Train Ranging And Collision Avoidance Method Based On Train To Train Communication Under Signal Faults

With the advantages of large transportation volume,high efficiency,punctuality and energy saving,rail transit has gradually become the main way for passengers to travel,and has played an important role in alleviating travel pressure and improving the convenience of life.The train control system is developing towards automation,intelligence and integration with the development of emerging technologies such as artificial intelligence and big data.As an important technology in the next generation of train control system,train to train communication has received extensive attention by domestic and foreign scholars.The existing train control system uses the train-ground communication to update the movement authority and other information to ensure the safe operation of the train.However,when the signal equipment fails,the train will run under the degraded mode of the automatic train protection system cutting off and manual visual driving,which is a risk of collision between the train and the front train.Different from the existing train-ground communication mode,the train to train communication does not use the ground equipment to realize the direct communication between trains,which can effectively reduce the communication delay and improve the communication rate and reliability.Therefore,it is of great practical significance to study train ranging and collision avoidance method under signal faults to ensure the train operation safety.This thesis focuses on the train ranging and collision avoidance method based on train to train communication under the typical scene of signal faults.The main research contents are as follows:Firstly,aiming at the typical scene of signal faults,an active ranging method of adjacent trains based on train to train communication is proposed.The train is at the risk of collision with the front train under signal faults,so the position of the front train should be obtained to protect the train actively.An active ranging method is proposed based on information transmission time of arrival,realizes direct communication between adjacent trains and obtain the distance information of two trains in real time.The evolution mechanism of the influence on the ranging accuracy is described under the different train running speeds and information transmission delay.Through measuring actual data in different operating speed scenes,the effectiveness of the proposed ranging method is verified.Secondly,an improved Kalman filtering is proposed to solve the problem of signal attenuation and noise interference which may lead to insufficient train accuracy of positioning.The original ranging data cannot be directly used for train protection due to the large errors between the original ranging data and the real data.An improved Kalman filtering is proposed based on the comprehensive consideration of the criterion of gross error,differential evolution algorithm and Kalman filter.This algorithm realizes the discrimination and elimination of gross errors in the original ranging data.Meanwhile,the processed data is optimized to improve the ranging accuracy.The transplantation of the ranging data processing algorithm is realized through programming.The result of the ranging data processing shows that the algorithm can effectively reduce the ranging errors.Finally,a train collision avoidance method based on Bayesian game is proposed to solve the problem of security errors caused by error code,packet loss and broken link in train to train communication process.An incorrect position information of the front train is obtained by train when the train to train communication status is abnormal.To determine the credibility of the front train’s position,a train collision avoidance method is proposed based on Bayesian game.By constructing and minimizing the train loss function,the position of the front train is judged and the optimal action strategy is determined.The movement authority is updated and the train protection speed curve is calculated based on the corrected position of the front train to protect the train.By building a train active ranging protection simulation module,it is verified that the proposed train collision avoidance method can effectively ensure the train operation safety.There are 69 figures,14 tables and 73 references in this thesis.