基于多级切换的动车组制动过程自适应控制

发布时间：2018-04-06 05:39

本文选题：高速动车组　切入点：系统辨识　出处：《华东交通大学》2017年硕士论文

【摘要】：高速动车组制动系统是动车组的重要组成部分,也是列车得以安全运行的保障。随着我国和谐系列(CRH)动车组的迅速发展,对列车的制动系统也提出了更高的要求。目前,列车制动主要是通过制动手柄人工机械操作的,自动化程度不高。因此,如何建立有效的动车组制动模型并实现对目标速度优化曲线(满足列车安全、舒适、准确停车的速度-位移曲线)的高精度跟踪控制对列车自动驾驶与停车过程自动化都具有重要的意义。本文首先从相关制动原理出发提出了一种列车制动过程的多级可切换方法,并结合实际运行情况运用基于RBF网络的模型参考自适应控制策略实现对目标速度优化曲线的跟踪控制。具体研究如下:1、根据CRH380AL型列车制动特性,构建了由静态特性和动态特性两部分组成的动车组制动模型,其中静态特性是模型的输入(制动级位)在当前速度下与模型中间量(制动力大小)的非线性关系,为便于分析本文用多个线性模型对其线性处理,并采用最小二乘的方法对多个静态线性函数辨识得到静态特性部分;动态特性则是制动装置作用下,模型中间量(制动力大小)对模型输出(列车速度)的作用,是由列车自身性能所决定的用动态传递函数描述,基于改进的速度变异粒子群算法即求动态性能方程参数,综合两部得到由多个线性模型描述的动车组制动模型。2、针对动车组制动过程的强非线性特点,传统的控制方法在设计算法时计算量过大。本文采用一种基于RBF网络的模型参考自适应控制策略避免了传统控制方法中复杂的数学分析,而待跟踪的目标优化曲线即可由一个稳定的参考模型来描述,结合神经网络的学习记忆功能使得跟踪精度大大提升。3、基于CRH380AL型列车制动特性曲线数据与实际运行过程中的数据对本文所提方法进行仿真与性能测试,仿真结果表明,本文建立的动车组制动模型能满足实际制动过程的要求,基于RBF网络的模型参考自适应控制策略可实现对目标优化曲线的高精度控制。
[Abstract]:High-speed EMU braking system is an important part of EMU, and it is also the guarantee of safe operation of the train.With the rapid development of CRH EMU in China, higher requirements are put forward for train braking system.At present, the train brake is mainly operated manually by the brake handle, and the degree of automation is not high.Therefore, how to establish an effective EMU braking model and achieve the target speed optimization curve (to meet the train safety, comfort,The high precision tracking control of accurate stopping speed-displacement curve is of great significance to the automation of automatic driving and stopping process of trains.In this paper, a multistage switchable method for train braking process is proposed based on the principle of relative braking.The model reference adaptive control strategy based on RBF network is used to realize the tracking control of the target speed optimization curve.The specific research is as follows: 1. According to the braking characteristics of CRH380AL train, the braking model of EMU composed of static characteristics and dynamic characteristics is constructed.The static characteristic is the nonlinear relation between the input (braking level) of the model and the middle quantity of the model (braking force) at the current speed. In order to analyze the linear relationship between the input (braking grade) and the braking force of the model, this paper uses multiple linear models to deal with them linearly.The static characteristic part is obtained by using least square method to identify many static linear functions, and the dynamic characteristic is the effect of the intermediate value (braking force) of the model on the model output (train speed) under the action of braking device.Based on the improved velocity variation particle swarm optimization algorithm, the parameters of the dynamic performance equation are obtained, which are determined by the performance of the train itself and described by the dynamic transfer function.In view of the strong nonlinear characteristics of the EMU braking process, the traditional control method has too much calculation in the design of the algorithm.In this paper, a model reference adaptive control strategy based on RBF network is used to avoid the complex mathematical analysis in the traditional control method, and the target optimization curve to be tracked can be described by a stable reference model.Combined with the learning and memory function of neural network, the tracking accuracy is greatly improved. Based on the data of braking characteristic curve of CRH380AL train and the data in actual running process, the simulation and performance test of the proposed method is carried out. The simulation results show that,The braking model of EMU in this paper can meet the requirements of actual braking process. The model reference adaptive control strategy based on RBF network can achieve high precision control of target optimization curve.
【学位授予单位】：华东交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U266;TP273.2

【相似文献】