| Heliostat field is one of the most important parts in the system or tower solar thermal power plant.Its function is to track the movement of the sun and to reflect sunlight to the collector which used to get solar energy.The tracking accuracy can affect the collection efficiency of collector which exists system errors.So the reflection angle error detection and automatic correction of heliostat are key technologies in the system of tower solar thermal power plant.Traditional methods exist different problems like too many parameters to solve,equations too complex to solve,spend long time and impractical and so on.Aiming at the current situation,this paper proposes a new method that it divided the whole work into two parts--zero-point deviation correction and reflection angle correction.The front one correct zero angle errors(α0,β0)and the latter one correct installed position errors(x0,y0,z0)of heliostat.During reflection angle correction,In order to determine the heliostat to be corrected in the image,a group of equations represent relationship between posture of heliostat and it’s shape parameters in the image have been derived.High Dynamic Range Imaging method has been used to grab image whose heliostat borders and solar facula are both clear.In order to fitting right lines of heliostat borders,this paper proposes a multiple linear fitting method based on linear array model.This method does not need to set a parameter and separate the multi-line fitting problem into two-stage estimation problems with high speed.Based on the above content,this paper wrote a zero-point deviation correction software,and a heliostat reflection angle correction software.The softwares have been used to do small-scale experiment in Deyang and Wenshui.And the result proves that the proposed correction method is useful and stable.Compared with traditional methods,this method can solve parameters easily and quickly and can keep several devices work at the same time.It can reduce costs and improve working efficiency. |