The invention discloses an MPC preview suspension control method based on a neurodynamics model, and the method comprises the steps: (1), periodically collecting a system state, inputting a current system state, a plurality of groups of system states closest to the current system state, and collected front road excitation into a trained neural network model, and predicting a system state at a next moment; taking the system state at the next moment as the current system state, and executing the step (1) until a plurality of groups of system states are predicted; and (3) inputting the predicted multiple groups of system states and front road excitation as state quantities into an MPC control model, and controlling the vehicle based on the control quantity at the next moment. A seven-degree-of-freedom suspension motion mathematical model is established by adopting a neural network, so that the nonlinear characteristics of the system can be accurately described, and iterative optimization can be carried out through data accumulation; the MPC model prediction fully utilizes visual preview to obtain a future road surface state and a future suspension motion state trend, and realizes optimal control in a period of time.
本发明公开一种基于神经动力学模型的MPC预瞄悬架控制方法,包括(1)周期性的采集系统状态,将当前系统状态和距当前最近的若干组系统状态、及采集到的前方路面激励输入训练好的神经网络模型,预测出下一时刻的系统状态;将下一时刻的系统状态作为当前系统状态,执行步骤(1),直至预测出若干组系统状态;(3)将预测出的若干组系统状态及前方路面激励作为状态量,输入MPC控制模型,基于下一时刻的控制量来控制车辆。采用神经网络建立七自由度悬架运动数学模型,不仅可以准确地描述系统非线性特性,而且可以通过数据积累进行迭代优化;MPC模型预测充分利用视觉预瞄得到未来的路面状态和悬架未来运动状态趋势,实现一段时间内的最优控制。
MPC preview suspension control method and system based on neurodynamics model
基于神经动力学模型的MPC预瞄悬架控制方法及系统
2024-03-22
Patent
Electronic Resource
Chinese
IPC: | B60G VEHICLE SUSPENSION ARRANGEMENTS , Radaufhängungen und Federungen für Fahrzeuge |
Neurodynamics With Spatial Self-Organization
NTRS | 1993
|Active Suspension With Preview Control
British Library Conference Proceedings | 1994
|Active Suspension With Preview Control
Taylor & Francis Verlag | 1994
|Suspension preview control method and suspension control device based on binocular vision technology
European Patent Office | 2021
|Neurodynamics of learning and network performance
British Library Online Contents | 1997
|