考虑前车动态扰动的改进车速在线预测方法  

Improved Velocity Online Prediction Method Considering the Dynamic Disturbance of the Front Vehicle

作　　者：郭兴 马彬[1,2,3] 姜文龙 陈勇 GUO Xing;MA Bin;JANG Wen-long;CHEN Yong(School of Mechanical and Electrical Engineering,Beijing Information Science and Technology University,Beijing 100192,China;Beijing Laboratory for New Energy Vehicle,Beijing 100192,China;Collaborative Innovation Center of Electric in Beijing,Beijing 100192,China;School Traffic Management,People's Public Security University of China,Beijing 100038,China)

机构地区：[1]北京信息科技大学机电学院,北京100192 [2]新能源汽车北京实验室,北京100192 [3]北京电动车辆协同创新中心,北京100192 [4]中国人民公安大学交通管理学院,北京100038

出　　处：《计算机仿真》2022年第12期173-179,316,共8页Computer Simulation

基　　金：北京市自然科学基金项目(3112005,3174049);国家自然科学基金项目(51608040);北京信息科技大学科研水平提高重点培育项目(2020KYNH203);公共安全行为科学实验室开放课题“重点项目”(2021SYS01)。

摘　　要：传统ARIMA车速预测模型忽略了前车动态扰动对自车车速的影响,导致自车车速在线预测误差较大。提出一种考虑前车动态扰动的改进车速ARIMA在线预测方法,提高动态扰动工况下车速在线预测精度。建立车辆运动学模型,基于EKF原理建立自车和前车的历史轨迹模型,通过实验验证模型的准确性;基于ARIMA原理建立自车和前车的轨迹预测模型,通过安全势场法来表征自车和前车预测轨迹的交互关系,有效识别自车危险势场;以危险势场交互点数、交互点与危险势场边界平均间距为输入,以自车减速量为输出,建立车速模糊调节器,对基于ARIMA模型的车速预测量进行修正。结果表明,在前车切入和减速两种行为下,改进预测精度分别提高了35.16%和56.62%,为构建预测能量管理策略提供理论基础。The traditional ARIMA(Autoregressive Integrated Moving Average) velocity prediction model ignores the influence of the front vehicle dynamic disturbance on the speed of the self-vehicle, which leads to a large error in the self-vehicle speed online prediction. Therefore, an improved velocity online prediction method was proposed considering the dynamic disturbance of the front vehicle in order to improve velocity online prediction accuracy under dynamic disturbance conditions. Firstly, the vehicle kinematic model was established, and the historical trajectory model of the self-vehicle and the front vehicle was established based on the EKF principle. The accuracy of the model was verified by experiments. Secondly, the trajectory prediction model of the self-vehicle and the front vehicle was established based on ARIMA theory. Thirdly, the safety potential field method was used to characterize the interaction between the predicted trajectory of the self-vehicle and the front vehicle, and the dangerous potential field of the self-vehicle was effectively identified. Finally, a velocity fuzzy regulator was established based on the input of interactionpoints and boundary of dangerous potential field, and the output of self-vehicle deceleration. It would correct velocity prediction values based on ARIMA model. Results showed that improved prediction accuracy increased by 35.16% and 56.62% respectively under two behaviors of front vehicle cut in and brake. This method provided a theoretical principle for constructing predictive energy management strategy.

关 键 词：车速预测 运动学模型 模糊控制 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]