为改善车辆的操纵安全性以及为自动转向系统和智能泊车系统的研究提供理论基础,本文给出一种新的汽车方向盘转角识别方法.建立了驾驶员—车辆闭环模型,并通过试验验证车辆模型.通过给定方向盘转角输入求解得到用于GRNN网络训练的车辆状态参数,运用GRNN网络建立以车辆状态参数来识别方向盘转角的映射模型.与RBF网络相比,GRNN神经网络具有更高的辨识精度.进行整车仿真,所建立的GRNN神经网络能较精确地识别方向盘转角并与仿真结果有较好的一致性. In order to provide a theoretical basis for improving the steering safety of vehicles and researching on the automatic steering system and the intelligent parking system, a new steering angle recognition method is presented in this paper. A closed-loop driver-vehicle model is established and verified by experiments The model is obtained by inputting the steering wheel angle input to obtain the vehicle state parameters for GRNN network training and using GRNN network to establish a mapping model which uses the vehicle state parameters to identify the steering wheel angle.Compared with the RBF network, the GRNN neural network has higher recognition Accuracy.After vehicle simulation, the established GRNN neural network can more accurately identify the steering wheel angle and has better consistency with the simulation results.