Research On Key Technology Of Passenger Environmental Comfort Monitoring In Intelligent Station

With the rapid development and construction of railway intelligent stations,passengers have put forward more comfortable travel demand for the station operating environment.Therefore,creating a warm and comfortable intelligent station operating environment is of great significance to improve the quality of passenger service.Providing a deeper level of comfort has also become an important part of building a intelligent station.Through on-site investigations at railway passenger stations such as Taiyuan Station,Changsha South Station and Tongliao Station,the existing railway station operating environment monitoring status and existing problems are analyzed,and combined with passengers’ urgent travel demand and intelligent station construction goals.In this paper,the following work is carried out for the intelligent station passenger environment comfort monitoring:(1)According to the current status and existing problems of railway passenger station operation at home and abroad,this paper establishes the overall structure of the intelligent station.In view of the increasing comfort and convenient travel needs of passengers,this paper proposes a definition and evaluation index system for the comfort of passengers in intelligent stations.Drawing on the idea of Artificial Intelligence Internet of Things(AIoT),this paper proposes the front-end structure of intelligent network in intelligent station,and builds an intelligent station passenger environment comfort monitoring system architecture based on intelligent network.(2)The number of passengers has an important influence on the passenger’s environmental comfort.This paper analyzes two main system factors affecting the number of passengers.Based on its time correlation,a long-short-term memory neural network(LSTM)-based passenger population prediction model is established.By predicting the number of passengers arriving at the station in advance,the number of passengers at the station at any time will be derived.This paper also compares the prediction results of LSTM model and BP model.By comparing the evaluation indicators,the analysis results show that the proposed LSTM model has higher prediction accuracy.(3)The temperature inside the station is an important comfort evaluation index that passengers can perceive.It is easy to be affected by many dynamic and complex factors in the station,and has problems such as time-varying,complexity andnonlinearity.In this paper,a support station vector regression(SVR)based on bacterial foraging optimization algorithm(BFOA)is proposed to predict the temperature of railway station environment,the passenger number and environmental data are input as feature vectors,the SVR model is optimized by BFOA,and temperature prediction is carried out based on the optimization model.This paper also compares the parameter optimization results of BFOA,GA and PSO algorithms.In terms of various evaluation indicators,the proposed BFOA-SVR has higher accuracy.In this paper,the actual deployment,testing and verification of the intelligent station environment comfort monitoring system was carried out in Taiyuan Station.The results show that the passenger environment comfort monitoring system can be deployed stably and flexibly in the railway intelligent station,and the monitoring data is accurate and reliable.The LSTM-based intelligent station passenger number prediction model and the BFOA-SVR ambient temperature prediction model proposed in this paper have higher prediction accuracy and robustness than other model methods,which can regulate the station passenger environment comfort.It can provide decision-making basis for station passenger environment comfort regulation and passenger equipment energy saving,and can also provide passengers with better,more efficient and more comfortable passenger transportation services.