Optimization of multi-layer absorbing systems in solar flat-plate collectors using cluster analysis

In this study, a recursive physical model of multi-layer absorbent systems is defined and validated by the experimental results. The comparison shows that this model has about 93% accuracy. Then, due to reducing the execution time, top-down and k-means clustering methods are used to find the optimum results of different absorbent systems. It was concluded that the best result is obtained for a system with width 100 nm that consists of an anti-reflection layer of SiN with width 10 nm, an absorbing layer of W with width 10 nm, a layer of Mo with width 10 nm and another layer of Mo with width 70 nm in which Mo is used as absorbing or reflective materials. Also, as the major result, it was determined that using cluster analysis for optimizing the studied systems will cause to reduce the execution time rather than a combination of Monte Carlo and finite element methods.