimum and has a complicated learning process, leading to relatively slow calculating speed. On the basis of existing literature, the authors carried out studies in an effort to optimize a new recurrent neural network by wavelet analysis to solve the previous problems.

The main technique the authors applied is referred to as echo state network (ESN). Detailed information has been acquired by the authors using wavelet analysis. After obtaining more information from original time series, different reservoirs can be built for each subsequence. The proposed method is tested by using hourly electricity load data from a southern city in China. In addition, some traditional methods are also applied for the same task, as contrast.

The experiment has led the authors to believe that the optimized model is encouraging and performs better. Compared with standard ESN, BP network and SVM, the experimental results indicate that WS‐ESN improves the prediction accuracy and has less computing consumption.

The paper develops a new method for short time load forecasting. Wavelet decomposition is employed to pre‐process the original load data. The approximate part associated with low frequencies and several detailed parts associated with high frequencies components give expression to different information from original data. According to this, suitable ESN is chosen for each sub‐sequence, respectively. Therefore, the model combining the advantages of both ESN and wavelet analysis improves the result for short time load forecasting, and can be applied to other time series problem.

Accurate prediction of seasonal power consumption trends with impact disturbances provides a scientific basis for the flexible balance of the long timescale power system. Consequently, it fosters reasonable scheduling plans, ensuring the safety of the system and improving the economic dispatching efficiency of the power system.

First, a new seasonal grey buffer operator in the longitudinal and transverse dimensional perspectives is designed. Then, a new seasonal grey modeling approach that integrates the new operator, full real domain fractional order accumulation generation technique, grey prediction modeling tool and fruit fly optimization algorithm is proposed. Moreover, the rationality, scientificity and superiority of the new approach are verified by designing 24 seasonal electricity consumption forecasting approaches, incorporating case study and amalgamating qualitative and quantitative research.

Compared with other comparative models, the new approach has superior mean absolute percentage error and mean absolute error. Furthermore, the research results show that the new method provides a scientific and effective mathematical method for solving the seasonal trend power consumption forecasting modeling with impact disturbance.

Considering the development trend of longitudinal and transverse dimensions of seasonal data with impact disturbance and the differences in each stage, a new grey buffer operator is constructed, and a new seasonal grey modeling approach with multi-method fusion is proposed to solve the seasonal power consumption forecasting problem.

The highlights of the paper are as follows:

1. A new seasonal grey buffer operator is constructed.

2. The impact of shock perturbations on seasonal data trends is effectively mitigated.

3. A novel seasonal grey forecasting approach with multi-method fusion is proposed.

4. Seasonal electricity consumption is successfully predicted by the novel approach.

5. The way to adjust China's power system flexibility in the future is analyzed.

A new seasonal grey buffer operator is constructed.

The impact of shock perturbations on seasonal data trends is effectively mitigated.

A novel seasonal grey forecasting approach with multi-method fusion is proposed.

Seasonal electricity consumption is successfully predicted by the novel approach.

The way to adjust China's power system flexibility in the future is analyzed.

Given the effects of natural and social factors, data on both the supply and demand sides of electricity will produce obvious seasonal fluctuations. The purpose of this article is to propose a new dynamic seasonal grey model based on PSO-SVR to forecast the production and consumption of electric energy.

In the model design, firstly, the parameters of the SVR are initially optimized by the PSO algorithm for the estimation of the dynamic seasonal operator. Then, the seasonal fluctuations in the electricity demand data are eliminated using the dynamic seasonal operator. After that, the time series after eliminating of the seasonal fluctuations are used as the training set of the DSGM(1, 1) model, and the corresponding fitted, and predicted values are calculated. Finally, the seasonal reduction is performed to obtain the final prediction results.

This study found that the electricity supply and demand data have obvious seasonal and nonlinear characteristics. The dynamic seasonal grey model based on PSO-SVR performs significantly better than the comparative model for hourly and monthly data as well as for different time durations, indicating that the model is more accurate and robust in seasonal electricity forecasting.

