Design And Development Of Camera System In Solar Power Tower Plant

Solar power is one of the most effective ways to solve energy crisis in the21th century. It has been a hot problem in power research area to increase both efficiency and reliability of the Solar Power Tower (SPT). Therefore, it appears to be necessary to design a control system, which can ensure the SPT runs safely, reliably and efficiently.Several chief problems related to the SPT in the world are researched in this paper, including enhancing accuracy of the controller, correction of control parameters, monitoring and detecting fault heliostats, layout design of cameras, etc. The detailed contents and main contributions of this paper are listed as following:(1) Through surveying relevant technology about SPT around the world, an error correction system for the heliostat controller based on image processing is presented. By improving the Tsai algorithm, transformation from2D to3D space has been achieved. Additionally, a method of binary conversion for gray scale image with self adapted threshold is proposed, as well as designing a filter based on mathematical morphology. Furthermore, algorithms like moment invariants and ellipse fitting are employed to locate the light-spot. Then the estimation and error correction of control parameters are realized through the efforts above. Especially, a new image enhancing algorithm is presented based on the Retinex theory; this algorithm, called MSR-Hist, performs a stronger enhancing and adaptive ability than MSR (Multi-Scale Retinex) and traditional histogram equalization. The MSR-Hist has solved the problem that the light-spot would be hard to be recognized under the condition of weak brightness or luminance fluctuation. Besides, a software system is developed to detect fault heliostats and alert illegal invasion, of course, the ultimate goal is to increase efficiency and stability of the SPT.(2) Since image-based controller parameter correction needs many cameras, a massive amount of data would be transmitted via the Ethernet. That image information would occupy relatively high network bandwidth, which might induce negative effect on real-time and reliability of the network. So, camera modeling and image compressing is studied in this paper. Then an optimized camera layout is designed in order to decrease network load.(3) Based on the researches mentioned above, two relevant applications are developed:1) Calculation of the sun azimuth parameters for heliostat control system;2) BCS parameter correction. Besides, this paper developed a multifunctional software application, which combines fault heliostat detection with illegal invasion alert; in this application, a new Fading Memory Gaussian Background Modeling is introduced to minimize the negative effect caused by small disturbance (e.g. fluttering trees) and long time span variant factors (e.g. illumination and shadow changes in a single day). These applications can communicate with the controller, increase control accuracy, and enhance reliability and efficiency of the SPT in a whole. Moreover, the applications developed in this paper have been employed in the SPT test area in Zijingang of Hangzhou. Long-time running has verified the reliability and stability of the algorithms presented in this paper. It can be safely concluded that the initial goal has been accomplished. Finally, the design of camera layout is simulated using Matlab, and the result showed that the design can satisfy the requirement of the SPT.In the end, summary of this paper and an outlook of the future work are provided.