AMT汽车动态模糊神经网络三参数换挡策略研究  

A Study on the Three Parameter Shift Schedule of AMT Vehicle Based on Dynamic Fuzzy Neural Network

作　　者：陈清洪[1] 秦大同[1] 叶心[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：《汽车工程》2010年第6期505-509,476,共6页Automotive Engineering

基　　金：国家863计划项目(2006AA110114)资助

摘　　要：提出了动态模糊神经网络三参数换挡控制原理及其控制器训练算法,将该控制器运用到长安羚羊AMT轿车上,进行神经网络三参数换挡控制仿真与试验,并与两参数换挡进行了比较。结果表明,采用动态模糊神经网络三参数换挡比两参数换挡更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。The dynamic fuzzy neural network-based control principle of three parameter shifting and the training algorithm for its controller are presented. The controller is used in ChangAn the simulation and test on neural network-based three parameter shifting control, and Lingyang AMT car to conduct their results are then compared with two parameter shifting. The results show that compared with two parameter shifting, three parameter shifting based on dynamic fuzzy neural network accords better with change of gear shifting surface, shifting and its algorithm for finding shifting experiences and habits schedule is simpler and of drivers with smooth easy to implement with stronger robustness.

关 键 词：汽车 机械式自动变速器 三参数换挡 动态模糊神经网络 

分 类 号：U463.212[机械工程—车辆工程]