A new braking force distribution strategy for electric vehicle based on regenerative braking strength continuity

Regenerative braking was the process of converting the kinetic energy and potential energy, which were stored in the vehicle body when vehicle braked or went downhill, into electrical energy and storing it into battery. The problem on how to distribute braking forces of front wheel and rear wheel for electric vehicles with four-wheel drive was more complex than that for electric vehicles with front-wheel drive or rear-wheel drive. In this work, the frictional braking forces distribution curve of front wheel and rear wheel is determined by optimizing the braking force distribution curve of hydraulic proportional-adjustable valve, and then the safety brake range is obtained correspondingly. A new braking force distribution strategy based on regenerative braking strength continuity is proposed to solve the braking force distribution problem for electric vehicles with four-wheel drive. Highway fuel economy test （HWFET） driving condition is used to provide the speed signals, the braking force equations of front wheel and rear wheel are expressed with linear equations. The feasibility, effectiveness, and practicality of the new braking force distribution strategy based on regenerative braking strength continuity are verified by regenerative braking strength simulation curve and braking force distribution simulation curves of front wheel and rear wheel. The proposed strategy is simple in structure, easy to be implemented and worthy being spread.     

汽车纵向主动避撞系统的研究现状与展望

针对现代交通安全问题,提出了一种辅助驾驶员行车安全的方法,即纵向主动避撞,主要围绕汽车纵向主动避撞系统中的行车信息感知与处理、安全距离模型、车辆动力学系统建模、车辆动力学控制四个关键技术进行叙述.各种先进的控制技术为研究汽车纵向主动避撞系统提供了新的机遇,它所衍生出的新功能新系统对汽车的安全性、舒适性及经济性等方面都有很大的提升.对汽车纵向主动避撞技术进行了综述,探讨了在研究汽车纵向主动避撞系统时所面临的问题,最后梳理出汽车主动避撞技术的发展趋势.     

基于计算机视觉的行人检测技术综述

行人检测是使用计算机视觉技术来判断图像或序列当中有没有行人经过,并同时对行人进行定位的技术,这项技术是无人驾驶领域中至关重要的一个研究方向.受行人个体差异、尺度姿态变化大、鲁棒性要求高等影响,使其具有挑战性,并得到了广泛关注.通过对建立在电脑视觉中的行人检测提箱进行目标分辨和分割的研究情况进行梳理和总结.首先,从阈值法、边缘检测法以及语义分割三个方面,对图像分割技术进行阐述,然后将特征提取的方法分为底层特征和机器学习的两种不同类别作出介绍,之后对分类器的构造方法进行归纳,最后通过目前存在的问题对未来的行人检测技术研究进行展望.     

针对一阶惯性加纯滞后过程的广义预测控制算法研究

针对典型的一阶生产过程，采用广义预测控制算法，完成了基于S7—200PLC的单步广义预测控制器的设计，并以JWS-6／3—4型过程控制装置为对象，以液位为被控变量，实现了单步预测控制。实验表明系统控制精度高，取得了实际误差小于5mm的控制效果，运行稳定，具有一定的实用性和推广性。