Optimizing Heliostat Positions with Local Search Metaheuristics Using a Ray Tracing Optical Model

Abstract

Optimizing heliostat positions of solar tower power plants The life cycle costs of solar tower power plants are mainly determined by the investment costs of its construction. Significant parts of these investment costs are used for the heliostat field. Therefore, an optimized placement of the heliostats gaining the maximal annual power production has a direct impact on the life cycle costs revenue ratio. We present a two level local search method implemented in MATLAB which is using the Monte Carlo raytracing software STRAL with a specific annual time scheme for the evaluation of the annual power output. The algorithm was applied to a solar tower power plant with a heliostat field of size 624. Compared to former work of Buck, we were able to improve both runtime of the algorithm and quality of the output solutions significantly. Using the same environment for both algorithms, we were able to reach the best solution of Buck with a speed up factor of 10.