Chapter Six - Renewable energy sources forecasting and integration using machine learning

Abstract

Conventional sources of electrical energy are depleting very fast in nature and also, the use of these causes environmental pollution. Therefore, the use of renewable energies has attracted the world. The use of renewable energy sources (RES) has dramatically risen in recent years due to an increase in power demand. The current rate of technological growth makes it economically feasible for wind, solar, tidal, geothermal, and many other renewable sources to tap electricity. As the solar and wind are highly intermittent, the integration of a large proportion of these RES makes the system unreliable and unstable. An intrinsic attribute in all renewable power plants is that the production of energy relies on environmental conditions such as temperature, pressure, wind speed, humidity, clouds, etc. Therefore, the power from RES cannot be completely regulated or preplanned. It is important to forecast the amount of electricity that can be produced in a power grid for future demand. If the renewable power plants continue to grow, their optimum sizes, positions, and configurations will still need to be calculated. In addition, smart grid management, which involves the incorporation of renewable energy plants, is also a challenging issue. Moreover, extracting the maximum power from the RES is also a major problem while integrating into the grid. Machine learning (ML) is a subset of artificial intelligence that plays a vital role in future energy systems. The challenges of variable supplies of electricity and variable load are now being tackled through ML. ML techniques are very useful in forecasting the power generation from RES considering the environmental effects. In this chapter, a review of numerous ML techniques such as decision trees, random forests, neural networks, support vector machines, multilayer perceptron, gradient boosting, k-means clustering, classification, and regression trees, etc. is used to resolve the above issues relating to the forecasting and integration of RES is included.