基于MPC的车车通信时间同步方法

Time Synchronization Method for Train-to-Train Communication Based on Model Predictive Control

车车通信已成为5G高速铁路通信系统的重要应用场景,因而实现车车之间的时间同步对于保障列控系统的实时性和安全性至关重要。针对现有车车时间同步方法易受报文丢失影响,导致同步精度低及收敛速度慢的问题,提出1种基于模型预测控制(MPC)的车车通信时间同步方法。首先,建立铁路专用移动通信系统(The Fifth Generation Mobile Communication-Railway,5G-R)下的车车通信同步时钟模型;其次,根据前后车同步时钟相位和频率偏移,引入观测器方程,构建面向MPC的前后车时钟状态空间模型;然后,通过多步预测、滚动优化等控制方式,求解出最优时钟偏移控制量,并通过负反馈校正,实现5G-R下不同车车通信场景的时间同步;最后,对有中继和无中继2种场景下车车通信同步过程进行仿真试验。结果表明:所提方法较其他方法具有更高的同步精度和收敛速度,能够有效完成5G-R下不同车车通信场景的时间同步。

Train-to-train communication technology has become an important application scenario for 5G highspeed railways. Therefore, realizing the time synchronization for train-to-train is critical in ensuring the realtiming and safety of the train control system. Aiming at the problems that existing train-to-train time synchronization methods are easily affected by message loss, resulting in low synchronization accuracy and slow convergence speed, a time synchronization method for train-to-train communication based on Model Predictive Control(MPC) is proposed. Firstly, the synchronous clock model of train-to-train communication under the Fifth Generation Mobile Communication-Railway(5G-R) is established. Secondly, according to the phase of the synchronous clock and the frequency offset of the front and the rear trains, the observer equation is introduced to construct the MPC-oriented spatial models of the clock states of the front and the rear trains.Then, through multi-step prediction, rolling optimization and other control methods, the optimal control quantity of clock offset is solved and the time synchronization under different 5G-R train-to-train communication scenarios is realized through negative feedback rectification. Finally, simulation tests are carried out for train-totrain time synchronization process under different communication scenarios with and without a relay. The results show that the proposed method performs higher synchronization accuracy and convergence speed than other comparison methods do, and can effectively complete the time synchronization in different train-to-train communication scenarios with the application of 5G-R.