【摘要】 为提高自动驾驶车辆传感数据融合后的精度,提出基于DCA的自动驾驶车辆传感数据融合方法研究。建立多尺度下多传感数据动态模型,确定传感数据的动态矩阵,实现动态评价。将获得的动态数据矩阵中引入DCA算法,利用其中的抗原对动态数据进行规范化处理。获得规范化后的传感数据后,利用平均加权方法,调整数据之间的权值,实现数据融合。为了验证设计的数据融合方法的可行性,设计对比实验,采用半实物的自动驾驶车辆,对其中传感数据进行融合处理,结果证明设计的融合方法在进行数据融合后,得出的数据精度更高,为99.8%,时延较短,平均为150ms左右,满足设计初衷。更多还原

【Abstract】 In order to improve the accuracy of automatic driving vehicle sensor data fusion, the research of automatic driving vehicle sensor data fusion method based on DCA is proposed.The dynamic model of multi-scale multi-sensor data is established, and the dynamic matrix of sensor data is determined to realize dynamic evaluation.The DCA algorithm is introduced into the dynamic data matrix, and the antigen is used to normalize the dynamic data.After obtaining the normalized sensing data, the average weighting method is used to adjust the weight between the data to achieve data fusion.In order to verify the feasibility of the designed data fusion method, a comparative experiment was designed, in which the sensor data were fused by using the semi physical automatic driving vehicle.The results show that after data fusion, the data accuracy of the designed fusion method is higher, which is 98%,and the time delay is short, which meets the original intention of the design.更多还原