Speaker
Description
The increase in demand for electricity all over the world has driven great interest in Concentrated Solar Power (CSP) technologies integrated with Thermal Energy Storage (TES) systems. This paper offers an ideal design approach for a CSP solar field with AI assistance, aiming for a nominal power production of 100 MW in Hurghada, Egypt’s climatic conditions. To determine the optimal design configurations, the proposed methodology integrates simulation, optimization, and artificial intelligence (AI) technologies. Initially, accurate simulations are carried out using the System Advisor Model (SAM) to determine the performance of the CSP system and generate detailed operational data. The simulation results are then exported to MATLAB and an optimization technique is used to determine the optimal design parameters in order to minimize the Levelized Cost of Energy (LCOE), using the solar field area and TES capacity as design parameters, subject to practical constraints such the TES capacity limits, the permitted land area, and the required energy production. An Artificial Neural Network (ANN) model is developed as a data-driven tool to predict ideal design parameters to further improve the design process. The ANN is trained using simulation data to capture the link between relevant design factors, such as solar field area and TES capacity, and significant environmental inputs, such as solar radiation and ambient temperature. This technique makes it possible to rapidly estimate optimal design configurations under a range of operating conditions.
| Topic Area | Computational Intelligence for Power Systems |
|---|