自动变速汽车神经网络三参数换挡策略  

Neural-networked three-parameter gear-shift schedule for automated-manual-transmission car

作　　者：陈清洪[1] 秦大同[1] 

机构地区：[1]重庆大学机械传动国家重点实验室,重庆400044

出　　处：《控制理论与应用》2010年第11期1580-1584,共5页Control Theory & Applications

基　　金：国家"863"高技术研究发展计划资助项目(2006AA110114)

摘　　要：为了使汽车自动变速器(AMT)换挡能够适应路况与车况变化,更好地满足汽车换挡平顺性和燃油经济性要求,提出了AMT汽车神经网络三参数换挡控制.论文首先给出了神经网络三参数换挡控制原理,接着给出了神经网络三参数换挡控制算法,并且叙述了在长安羚羊AMT轿车上进行神经网络三参数换挡仿真与试验的结果.与神经网络两参数换挡进行的比较表明:采用神经网络三参数换挡比采用神经网络两参数换挡更加符合驾驶员的换挡经验和习惯,挡位切换曲面变化平滑,比传统计算法求解换挡规律更简便、易于实现、鲁棒性更强.For the gear-shift of automated manual transmission（AMT） car to accommodate the road conditions and vehicle states, and to provide the smoothness in shifting, we present a neural-networked three-parameter gear-shift control scheme. After an introduction of the control principle of three-parameter gear-shift based on neural network, we present the control algorithm of the neural network, and provide the simulation and test results on ChangAn Lingyang AMT cars. The comparison with the neural-networked two-parameter gear-shift shows that the proposed gear-shift is better conformed to driver＇s experience or practice and is more smooth for gear-shift surface changes. The gear-shift is also simpler, easier and more robust in operations than the gear-shift controlled by traditional methods.

关 键 词：神经网络 自动控制 汽车工程 自动变速器 三参数换挡 

分 类 号：U463.212[机械工程—车辆工程]