The modernisation of the world has significantly reduced the prime sources of energy such as coal, diesel and gas. Thus, alternative energy sources based on renewable energy have been a major focus nowadays to meet the world's energy demand and at the same time to reduce global warming. Among these energy sources, solar energy is a major source of alternative energy that is used to generate electricity through photovoltaic (PV) system. However, the performance of the power generated is highly sensitive on climate and seasonal factors. The unpredictable behaviour of the climate affects the power output and causes an unfavourable impact on the stability, reliability and operation of the grid. Thus an accurate forecasting of PV output is a crucial requirement to ensure the stability and reliability of the grid. This study provides a systematic and critical review on the methods used to forecast PV power output with main focus on the metaheuristic and machine learning methods. Advantages and disadvantages of each method are summarised, based on historical data along with forecasting horizons and input parameters. Finally, a comprehensive comparison between machine learning and metaheuristic methods is compiled to assist researchers in choosing the best forecasting technique for future research.

https://ietresearch.onlinelibrary.wiley.com/doi/epdf/10.1049/iet-rpg.2018.5649

Review on forecasting of photovoltaic power generation based on machine learning and metaheuristic techniques

Global solar radiation is generally measured on a horizontal surface, whereas the maximum amount of incident solar radiation is measured on an inclined surface. Over the last decade, a number of models were proposed for predicting solar radiation on inclined surfaces. These models have various scopes; applicability to specific surfaces, the requirement for special measuring equipment, or limitations in scope. To find the most suitable model for a given location the hourly outputs predicted by available models are compared with the field measurements of the given location. The main objective of this study is to review on the estimation of the most accurate model or models for estimating solar radiation components for a selected location, by testing various models available in the literature. To increase the amount of incident solar radiation on photovoltaic (PV) panels, the PV panels are mounted on tilted surfaces. This article also provides an up-to-date status of different optimum tilt angles that have been determined in various countries.

Estimation of Hourly, Daily and Monthly Global Solar Radiation on Inclined Surfaces: Models Re-Visited

This paper reviews the technological and economical feasibilities as well as sustainable assessment of approaches (thermochemical and biochemical) applied for sustainable “drop-in” fuel production from lignocellulosic sources. The challenges for each pathway to produce “drop-in” fuels are covered. Currently “drop-in” fuel production cost is approximately 2 times (~5–6$/gallon) higher than fossil fuels (3$/gallon), especially with the use of 2nd generation feedstocks. The primary sources of cost with “drop-in” fuel production are feedstock cost (40–60% of the total production cost), syngas cleaning and conditioning to meet Fischer-Tropsch synthesis requirement (12–15% of the total production cost) and bio oil upgrading (14–18% of the total production cost) in the case of pyrolysis and hydrothermal liquefaction (HTL). The most influential factors on the life cycle analysis (LCA) were biomass cultivation, harvesting, biomass pre-treatment, and transportation. Therefore, robust processes that can use local waste biomass are far more environmental and economically viable, especially as biofuel from second generation have a greater potential to reduce greenhouse gas emissions (50–100%) than first generation biofuels (50–90%) when land use changes are omitted in the LCA. The sustainability of biofuels is pre-dominantly dependant on the sustainability of the initial biomass, with 2nd generation feedstocks being more sustainable than 1st generation. Gasification-FTS is considered as the most promising technique for “drop-in” fuel production over pyrolysis and HTL due to its flexibility towards feedstock acceptance and the ability to produced high yields of liquid fuel together with other economically viable biofuels such as electricity and heat. Biochemical routes (i.e.fermentation) to “drop-in” fuels are still in their early development stages, and therefore require more studies and pilot-scale experiments in order to discover an economic and sustainable means of using these methods.

“Drop-in” fuel production from biomass: Critical review on techno-economic feasibility and sustainability

Short-term power prediction for photovoltaic power plants using a hybrid improved Kmeans-GRA-Elman model based on multivariate meteorological factors and historical power datasets

Over the past decade, significant progress in solar forecasting has been made. Nevertheless, there are concerns about duplication, long-term value, and reproducibility; this is referred to as the “solar forecasting bubble.” There is an urgent need to reconcile and improve the current solar forecasting research practice. This discussion paper proposes a so-called “ROPES” guideline that describes the desirable characteristics of future solar forecasting studies. In short, ROPES stands for reproducible, operational, probabilistic and/or physically based, ensemble, and skill. This set of characteristics is intended to facilitate comparison, comprehension, and communication within the solar forecasting field and speed up its development. Compliance with ROPES is evaluated on 79 solar forecasting references published during 2017 January to 2018 July in 6 Elsevier energy journals. Although most current papers fall short of complying with ROPES, evidence suggests that a consensus is forming.

