INFORMATION TECHNOLOGY

Put a pair of “smart eyes” on the city rail train


Team members conduct field tests courtesy of Soochow University

A few days ago, a research team composed of a number of students from the School of Rail Transit of Soochow University developed the “Urban Rail Train Obstacle Early Warning System” and installed a pair of “smart eyes” for the urban rail train.

It is reported that the system can not only detect foreign objects invading the track limit all day long, accurately determine the hazard level of obstacles in real time, but also link the existing train signal system to report road condition information, effectively solving the obstacle warning problem.

The urban rail train obstacle early warning system adopts active intelligent environment perception, and its core technology has a number of advantages, which can improve the completeness of information to more than 95% by fusing multi-source data such as visible light, thermal infrared, and lidar, and can effectively detect key targets within 350 meters of the rail area, and the response time of driver detection can be shortened by 90%, so as to achieve high dynamic obstacle warning of trains and the accuracy rate is as high as 99.9%.

The development of the system lasted nearly 2 years, under the guidance of Wu Cheng, associate professor of the Department of Signal and Control Engineering, School of Rail Transit, Soochow University, the students innovated data fusion, target detection and other technologies. At present, the research results have applied for five national invention patents and published high-level papers in the International Journal of Intelligent Systems.

In the process of actual measurement, in addition to solving academic problems such as debugging algorithms, the team also encountered practical engineering problems such as visible light equipment selection, infrared calibration, radar acquisition parameter adjustment, unstable equipment connection, obstacle simulation, etc. The team actively docked with technical personnel, which was solved after many discussions and practices, which also effectively enhanced the engineering practice ability of team members. “We have completed nearly 20 experiments on Line 1 to collect data and adjust the system, and only after experiencing the overall process of train operation and understanding the various environments inside the subway can we optimize the algorithm more specifically.” Team member and graduate student of traffic information and control engineering Yuan Hao introduced.

It is worth mentioning that the team continued to develop a new subway scene on the basis of the previous results. Since 2020, it has been committed to algorithm research and development and supporting hardware integration, and after multiple rounds of testing and improvement, the hardware has passed the EN50155 railway industry standard test, and finally achieved a technological breakthrough. At the expert review meeting, Professor Yang Shucai, chief engineer of Nanjing Metro, Lu Hongbiao, director of Jiangsu Civil Engineering and Architecture Society, and many other experts showed that the technical route of the project is reasonable, the relevant achievements of the project have reached the expected goals, and the scene application demonstration can be further carried out in the future.

It is understood that the team has previously conducted relevant research in the tram and low-speed freight rail, and its cooperative unit Suzhou Fuxin Intelligent Traffic Control Co., Ltd. li Guangbin, head of the acceptance expert group, commented: “The project adopts the scientific innovation model of industry-university-research-application, independently explores and realizes the exploration and ideas of tram obstacles based on video recognition technology that is still blank in the world. ”

At present, this set of equipment has been installed on some trains of Su Rail Line 5 for trial operation, and has performed well in the actual measurement of obstacle engineering.

Chen Yihao, the person in charge of the project and a graduate student of traffic information and control engineering at Soochow University, said that in the future, the team will continue to adhere to the artificial intelligence technology as the guide, continue to deepen the intelligent driving and vehicle-road coordination, and use technological innovation in the field of rail transit to control operational risks and escort the safe operation of trains. (Source: China Science Daily, Wen Caifei, Yang Shuting)

Related paper information:https://doi.org/10.1002/int.22801



Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button