基于XGBoost预测及弹性网误差补偿的室内定位算法

为解决室内定位系统中因环境动态变化而导致定位精度下降的问题，提出一种基于XGBoost并融合弹性网的误差补偿算法。采用XGBoost定位模型对目标位置进行初步预测，当室内环境改变后，再采用弹性网算法构建误差补偿模型，修正XGBoost定位模型的定位误差，并与基于K近邻、支持向量机、随机森林、梯度提升决策树等定位算法做对比。实验结果表明：在更新15%指纹数据库样本的情况下，该算法在80%分位处的定位精度控制在0.73 m以内，明显优于其他定位算法，且较基于XGBoost的定位算法精度提高了25.5%。

Indoor Positioning Algorithm Based on XGBoost Prediction and Elastic Net Error Compensation

Arming at the decreased positioning accuracy caused by the environment dynamic change of indoor positioning system, an error compensation algorithm based on XGBoost fusion elastic net is proposed. XGBoost positioning model is used to make a preliminary prediction on the target position. When the indoor environment changes, the elastic net algorithm is used to construct an error compensation model to correct the positioning error of XGBoost positioning model. The experimental results show that when only 15% of the fingerprint database samples need to be updated, the positioning accuracy of the proposed algorithm is controlled in 0.73m at the 80% percentile, which is significantly better than those of the K-nearest neighbor (KNN), support vector machine (SVM), random forest (RF) and gradient boosting decision tree (GBDT) positioning algorithms, and the accuracy increases 25.5% than XGBoost.