Operation Safety Assessment For Self-driving Trams

Due to the advantages of low cost,low energy consumption,and small pollution,trams have developed rapidly and become one of the important public transportation tools.With the rapid development of artificial intelligence,driverless tram has gradually becom e a research hotspot.However,for the operation condition of trams,the safety of driverless trams during running is still the focus of researchers.How to ensure the safe and effective drivin g of the driverless tram in the urban road,and how to ensure the safety performance of the tram while stable driving has become the focus of the research.Considering the driving characteristics of driverless trams during operation and surrounding environment,this paper mainly launched the following work for the operation safety assessment of driverless trams:1.Analyzing the relationship between the safety assessment method of the driverless vehicle and the driverless tram,and the GPS data contain noise when the tram received information,the surrounding vehicle information is pre-processed and the valid data is obtained;for the requirement of real-time and high accuracy of trajectory prediction,the trajectory prediction m odel is established that merged the characteristics of convolutional neural network(CNN)and long and short-term memory network(LSTM);for the super parameters have a greater impact on prediction results,the grid search algorithm is adopted to optimizes different super parameters.The method is validated by the real datasets and the results show the feasibility of this method;2.Considering the characteristics of the driverless tram running route is fixed at the intersection,the driverless tram dynamic model is established,and the time that the driverless tram at current speed and surrounding vehicle reached the conflict point at the same time is recorded.Combining the Monte Carlo simulation method,the time distribution that the driverless tram reached to the conflict point is counted,and the collision probability between the driverless tram and the surrounding vehicle is calculated based on statistic technique;The risk assessment model of driverless tram at crossroad conflict point under different speeds is established,and the effectiveness of the method is verified by comparing with the experimental parameters set in advance.