【作者】 孙文秀；

【导师】 张润彤；

【作者基本信息】 北京交通大学 ， 信息管理， 2010， 硕士

【摘要】 近年来城市轨道交通发展迅速,但其设备主要依靠进口,因此国产化成为发展瓶颈。其中轨道交通的自动控制系统为难点之一,重点在于利用计算机控制技术实现列车安全平稳、舒适、正点、节能、精确定位停车等运行控制。与单项产品的开发不同,城市轨道交通自动防护系统(Automatic Train Protection, ATP)是一个复杂系统,涉及机车车辆、地车通信、信号联锁、行车指挥以及司机驾驶等诸多因素,系统开发设计中的疏漏乃至错误都在所难免,所研制设备的性能测试与改进也不可能一蹴而就。建立起一套相对合理完整的ATP性能评价指标体系,并应用计算机仿真技术模拟复杂的机车设备环境和列车运行环境,以研究列车运行防护理论与方法,作为一种实验手段来完成对各种车载设备的测试和评价,是一种十分经济而又安全可靠的方法。本文借鉴复杂过程控制领域的模糊预测控制原理,对于列车自动控制系统(Automatic Train Control, ATC)的子系统——列车自动防护系统进行建模,并构建一套完整的基于隶属函数的ATP模型评价指标体系和ATP测试系统,对ATP模型进行测试评价。本文详细说明了该模型的基本原理及实现步骤：模型分为两部分,ATP系统预测模型和满意优化模型,其中预测模型分为ATP系统的单步和多步预测模型,主要基于预测控制理论,为ATP防护曲线提供依据；满意优化模型是在模糊控制理论基础上引入预测控制原理,用模糊隶属函数描述列车自动防护目标的满意度,建立模糊规则库,构建列车自动防护的多目标优化模型。经过ATP测试系统模型评价指标体系的验证,该模型具有较好的防护效果,其结果是可以接受并令人满意的,从而证明了在列车优化控制原则基础之上的,基于模糊预测控制算法的ATP模型的正确性、通用性和可靠性。更多还原

【Abstract】 In recent years, urban rail transit develops rapidly, however the equipments mainly rely on imports. So the localization of equipments, in which the automatic control system for rail transit is one of the difficulties, has become the development bottleneck. The main point is how to achieve a smooth safety, comfort, punctuality, energy saving, and precise parking control to the train by means of computer-controlled technology. Different with single product development, urban rail transit ATP is a complex system, involving locomotives, communications, interlocking commanding and driver factors in the design and development. As system errors and omissions are unavoidable, develop equipment and improve the performance test may not come easily. Establish a relatively reasonable complete ATP performance evaluation index system, and introduce computer simulation technology of complex locomotive equipment and train operation environment, as an experimental method to finish all kinds of on-board equipment testing and evaluation, is a kind of economy method.This paper presents a model of automatic train protection system (ATP), by drawing the principle of fuzzy predictive control in the field of the complex process control, and constructs the model on a complete set of subordinate function of ATP model based on the evaluation index system and test system of ATP. This paper explains the basic principle of this model and realize the steps:the model is divided into two parts, predictive model and satisfactory optimization model. Predictive model, which is mainly based on predictive control theory, providing theoretic foundation for the protection curve of ATP, divides into the ATP system single-step and multi-step predictive model; Satisfactory optimization model is for the automatic protection of the multi-objective optimization, it introduces predictive control principle based on fuzzy control theory, describes the satisfaction of train automatic protection objectives with fuzzy membership functions and establishment of the fuzzy rules. After simulation and comparative analysis, it shows that the model has a better protective effect, and the result is acceptable and satisfactory. So the accuracy, versatility and reliability of the model of ATP system, which based on the principle of train optimal control and fuzzy prediction control, are proved.更多还原