%0 Journal Article
%T 基于改进Bi-LSTM的航迹预测模型设计<br>Design of Trajectory Prediction Model Based on Improved Bi-LSTM
%A 雷妮
%J Computer Science and Application
%P 1557-1568
%@ 2161-881X
%D 2023
%I Hans Publishing
%R 10.12677/CSA.2023.138154
%X <span style=\"font-size:14px;\">随着我国航空事业的发展，国家航空安全形势不容乐观，飞机航迹精准预测对执行反劫机任务场景时提高指挥员战场决策能力至关重要，可使指挥员迅速掌握战场态势、精准把握作战时间、快速作出优化决策，对有效打赢反恐战争具有重要意义。针对航迹特征提取过程的梯度消失影响预测精确性的问题，提出了一种基于残差门的Bi-LSTM改进预测模型。在单个LSTM模型中引入残差结构，由正反序两个残差门LSTM构建出Bi-LSTM模型，使得Bi-LSTM模型可以更好的记忆和训练数据特征，避免轨迹数据梯度消失，对飞机航迹数据具有较高的预测精度。</span><br />
<span style=\"font-size:14px;\">With the development of China’s aviation industry, the national aviation safety situation is not optimistic, and accurate prediction of aircraft tracks is crucial to improve commanders’ battlefield decision-making ability when performing anti-hijacking mission scenarios, so that commanders can quickly grasp the battlefield situation, accurately grasp the combat time, and quickly make optimization decisions, which is of great significance to effectively winning the war against terrorism. Aiming at the problem that gradient disappearance affects the prediction accuracy of track feature extraction process, an improved prediction model of Bi-LSTM based on residual gate is proposed. The residual structure is introduced into a single LSTM model, and the Bi-LSTM model is constructed from the two residual gate LSTMs in forward and reverse order, so that the Bi-LSTM model can better remember and train data features, avoid the disappearance of trajectory data gradient, and have high prediction accuracy for aircraft track data.</span>
%K 残差门，航迹预测，Bi-LSTM神经网络<br>Residual Doors
%K Track Prediction
%K Bi-LSTM (Bi-Directional Long Short-Term Memory) Neural Network
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=70748