摘要
针对试运线列车运行的安全性与停车准确度的要求,提出一种基于ARM与多模型实现的列车超速防护系统。通过车载设备传感器采样时间序列数据,并将其进行小波去噪,从而得到基于傅里叶模型的目标跟踪运行曲线。设计以模糊神经网络为动态预测模型的速度控制器,利用预测控制的滚动优化与误差矫正特性增加速度控制器在不同运行环境下的鲁棒性。为加快模糊神经网络的训练速度,将改进型粒子群模糊聚类算法的聚类结果作为模糊神经网络的前件规则构建参数。以中车试运线数据为例对其进行仿真,并通过基于曲线面积误差的评价指标对全局速度下的停车精确度进行分析。仿真结果表明:所提出的以傅里叶模型作为目标函数实现的基于模糊网络的预测控制策略具有明显的优势。
This paper proposes an ARM-based and multi-model implementation for the safety of train operation and the accuracy of parking accuracy.The train overspeed protection system controls the speed.The control system samples the time series data through the in-vehicle device sensor and performs wavelet denoising to obtain the target tracking operation curve based on the Fourier model.The fuzzy controller is designed as the dynamic controller of the dynamic prediction model.The rolling control and error correction characteristics of the predictive control are used to increase the robustness of the speed controller in different operating environments.In order to speed up the training speed of fuzzy neural networks,the clustering results of the improved particle swarm fuzzy clustering algorithm are used as the parameters of the fuzzy rules.Taking the data of the trial line of the medium vehicle as an example,the simulation is carried out,and the parking accuracy at the global speed is analyzed by the evaluation index based on the curve area error.The simulation results show that the fuzzy networkbased predictive control strategy based on the Fourier model as the objective function has obvious advantages.
作者
姜俊彤
李鸿
苏醒
JIANG Juntong;LI Hong;SU Xing(School of Electrical and Information Engineering,Changsha University Of Science&Technology,Changsha 410114,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2021年第1期180-187,共8页
Journal of Chongqing University of Technology：Natural Science
