中文摘要：

大型槽式太阳能聚光器抛物型镜面尺寸大,要求测量精度高,需要合理稳定的摄影测量网络完成对抛物面形的三维重构。为了提高对聚光器面形的检测精度,避免测量网络选择的盲目性,提出并实现了一种基于遗传算法的摄影测量网络规划方法,研究空间点目标的三维重构过程,将空间点三维重构不确定度作为评价网络优劣的依据并设计遗传算法。针对实验室控制场及太阳能聚光器镜面实际场,首先根据实际测量环境约束进行仿真研究,得到优化后的摄影测量网络,然后通过实测实验对仿真结果进行验证。实验结果表明,在给定测量条件下,利用仿真得到的摄影测量网络进行布站,可以提高面形测量精度,验证了算法的可行性。

英文摘要：

Large-scale trough solar concentrator parabolic mirror measurement need reasonable and stable photogrammetry network to realize the three-dimension reconstruction of parabolic in view of its large size and demand for high precision. A planning method of photogrammetric network based on genetic algorithm is presented to avoid the blindness of measuring network＇s selection. On analysis of 3D points＇ reconstruction and related error propagation theory,space coordinate uncertainties are estimated and selected as quantity indicator of photogrammetric network based on which the network-design-genetic-algorithm is proposed. Measurement and comparison experiments are carried out for both a laboratory field and a practical trough solar concentrator parabolic mirror to prove effectivity of the genetic algorithm. Firstly,the optimal photogrammetry network is obtained by the proposed genetic algorithm in the simulated measuring environment which has many practical photographic constraints. Then,real photogrammetric measurements of the environments are conducted by both the optimal network and randomly selected networks. Comparisons of measurement quantity indicators show that under the given measurement conditions,the optimal network obtained by the genetic algorithm do improves the surface shape measurement accuracy which proves its feasibility.