论文部分内容阅读
电控机械式自动变速器的最佳换档规律是车辆状态与最佳档位间的一个非线性关系，往往以数据表的形式给出。用这些数据去训练一个神经网络，就可使表中反映的换档规律存到网络中；在线应用时，就可用其计算最佳档位。本文介绍了一个基于这种思想的最佳档位判别方案，并以两参数换档规律为例进行了网络训练的过程。
The optimal shift schedule for an electrically controlled mechanical automatic transmission is a non-linear relationship between the vehicle’s condition and the best gear, often given as a datasheet. Use these data to train a neural network, you can make the shift table reflected in the law stored in the network; online applications, you can use it to calculate the best gear. In this paper, we introduce a best gear discrimination scheme based on this idea, and take the two-parameter shift schedule as an example to conduct the network training process.