The invention discloses an automobile track prediction method based on an LSTM technology. The automobile track prediction method comprises the following steps: step 1, collecting real-time information; 2, extracting a track segment; step 3, data conversion and classification; step 4, data set selection processing; step 5, neural network training; 6, predicting a vehicle track; the method comprises the following steps: selecting a road section with clear lane lines and a plurality of turnouts, acquiring information, and acquiring the information at a time point with moderate traffic flow at asampling frequency of 5 Hz; extracting vehicle track segments of going straight, turning left and turning right, and processing the vehicle track segments; the invention discloses an automobile trackprediction method based on an LSTM technology. The straight-going, leftward or rightward probabilities of the adjacent vehicles in the visual field are predicted by utilizing the angle change trend ofthe adjacent vehicles deviating from the original linear track in unit time in the predicted visual field and opening and closing of the steering lamps, so that the accuracy of predicting the vehicletrack for a long time is obviously improved, and the real prediction of the vehicle track of the adjacent vehicles is realized.

    本发明公开了一种基于LSTM技术的汽车轨迹预测方法,包括以下步骤:步骤一,实时信息采集;步骤二,提取轨迹片段;步骤三,数据转换分类;步骤四,数据集挑选处理;步骤五,神经网络训练;步骤六,车辆轨迹预测;选取车道线清晰、同时具有多个岔道的路段,并进行信息采集,在车流量适中的时间点进行信息采集,采样频率为5赫兹;提取直行、左转和右转的车辆轨迹片段,并对其进行处理;该基于LSTM技术的汽车轨迹预测方法,利用对预测视界内相邻车辆在单位时间内偏离原先直线轨迹的角度变化趋势,辅以转向灯的开闭进行预测视界内相邻车辆的直行、向左或向右的概率,明显提高了长时预测车辆轨迹的准确率,实现了相邻车辆车辆轨迹的真实预测。


    Access

    Download


    Export, share and cite



    Title :

    Automobile track prediction method based on LSTM technology


    Additional title:

    一种基于LSTM技术的汽车轨迹预测方法


    Contributors:
    CHEN GUANGXI (author) / HU LING (author) / XIE CHEN (author)

    Publication date :

    2020-06-19


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    G06K Erkennen von Daten , RECOGNITION OF DATA / B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G07C TIME OR ATTENDANCE REGISTERS , Zeit- oder Anwesenheitskontrollgeräte



    Automobile track prediction system based on an LSTM technology and prediction method thereof

    CHEN GUANGXI / HU LING / LI QIAOCHU | European Patent Office | 2020

    Free access

    Prediction of ship track based on ARIMA-LSTM

    Yu, Chen / Fu, Yuhui | SPIE | 2022


    Automobile track prediction method

    QIU HAITAO / ZHU AIJUN | European Patent Office | 2022

    Free access

    Prediction of ship track based on ARIMA-LSTM

    Yu, Chen / Fu, Yuhui | British Library Conference Proceedings | 2022


    Automobile track prediction system and method

    CONG YANCHAO / WANG ZIJUN / LI ZHIGANG et al. | European Patent Office | 2022

    Free access