基于优化模型车辆路径跟踪控制研究

Vehicle Path Tracking Based on Optimized Model

路径跟踪是实现自动驾驶的基础,而根据模型的几何关系实现路径跟踪是一种非常实用的方法.传统的几何控制模型简单可靠、易于改进,但向前预瞄的局限使得此模型在大曲率、高速等极限工况下的跟踪效果较差.为了使几何控制模型的鲁棒性更好、控制精度更高,本文同样基于几何关系控制的原理提出一种优化模型,取消了固定预瞄的计算方式,采用自适应预瞄在预瞄区间中寻找最佳目标点,在转角控制上采用了前馈控制对转角控制进行修正,联合MATLAB和Prescan仿真软件进行仿真测试,并进行实车实验.实验结果表明:该优化模型能够适应多数场景,相较于传统的控制模型,改进了一些缺陷,增加了控制的稳定性,提高了跟踪精度,能够稳定地用于实车开发.

Path tracking is the basis of autonomous driving,and it is a very practical method to realize vehicle path tracking according to geometric relationship.Traditional geometric control model is simple,reliable and easy to improve,but the limitation of forward preview makes the tracking effect worse under the extreme conditions of large curvature and high speed.In order to make the geometric control model more robust and more accurate,in this article we propose an optimized model based on the principle of geometric relationship,which cancels the calculation method of fixed preview.In our optimized model we first used adaptive preview to find the best target point in the preview interval and feed-forward control to modify the angle control.Then we conducted simulation tests with MATLAB and Prescan simulation software as well as actual vehicle experiment.The experimental result showed that the optimized model could adapt to many scenarios.Compared with the traditional control model,our optimized model has improved some defects,increased the stability of the control and improved the tracking accuracy.It can be stably used in the development of actual vehicles.