基于深度学习的汽车动力学建模与预测  

Modeling and Prediction of Automotive Dynamics Based on Deep Learning

作　　者：葛小菡 GE Xiaohan(Qingdao University of Technology(Linyi),Linyi,Shandong Province,273400 China)

机构地区：[1]青岛理工大学(临沂),山东临沂273400

出　　处：《汽车知识》2023年第8期217-219,共3页Auto Know

摘　　要：该文针对基于深度学习的汽车动力学建模与预测问题进行研究,通过收集和处理实际汽车传感器数据,构建了数据集,基于深度学习的预测模型,利用卷积神经网络(CNN)和循环神经网络(RNN)等技术从原始输入数据中学习特征表示,并预测未来时刻的汽车动力学状态或行为。在模型训练阶段,采用有效的优化算法、合适的损失函数和验证策略,通过迭代优化模型参数来提高预测性能。最后,通过对实际数据集上的实验评估,展示了所提出方法的有效性和准确性。This paper studies the problem of vehicle dynamics modeling and prediction based on deep learning.By collecting and processing the actual vehicle sensor data,a data set is constructed.The prediction model based on deep learning uses Convolutional neural network(CNN)and Recurrent neural network(RNN)technologies to learn feature representation from the original input data and predict the future vehicle dynamics state or behavior.In the model training phase,effective optimization algorithm,appropriate Loss function and verification strategy are used to improve the prediction performance by iteratively optimizing model parameters.Finally,the effectiveness and accuracy of the proposed method are demonstrated through the experimental evaluation on the actual data set.

关 键 词：汽车动力学 深度学习 建模 预测 

分 类 号：U461.1[机械工程—车辆工程]