The rapid expansion of renewable energy sources such as solar and wind power is crucial to meet the increasing global demand for clean and sustainable power generation. However, the inherent intermittency and volatility of these energy sources bring forth challenges in maintaining the operation and stability of energy systems. Efficiently storing surplus energy during periods of peak generation and releasing it during peak demand is vital for ensuring a stable and reliable energy supply. While traditional methods of forecasting and optimizing renewable energy systems have their merits, they often fall short when it comes to handling the complexity and uncertainty of renewable energy generation and utilization processes. Machine learning techniques emerge as a potent solution, enabling the development of accurate predictive models and optimization strategies derived from extensive historical data analysis. This study focuses on applying machine learning to optimize solar and renewable energy systems, aiming to boost power generation, enhance grid stability, and ensure seamless integration with conventional power sources. In our comprehensive review of recent research utilizing learning-based methods in the renewable energy domain, we pay special attention to deep learning and machine learning algorithms. We conducted a feasibility study to evaluate the effectiveness of machine learning techniques in tackling the challenges associated with solar and renewable energy system optimisation and energy storage. Our findings indicate that hybrid learning techniques, which integrate machine learning and deep learning with other optimization methods, show great promise in enhancing the development and optimization of these systems. We propose that future endeavours in solving energy production challenges adopt blended learning techniques, leveraging the strengths of multiple approaches to achieve more accurate predictions and efficient energy utilisation.