Mirror Surface Assessment in Solar Power Applications by 2-D Coded Light

Abstract

Concentrative mirrors are typically used to provide the reliable thermal source in solar thermal power applications. Their geometrical optical characteristics are of essential importance to the sun-to-electricity conversion efficiency. However, there is a lack of cost-effective approach for measuring these characteristics in field applications. In this article, we propose a color-coded grid pattern-based method to extract optical properties of paraboloidal mirrors. Comparing with the existing techniques, our assessment system only consists of a digital camera, a color-coded grid plane, and a flat mirror with chess-board pattern, and no additional equipments are required. In our system, the image of the grid pattern is captured by the camera through the reflection of the test mirror. Benefiting from the representation of grid positions with colors, the image points can be easily paired with their source points on the grid plane. Applying ray-tracing and reflection principle, the normal vectors of the reflecting points on the test mirror surface can be determined. These normals facilitate to construct the surface model in the form of a second-order polynomial. The gradient descent algorithm is used to minimize the discrepancy between the measured normals and those calculated from a guess model, by which we improve the accuracy of the model parameters and, finally, obtain the optical properties of the test mirror. The accuracy and efficiency of the proposed method are verified by a set of numerical simulations with respect to mirrors with known parameters. Also, for a real concentrative mirror, the proposed method helps to retrieve quantities of interest and show little dependence on mirror position.