【作者】 姚精明；

【导师】 刘虎兴； 栗学仁；

【作者基本信息】 铁道部科学研究院 ， 交通信息工程及控制， 2001， 硕士

【摘要】 车组溜放速度控制是驼峰自动化的重点和核心内容。现行的间隔调速法主要有查表法和等时间间隔法，它们通过分析车组溜放的物理过程，建立减速器出口定速的物理模型，采取一次性定速的方式。但是由于车组溜放过程是一个受多因素影响的、非线性时变过程，使得现行的定速法有时存在较大误差，且一次性定速方式不能根据车组溜放的实时状态动态改变出口定速，在前后车组间隔较小时容易产生不良的控制效果。 本文在研究人工调速过程的基础上，应用模糊控制和神经网络理论建立了仿人的间隔调速模型，它将前后车组间的间隔引入减速器出口定速模块，能根据现场的实时情况动态改变出口定速。计算机仿真表明，该模型在经过学习后，能达到较好的控制效果。更多还原

【Abstract】 The core of automated hump is to control the speed of a rolling car. The existing models of interval adjustment are based on the study of rolling process. In these models, the efined speed?is unchangeable, that is, it can be changed according to the space between two neighboring rolling cars. These models are tending to lead bad control results because the rolling process is a time-varying, nonlinear and multi-factor affected process.This paper establishes a speed control model, which is based on theories of fuzzy control and neural network. Space between two neighboring rolling cars is introduced into this model. In this model, efined speed? the output can be changed with the change of the rolling condition. Simulation shows that this model can lead to good control result after it has been trained.更多还原