The invention discloses a vehicle motion prediction control method based on a recurrent neural network, and the method comprises the steps: (1) constructing a vehicle kinematics equation, and collecting an input and output data set of vehicle motion; (2) designing a vehicle motion model based on a recurrent neural network, and training by using the input and output data set; (3) designing a prediction controller of the vehicle motion model, and establishing a vehicle motion prediction model and a target function of the prediction controller; (4) converting the vehicle motion prediction model and the target function into a mixed integer linear programming problem; and (5) solving optimal control and feeding back the optimal control to the controlled vehicle to realize track tracking control. According to the method, a vehicle prediction model based on the recurrent neural network is adopted, a piecewise function is designed to serve as an activation function of the neural network, a vehicle motion model prediction controller based on the recurrent neural network is designed, the prediction model and the target function of the controller are linearized, the calculation problem of the controller is simplified, and the stable control effect is achieved.
本发明公开了一种基于递归神经网络的车辆运动预测控制方法,包括:(1)构建车辆运动学方程,采集车辆运动的输入输出数据集;(2)设计基于递归神经网络的车辆运动模型,并利用输入输出数据集进行训练;(3)设计车辆运动模型的预测控制器,建立预测控制器的车辆运动预测模型和目标函数;(4)将车辆运动预测模型和目标函数化为混合整数线性规划问题;(5)求解最优控制并反馈给被控车辆,实现轨迹的跟踪控制。本发明采用基于递归神经网络的车辆预测模型,设计分段函数作为神经网络的激活函数,设计基于递归神经网络的车辆运动模型预测控制器,将控制器的预测模型和目标函数线性化,简化了控制器的计算问题,实现稳定的控制效果。
Vehicle motion prediction control method based on recurrent neural network
一种基于递归神经网络的车辆运动预测控制方法
2023-07-28
Patent
Elektronische Ressource
Chinesisch
IPC: | 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 |
Vehicle Cornering Motion Modeling Based on a Recurrent Neural Network
British Library Conference Proceedings | 1996
|Europäisches Patentamt | 2019
|Traffic accident severity prediction method based on recurrent neural network
Europäisches Patentamt | 2024
|