基于MEMS传感器的汽车试验场道路识别系统

Proving ground road identification systembased on MEMS sensor

为实现汽车试验场结构道路的自动识别,以MPU6050为基础设计了道路结构特征采集系统。通过获取试验车辆轴架处、货舱内和驾驶室内的三轴加速度和角速度;分别采用四元数方法解算得到车辆俯仰角和横滚角;结合三轴加速度得到车辆行驶过程中的垂向加速度;再结合试验车速利用傅里叶变换推导得到道路的空间域特征;选择合适的空间长度进行数据截断;通过DFT得到道路的空间频域特征向量;采用模板匹配方法对试验场6种不同结构道路进行分类。实验结果证明:该系统对试验场道路的分类综合准确率超过95%,且对传感器安装位置、不同车辆和车速具有很高的鲁棒性。

In order to realize the automatic identification of structural road of proving ground,a road structural characteristics acquisition system based on MPU6050 is designed.Tri-axial acceleration and angular velocity are acquired from acquisition system and the quaternion method is applied to calculate the pitch and roll angle of the vehicle.Vertical acceleration of test vehicle can be calculated through these data.A set of uniform interval road characteristics in spatial domain can be derived with vehicle speed based on Fourier transform.Data truncation is carried out and the spatial frequency domain eigenvector can be obtained with DFT algorithm.Template matching method is adopted in order to realize the classification of six roads.Test result shows that comprehensive accuracy is over 95%and it is robust to sensor location,different vehicles and different speeds.