高速铁路列控系统车载模式显示识别研究

Recognition of Modes of Train Control System for High-speed Railway

高速铁路列车运行控制系统车载设备通过人机界面(DMI)图像显示和按键点击等方式和司机进行交互。通过DMI信息自动识别实时获取列控车载设备工作模式,对实现车载设备状态监控、自动测试等,均具有重要意义。本文基于支持向量机(SVM)和粒子群算法(PSO)等方法,对在DMI上显示的列控车载工作模式的分类识别进行研究。在对DMI图像进行预处理得到包含车载工作模式的图像区域后,首先对图像采用2DPCA方法进行降维并提取特征,然后采用支持向量机(SVM)进行训练和学习,其中SVM参数的优化采用改进的粒子群算法(PSO)。仿真实验表明,经过训练后的分类器可快速准确识别DMI显示的车载工作模式,平均识别率达到98.3%。该方法对DMI其它显示信息的识别具有参考意义。

The on-board equipment of Chinese Train Control System （CTCS） interacts with drivers through the image display and the buttons of the Driver-Machine Interface（DMI）. Through the automatic recognition of DMI information, real-time acquisition of the working modes of the on-board equipment of CTCS is significant to accomplish the monitoring and automatic test of the status of the on-board equipment. Based on the SVM （Support Vector Machine） and PSO （Particle Swarm Optimization） algorithms, in this paper, a study was con- ducted on the classification and recognition of the working modes of the on-board equipment displayed on the DMI. Firstly the 2DPCA （2 Dimension Principal Component Analysis） was used to extract the feature of the images and reduce the dimension of the feature after the pretreatment of the DMI images. Then the SVM （Sup- port Vector Machine） was applied to build the classifier and the parameters of the SVM were optimized using improved PSO （Particle Swarm Optimization）. The simulation results showed that the classifier trained by sample data can recognize automatically and accurately the Chinese character, or the working modes of the CTCS system, with the average recognition rate of 98.3%. The method discussed in this paper can be used for the recognition of the other information displayed on the DMI of CTCS system.