Considering the seasonal and nonlinear fluctuation characteristics of electricity data. In this paper, a dynamic seasonal grey model based on PSO-SVR is established to predict the consumption and production of electric energy.

Load forecasting is important to any electrical grid, but for the developing and third-world countries with power shortages, load forecasting is essential. When planed load shedding programs are implemented to face power shortage, a noticeable distortion to the load curves will happen, and this will make the load forecasting more difficult.

In this paper, a new load forecasting model is developed that can detect the effect of planned load shedding on the power consumption and estimate the load curve behavior without the shedding and with different shedding programs. A neuro-Fuzzy technique is used for the model, which is trained and tested with real data taken from one of the 11 KV feeders in Najaf city in Iraq to forecast the load for two days ahead for the four seasons. Load, temperature, time of the day and load shedding schedule for one month before are the input parameters for the training, and the load forecasting data for two days are estimated by the model.

To verify the model, the load is forecasted without shedding by the proposed model and compared to real data without shedding and the difference is acceptable.

The proposed model provides acceptable forecasting with the load shedding effect available and better than other models. The proposed model provides expected behavior of load with different shedding programs an issue helps to select the appropriate shedding program. The proposed model is useful to estimate the real demands by assuming load shedding hours to be zero and forecast the load. This is important in places suffer from grid problems and cannot supply full loads to calculate the peak demands as the case in Iraq.

The purpose of this paper is to evaluate the forecasting performance of grey prediction models on educational attainment vis-à-vis that of exponential smoothing combined with multiple linear regression employed by the National Center for Education Statistics (NCES).

An out-of-sample forecasting experiment was carried out to compare the forecasting performances on educational attainments among GM(1,1), GM(1,1) rolling, FGM(1,1) derived from the grey system theory and exponential smoothing prediction combined with multivariate regression. The predictive power of each model was measured based on MAD, MAPE, RMSE and simple F-test of equal variance.

The forecasting efficiency evaluated by MAD, MAPE, RMSE and simple F-test of equal variance revealed that the GM(1,1) rolling model displays promise for use in forecasting educational attainment.

Since the possible inadequacy of MAD, MAPE, RMSE and F-type test of equal variance was documented in the literature, further large-scale forecasting comparison studies may be done to test the prediction powers of grey prediction and its competing out-of-sample forecasts by other alternative measures of accuracy.

The findings of this study would be useful for NCES and professional forecasters who are expected to provide government authorities and education policy makers with accurate information for planning future policy directions and optimizing decision-making.

As a continuing effort to evaluate the forecasting efficiency of grey prediction models, the present study provided accumulated evidence for the predictive power of grey prediction on short-term forecasts of educational statistics.

Study conditions of empirical publications on time series modeling and forecasting of electricity prices vary widely, making it difficult to generalize results. The key purpose of the present study is to offer a comparison of different model types and modeling conditions regarding their forecasting performance.

The authors analyze the forecasting performance of AR (autoregressive), MA (moving average), ARMA (autoregressive moving average) and GARCH (generalized autoregressive moving average) models with and without the explanatory variables, that is, power consumption and power generation from wind and solar. Additionally, the authors vary the detailed model specifications (choice of lag-terms) and transformations (using differenced time series or log-prices) of data and, thereby, obtain individual results from various perspectives. All analyses are conducted on rolling calibrating and testing time horizons between 2010 and 2014 on the German/Austrian electricity spot market.

The main result is that the best forecasts are generated by ARMAX models after spike preprocessing and differencing the data.

The present study extends the existing literature on electricity price forecasting by conducting a comprehensive analysis of the forecasting performance of different time series models under varying market conditions. The results of this study, in general, support the decision-making of electricity spot price modelers or forecasting tools regarding the choice of data transformation, segmentation and the specific model selection.