蛇行运动是列车实现高速运行的一大障碍。现有蛇行监测系统都是在大幅蛇行已充分显现时才能识别并报警，且往往仅借助一个观测量来进行判断，其系统的实时性和准确性有待进一步提高。随着高速列车的运行速度不断提高，在大幅蛇行异常现象充分显现之前，利用多表征量对蛇行运动演变趋势进行预测是建立准确、快速的蛇行异常预警系统的有效途径。本项目将理论仿真和台架试验以及机器学习等方法相结合，对高速列车小幅蛇行演变机理及预测方法进行研究。通过建立高速列车整车数学模型和动力学模型，探索高速列车小幅蛇行运动的演变规律；利用机车车辆整车滚动振动试验台，模拟高速列车小幅蛇行演变的过程，揭示高速列车各关键表征参数对小幅蛇行演变趋势的影响规律；利用机器学习，建立基于多个表征参数融合的小幅蛇行演变趋势的预测方法。相关研究有助于丰富蛇行失稳监测的相关理论，为建立高速列车蛇行预警系统提供重要的理论支撑。

Hunting is a serious obstacle to the high-speed operation of trains. All the monitoring systems are designed to detect hunting only after hunting has developed sufficiently. Besides in most cases, the recognition result is obtained by only one observation. As a result, the accuracy and real-time performance of the monitoring system need to be further improved. With the increasing speed of high-speed trains, it is an effective way to establish an accurate and fast warning system for hunting by predicting evolution trend of small amplitude hunting through multi-characterizations before the hunting phenomenon has developed sufficiently. This project combines theoretical simulation, bench test and machine learning to study the evolution mechanism and prediction method of small amplitude hunting of high-speed trains. The evolution law of small amplitude hunting of high-speed trains will be explored by establishing mathematical models and dynamic models of high-speed trains. By using the roller testing rig of railway vehicles, the evolution process of small amplitude hunting of high-speed trains will be simulated to reveal the coupling law between the key characterization parameters of high-speed trains and the evolution trend of small amplitude hunting. On this basis, a prediction method of small amplitude hunting evolution based on multiple characterization parameters will be developed by machine learning theory. Relevant research is helpful to enrich the theory of hunting instability monitoring and will provide important theoretical support for an accurate and fast hunting warning system of high-speed trains.