欢迎访问《兵工学报》官方网站,今天是 分享到:

兵工学报 ›› 1997, Vol. 18 ›› Issue (1): 80-83.

• 综述 • 上一篇    下一篇

电控自动变速器档位决策神经网络方法

秦贵和、葛安林、魏达等   

  1. 吉林工业大学,吉林长春,130025
  • 收稿日期:1996-06-01 修回日期:1997-01-01 上线日期:2014-12-25
  • 通讯作者: 秦贵和

A Neural Network Approach to Optimal Gear Position Decision for an Electronically Controlled Mechanical Automatic Transmission

Qing Guihe, Ge Anlin ,et al.   

  1. Jilin University of Technology, Changchun, 130025
  • Received:1996-06-01 Revised:1997-01-01 Online:2014-12-25
  • Contact: Qing Guihe

摘要: 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曲过程。

关键词: 自动变速 , 智能控制 , 神经网络

Abstract: Optimal shifting in an electronically controlled automatic transmission turns out to be a nonlinear function between the vehicle state and gear position, and is usually described through a data table. It is proper to train a neural network based on the data, and to store the rules of shifting in the network. When used online the network can calculate the optimal gear position. An approach based on this concept is introduced and a neural network training course is illustrated.

Key words: automatic transmission , intelligent control , neural network