The accuracy and reliability of autonomous speed measurement and positioning for subway trains are prerequisites for ensuring their driving safety and efficiency.This paper builds on the commonly used speed measurement methods for subway trains,and studies the selection of sensors and corresponding algorithms for multi-source data fusion.On this basis,it proposes an intelligent speed measurement method for subway trains using artificial neural networks and combining photoelectric sensors,Doppler radars and accelerometers.The proposed method fully utilizes the self-learning,adaptive,and non-linear capabilities of artificial neural networks.It uses the measured data from three types of sensors,i.e.photoelectric sensors,Doppler radars and accelerometers as its input.It utilizes RBF artificial neural networks for intelligent fusion and rapid optimization.Thus,the weights of the measured data from these sensors are adaptively adjusted to obtain the real-time speed values of subway trains,to achieve the goal of further improving the accuracy and reliability of train speed measurement.
speed measurement methodmultiple-source information fusionRBF nerve networksubway train
维普
万方数据
