Thesis title: Energy Management and Optimization for Smart Grids
Nowadays, wind and solar power generation have a major impact in many microgrid hybrid energy systems based on their cost and pollution. On the other hand, accurate forecasting of wind and solar power generation is very important for energy management in microgrids. The trend for a cost effective and efficient wind energy production leads to larger wind turbine generators and drive for more advanced forecast models to increase their accuracy. In addition, electricity price forecasting is a key aspect for market participants to maximize their economic efficiency in deregulated markets. Nevertheless, due to its non-linearity and non-stationarity, the trend of the price is usually complicated to predict. On the other hand, the accuracy of short-term electricity price and load forecasting is fundamental for an efficient management of electric systems. An accurate prediction can benefit future plans and economic operations of the power systems’ operators.
In chapter one, a novel prediction interval model, consisting of several sections (wavelet transform, hybrid feature selection, Group Method of Data Handling neural network, and modified multi-objective fruit fly optimization algorithm), has been developed to short-term predict wind speed and solar irradiation and to investigate the energy consumption of microgrids. The renewables prediction and the energy consumption analysis have been applied to the Favignana island microgrid, in the south of Italy, using the new proposed point forecast model (Group Method of Data Handling neural network and modified fruit fly optimization algorithm - GMDHMFOA) and a Pareto analysis. The results show that the proposed interval prediction model has a good performance in different confidence levels (95%, 90%, and 85%) to predict wind speed and solar irradiation than other already existing methods. In addition, the proposed point forecast model (GMDHMFOA) has an acceptable error and better performance than the other ones commonly used in predicting energy consumption. Lastly, the monthly energy consumption in different stations of the microgrid can be predicted by using the proposed model and provides suitable solutions for energy management of the microgrid.
In chapter two, a combined forecasting model consists of empirical mode decomposition, fuzzy GMDH neural network, and grey wolf optimization algorithm (GWO) has been proposed. In the proposed forecasting model, a combined K-means and Identifying Density-Based Local Outliers (LOF) is applied to detect and clean the outliers of the raw SCADA data. Moreover, the empirical mode decomposition is employed to decompose signals and pre-processing data. Fuzzy GMDH neural network is a forecaster engine to estimate the future amount of wind turbines energy production, where the GWO is used to optimize the parameters of the fuzzy GMDH neural network to achieve a lower forecasting error. Moreover, the model has been applied using real data from a pilot onshore wind farm in Sweden. Obtained results indicate the proposed model has higher accuracy than others in literature and provides single and combined forecasting models in different time-steps ahead and seasons.
In chapter three, a novel hybrid strategy based on intelligent approaches to forecast electricity consumptions has been provided. The proposed forecasting strategy includes three main steps: of a) the evaluation of a correlation coefficient for socio-economic indicators on electric energy consumptions, b) the classification of historical and socio-economic indicators using the proposed feature selection method, c) the development of a new combined method using Adaptive Neuro-Fuzzy Inference System and Whale Optimization Algorithm to predict electrical energy consumptions. The simulation results have been tested and validated by real data sets achieved within 1992 and 2010 in two pilot cases in a developing country (Iran) and a developed one (Italy). The research findings pinpointed the greater accuracy and stability of the new developed method for electrical energy consumption forecasting compared to existing single and hybrid benchmark models.
In chapter four, a new and accurate combined model has been proposed for short-term load forecasting and short-term price forecasting in deregulated power markets. It includes variational mode decomposition, mix data modeling, feature selection, generalized regression neural network and gravitational search algorithm. A mixed data model for the price and load forecast has been considered and integrated with the original signal series of price and load and their decomposition. Throughout this model, the candidate input variables are chosen by a distinct hybrid feature selection. Two reliable electricity markets (Pennsylvania-New Jersey-Maryland and Spanish electricity markets) have been used to test the proposed forecasting model and the obtained results have been compared with different valid benchmark prediction models. Lastly, the real load data of Favignana Island's power grid have been considered to test the proposed model. The obtained results pinpointed that the proposed model’s precision and stability is higher than in other benchmark forecasting models.
In chapter five, a novel optimization strategy to optimal design of a hybrid renewable-based microgird in order to minimize the cost of electricity and loss of power supply probability as economy and reliability parts respectively has been proposed. The hybrid energy system is composed of PV panel, wind turbine, diesel generator, and battery storage system. The proposed optimization strategy consists of Taguchi method, a novel multi-objective optimization algorithm, and a fuzzy decision-maker. In other words, a) hybrid renewable-based microgrid management system (HRMS) is applied, b) Taguchi method is implemented in order to optimize the values of decision variables, c) multi-objective moth flame optimization algorithm has been employed to optimize the size of the renewable-based microgird components, and c) the best Pareto front will be found by fuzzy decision-making approach. Finally, to evaluate the effectiveness and efficiency of the proposed model, different case studies (scenarios) including various house numbers located in Sønderborg, Denmark have been applied. In addition, the effectiveness of the proposed multi-objective optimization algorithm (MOMFO) in solving the optimization problem is examined and the results are compared with Non-Dominated Sorting Algorithm II (NSGA-II) and Multi-Objective Social Engineering Optimizer (MOSEO).