【摘要】 受大气折光等因素的综合影响，测量机器人测距观测值存在较大误差。首先介绍了斜距的气象改正模型和后视差分改正模型，提出了一种基于回归分析的建模方法；然后对某大坝观测数据进行了实验分析，以基准点历史数据为数据集训练回归模型；最后利用不同改正模型对监测点的斜距观测值进行改正和对比分析。结果表明，以经验回归模型为代表的数学回归建模方法对斜距观测值的改正效果较好，可有效克服后视差分改正因距离和高差差异带来的影响，降低了传统方法需要现场采集气象元素的劳动强度。更多还原

【Abstract】 Affected by factors such as atmospheric refraction, there is a significant error in the distance measurement observations of georobots.Firstly, we introduced the meteorological correction model and the back-sight differential correction model of slope distance, and proposed a modeling method based on regression analysis. Then, taking the historical data of benchmark point as the dataset to train the regression model,we conducted experiments on the observation data of a certain dam. Finally, we used different correction models to correct the slope distance observations of monitoring points, and conducted comparative analysis. The experimental results show that the mathematical regression modeling method represented by the empirical regression model has a good correction effect on the slope distance observations, which can effectively overcome the influence of back-sight differential correction due to distance and height difference, and reduce the labor intensity required by traditional methods to collect meteorological elements on site.更多还原