Diesel power plants are still the main choice for supplying isolated grids in Indonesia. Although this kind of power plant is easy to install, it has a high cost of energy (COE) mainly due to the cost of diesel fuel. Besides, Indonesia as a tropical country has a high intensity of solar radiation. Moreover, the investment cost of PV power plant is getting lower and it does not require high operation cost. Thus, the implementation of PV power plant is considered promising in Indonesia, and the idea of combining diesel power plants with PV in isolated grids arises to increase the efficiency of the COE. Apart from the economical aspect, the technical aspect of a hybrid diesel-PV power system implementation needs to be studied as well. In this study, the impact of the PV power plant interconnection to the existing grid is analysed in terms of the power flow and the transient stability using the DIgSILENT PowerFactory software. Furthermore, a load-sharing scheme is applied to some diesel generator units. According to the simulation result, the hybrid power system operated within the allowable voltage limits. After some transient events occurred, the hybrid power system was able to maintain its stability.
The solar power plant is an alternative to the provision of environmentally friendly renewable electricity, especially in the tropics, which are sufficiently exposed to the sun throughout the year. However, environmental conditions such as rainfall, solar radiation, or clouds may affect the output power of photovoltaic (PV) systems. These factors make it difficult to know whether PV can meet the needs of the existing load. This research develops a model to predict the output power of a 160 x 285W PV system located in the tropics and has certain environmental conditions. The prediction development is supported by the Python programming language with a single hidden layer and two hidden layers Neural Network, as well as the traditional Multiple Linear Regression tools. The simulation results show that the two hidden layers Neural Network method has a higher level of accuracy compared to the single hidden layer and Multiple Linear Regression as seen from the value of R2, MSE, and MAE.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.