A guideline to solar forecasting research practice: Reproducible, operational, probabilistic or physically-based, ensemble, and skill (ROPES)

The energy value and economic efficiency of solid biofuels produced from digestate and sawdust

The paper presents the use of classical statistical methods and methods based on neural modeling in short-term forecasting of electric energy from photovoltaic conversion. A detailed analysis of the input data measured in central Poland (Poznan, 52°25′ N, 16°56′ E) showed that some variables like air pressure and the length of the day are statistically insignificant. The values of kurtosis, skewness and results of applied tests, to check the normality of the distribution of dependent variable in the form of daily electricity production, indicate that the linear regression models should not be the only method in forecast process. The result of neural modeling using implemented network designer is RBF 6: 6-5-1: 1 model with quality test approximately 93% and the RMS error of 0.02%. The input parameters necessary for the operation of proposed ANN model are: number of sunny hours, length of the day, air pressure, maximum air temperature, daily insolation and cloudiness.

Short-term forecast of generation of electric energy in photovoltaic systems

The paper presents selected isotropic and anisotropic mathematical models to calculate the distribution of solar radiation on the photovoltaic module plane with any spatial orientation. A comparison of solar radiation models for Poland is based on measured data and received from the Institute of Meteorology and Water Management. Power density of solar radiation for different angular positions, especially for representative days of each month, was calculated. Based on the statistical analysis of the percentage root-mean-square error (RMSE%), mean-bias error (MBE%) and the Pearson correlation coefficient of an individual mathematical model, our own correction factor for diffuse radiation was proposed. A modified Liu-Jordan model was compared with six common mathematical models showing better agreement of measured and calculated values of solar radiation density. The presented analysis explains which mathematical model is the most suitable for central Poland (Poznan, 52°25’ N, 16°56’ E) and shows the validity of applying the modified model to improve the accuracy of determination of the radiation power density for a given elevation and azimuth angle using values for a horizontal plane.

https://www.mdpi.com/1996-1073/8/2/1025/pdf

Modeling the Distribution of Solar Radiation on a Two-Axis Tracking Plane for Photovoltaic Conversion

The paper presents the methodology of minimisation of the unit cost of production of energy generated in the hybrid system compatible with the lead-acid battery, and used to power a load with the known daily load curve. For this purpose, the objective function in the form of the LCOE and the genetic algorithm method were used. Simulation tests for three types of load with set daily load characteristics were performed. By taking advantage of the legal regulations applicable in the territory of Poland, regarding the energy storing in the power system, the optimal structure of the prosumer solar-wind system including the lead-acid battery, which meets the condition of maximum rated power, was established. An assumption was made that the whole solar energy supplied to the load would be generated in the optimised system.

/pdf/minimisation-of-the-lcoe-for-the-hybrid-power-supply-system-27wc6hlumd.pdf

Minimisation of the LCOE for the hybrid power supply system with the lead-acid battery

The paper presents the most representative – from the three-year measurement time period – characteristics of daily and monthly electricity production from a photovoltaic conversion using modules installed in a fixed and 2-axis tracking construction. Results are presented for selected summer, autumn, spring and winter days. Analyzed measuring stand is located on the roof of the Faculty of Electrical Engineering Poznan University of Technology building. The basic parameters of the statistical analysis like mean value, standard deviation, skewness, kurtosis, median, range, or coefficient of variation were used. It was found that the asymmetry factor can be useful in the analysis of the daily electricity production from a photovoltaic conversion. In order to determine the repeatability of monthly electricity production, occurring between the summer, and summer and winter months, a non-parametric Mann-Whitney U test was used as a statistical solution. In order to analyze the repeatability of daily peak hours, describing the largest value of the hourly electricity production, a non-parametric Kruskal-Wallis test was applied as an extension of the Mann-Whitney U test. Based on the analysis of the electric energy distribution from a prepared monitoring system it was found that traditional forecasting methods of the electricity production from a photovoltaic conversion, like multiple regression models, should not be the preferred methods of the analysis.

/pdf/statistical-analysis-of-the-electric-energy-production-from-t8nfe5o45x.pdf

Artur Bugała

Papers

The energy value and economic efficiency of solid biofuels produced from digestate and sawdust

Short-term forecast of generation of electric energy in photovoltaic systems

Modeling the Distribution of Solar Radiation on a Two-Axis Tracking Plane for Photovoltaic Conversion

Minimisation of the LCOE for the hybrid power supply system with the lead-acid battery

Statistical analysis of the electric energy production from photovoltaic conversion using mobile and fixed constructions