A Train Environment Monitor System Based On Multi-sensor Data Processing

Due to its powerful transport capacity,safety and convenience,punctual,eco-friendly,efficiency,rail transportation is very popular among in china and plays an important role in China's transportation system.With the continuous development of rail transit science and technology,in the past decade,China's rail transit industry has achieved significant development and is now in a world-class position.With the construction of an overall well-to-do society,people's requirements for the quality of travel have also been continuously increased.China's rail transit is also developing in a direction that is faster,safer,more environmentally friendly,and healthier.How to improve the train's micro environment and ensure the comfort experience of passengers has become a hot topic of research.However,the micro-environment monitoring system of trains is still relatively primitive.There are many problems such as lack of monitoring data,complexity of the monitoring methods,and inconvenient deployment.The environmental quality of the train is mainly related to the temperature,humidity,and carbon dioxide concentration in the train compartment.This paper uses a temperature and humidity sensor,a carbon dioxide sensor,and zigbee technology to develop a train compartment environment monitoring system.Combined with the needs of environmental flow monitoring,this paper uses STM32 F103 development board,based on ucos real-time embedded operating system and SEGGER's em Win graphics library to develop a mobile monitoring end device,which can curve diagram in real-time,showing the three environmental parameters of the trend.In order to meet the needs of system data storage,query,visualization,backup and evaluation processing,this article developed a host computer based on Qt,and used SQLlite to store data.In trains that is in operation,the micro environment of the train compartment is complex,and the range of temperature,humidity,and carbon dioxide concentration change is relatively large.Under the relatively harsh conditions such as long-term vibration,the reliability of sensor data may be affected and a series of Abnormal jump data may be generated.Combined with the multi-sensor and multi-node characteristics of the wireless monitoring system of the train compartment,this article uses Python to find and filter out abnormal data points based on DBSCAN clustering method to reduce the impact of abnormal data on the monitoring system and compared with the traditional limiting filter,moving average filter and first-order lag filter.The shows that abnormal data process can improved system reliability.Based on ridge regression,this paper realizes the regression of monitoring data and the prediction of the train microenvironment trend.Based on the fuzzy comprehensive evaluation method,the comprehensive evaluation of train microenvironment is realized.Finally,in the laboratory environment,a simulation environment was set up to analyze and verify the data collection ability of the system,the data processing ability of the algorithm,the ability to filter out erroneous and abnormal data,and the ability of environmental evaluation.According to simulation experiments,the basic design goals of the system are achieved.