OPTIMAL SIZING AND CONTROL OF RENEWABLE ENERGY BASED HYBRID POWER GENERATING SYSTEM

Abstract

In India, the demand for electricity has been surging due to population growth and increased technological usage in homes, industries, and agriculture. To meet this escalating demand and align with the government's objectives of promoting green and clean energy while ensuring 24x7 availability, renewable energy sources are gaining popularity as a promising solution. India's tropical location offers excellent opportunities for biomass, solar, and wind power generation. Leveraging the country's vast agricultural capacity, substantial amounts of agricultural waste can be harnessed to generate electricity, particularly benefiting communities facing frequent power outages. However, relying solely on renewable energy sources poses challenges due to their sporadic and variable supply. To overcome these issues, the concept of hybrid renewable energy systems (HRES) has emerged as a viable approach to integrate and optimize multiple renewable sources for a more stable and reliable power generation. HRES involves the combination of multiple renewable energy sources like solar, wind, hydro, biomass, or geothermal, alongside energy storage technologies. The primary goal of HRES is to fulfill the energy requirements of a specific system or community while ensuring reliability and sustainability. By harnessing the potential of various renewable sources and mitigating their intermittency challenges, HRES offers a reliable energy supply. If one energy source experiences fluctuations or downtime, other sources can step in to compensate and maintain a continuous and uninterrupted power output. As a result, the overall energy system becomes more reliable, minimizing the risk of power outages. xxxiii The optimal sizing of a HRES is a crucial aspect of its design, involving the determination of appropriate capacities for each renewable energy source, energy storage system, and other components. This process revolves around matching the system's capabilities with the community's energy demand. Over sizing the components can lead to unnecessary costs and resource underutilization, while under sizing may result in inadequate energy production, compromising the system's reliability. Therefore, the present work focuses on optimizing the sizing of grid-connected HRES, combining renewable energy sources with storage devices and the grid. Initially, the feasibility studies, techno-economic analyses, and develops HRES models for two remote sites in the Indian provinces of Haryana and Uttar Pradesh has been conducted . To aid in this analysis, the Hybrid Optimization Model for Electrical Renewable Pro (HOMER Pro) software has been utilized, which effectively assesses feasibility and designs hybrid models based on renewable energy source availability. Given the unpredictable nature of most renewable energy sources in their generation, integrating them into the electricity grid can be challenging and time consuming. This approach presents various technical and non-technical difficulties. The thesis also covers the benefits, challenges, and suggested solutions arising from the process of integrating diverse renewable energy sources into the grid. In recent years, researchers have extensively explored intelligent techniques inspired by natural phenomena to optimize various engineering and technological fields. Therefore, this research utilizes three recently developed optimization algorithms: Aquila Optimization (AO), Colony Prediation Algorithm (CPA), and Tunicate Swarm Algorithm (TSA) for sizing and optimizing HRES. xxxiv Further, to enhance the accuracy of size optimization in HRES, precise weather data obtained through forecasting plays a crucial role. In this study, four machine learning (ML) techniques Gaussian Process Regression (GPR), Support Vector Regression (SVR), Extreme Gradient Boosting (XGBT), and Decision Tree (DT) are employed for hourly forecasting of solar radiation, temperature, and wind speed in selected areas. Comparing the results of these forecasting models (FM) reveals that GPR outperforms the other techniques for both sites. The forecasted data from GPR for solar, wind, and temperature are then utilized for sizing the HRES, catering to the energy needs of remote sites in the Indian provinces of Haryana and Uttar Pradesh. Afterward, the study proceeds with a thorough comparison of the three optimization algorithms: CPA, AO, and TSA. Among these algorithms, TSA stands out as the most promising option, offering superior results and better outcomes in the optimization process.