Simulation of a CSP Solar Steam Generator, Using Machine Learning

Developing an accurate concentrated solar power (CSP) performance model requires significant effort and time. The block (PB) is the most complex system, its modeling clearly complicated time-demanding part. Nonetheless, PB layouts are quite similar throughout CSP plants, meaning that there enough historical process data available from commercial plants to use machine learning techniques. These algorithms allowed development of a very black-box in short amount This could be easily integrated as into PM. technique selected was SVR (support vector regression). trained using complete year plant situated southern Spain. With limited set inputs, results were accurate, according their validation against new data. not only fit well on aggregate basis, but also transients between operation modes. To validate applicability, same methodology used with different Plant, located MENA region more than double nominal electric power, obtaining excellent fitting validation.