Indoor trajectory error evaluation method based on LSTM neural network

Abstract

The invention discloses an indoor track error evaluation method based on an LSTM neural network. The method comprises the following steps: collecting a track step number serial number and a reconstruction track point corresponding to the track step number serial number; constructing multi-dimensional feature vectors of the reconstructed trajectory points, obtaining a multi-dimensional feature vector set according to the multi-dimensional feature vectors of the reconstructed trajectory points, and abstracting the indoor trajectory into a sequence; for each target trajectory step number serial number, inputting the multi-dimensional feature vector set corresponding to the target trajectory step number serial number into a trained LSTM neural network model to obtain an indoor trajectory deviation value corresponding to each target trajectory step number serial number, the target trajectory step number serial number being used for representing a trajectory step number serial number greater than or equal to a preset value K, and the target trajectory step number serial number being used for representing a trajectory step number serial number greater than or equal to the preset value K; the mapping relation between the feature vector of each track step number sequence number of the indoor track and the real track is automatically learned and evaluated through a machine learning model, and an indoor track error evaluation result with finer granularity and higher precision can be obtained. 本发明公开了一种基于LSTM神经网络的室内轨迹误差评估方法，方法包括：获取轨迹步数序号和与轨迹步数序号对应的重建轨迹点；构建重建轨迹点的多维特征向量，并根据重建轨迹点的多维特征向量，得到多维特征向量集，将室内轨迹抽象为一个序列；对于每个目标轨迹步数序号，将目标轨迹步数序号对应的多维特征向量集输入已训练的LSTM神经网络模型，得到每个目标轨迹步数序号对应的室内轨迹偏差值，其中，目标轨迹步数序号用于表征大于或者等于预设值K的轨迹步数序号，通过机器学习模型自动学习和评估室内轨迹每个轨迹步数序号特征向量与真实轨迹的映射关系，能够取得更细粒度、更高精度的室内轨迹误差评估结果。