基于人-车-路-环境系统的混合动力汽车整车控制方法研究

项目名称： 基于人-车-路-环境系统的混合动力汽车整车控制方法研究

项目编号： No.61473057

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 自动化技术、计算机技术

项目作者： 连静

作者单位： 大连理工大学

项目金额： 81万元

中文摘要： 目前混合动力汽车的整车控制常基于车辆动力系统模型和典型工况，对道路、环境及驾驶员的适应能力不足，且未考虑行驶的安全性。为此，本项目首先对影响HEV工作模式和工作点的道路特征、环境信息和驾驶员行为参数进行分析，并立足车载设备，研究相关参数获取方法。从道路和环境角度，提出基于车载信息采集的路面坡度、道路曲率、自由驾驶空间和交通状态检测算法，为未来车速轨迹预测提供环境反馈；从驾驶员的角度，在驾驶风格识别及当前意图识别基础上，提出基于环境反馈及RBF神经网络在线学习的未来车速轨迹预测方法，使车辆具备预知能力。然后，建立基于混杂系统理论的整车控制模型，提出基于约束优先级的SOC序列全局优化求解方法，并进一步计算瞬时最优的功率分配，实现控制策略随道路环境及驾驶员变化的自适应寻优。最终，构建以智能交通信息为基础，集节能、环保和安全为一体的HEV控制方法实施框架，推动新能源汽车的智能化发展。

中文关键词： 混合动力汽车；智能交通；整车控制；计算机视觉；智能控制

英文摘要： Nowadays,the Hybrid Electric Vehicle(HEV) control strategy is usually based on the vehicle dynamic model and typical driving cycle. It can't adapt to the road, environment, and driving style variation, and also lack of considering on the driving safety. Therefore,the road characteristic, environment information and driving style parameters which influence the HEV working point are first analyzed in this project,the way to obtain these parameters are also determined based on the on-board equipment. As for the road and environment,the road slope detection,the road curvature,the free driving space detection and traffic state detection algorithm based on the on-board information collection will be developed to provide the environmental feedback for the future speed prediction. As for the driver, based on the driving style and current intention recognition, an on-board self-learning future vehicle speed prediction algorithm utilized the Radical Basis Function (RBF) neural network is proposed to equip the vehicle with the prediction ability. To realize the adaptive adjustment of the control strategy with the variation of road, environment and driver, the vehicle control model is constructed based on the hybrid system theory; the State Of Charge (SOC) sequence global optimization solving method based on the constraints priority is proposed; the real-time optimal power split ratio is calculated. Finally, a general implementation framework of HEV control strategy based on the intelligent transport information and integrated with safety, environmental protection and energy efficiency is constructed. It will promote the intelligent development of new energy vehicles.

英文关键词： Hybird Electric Vehicle;Intelligent Transport;Vehicle Control;Computer Vision;Intelligent Control