Fault Diagnosis And Hazard Level Assessment Of High-speed Rail Train Control On-board Equipment Based On Multimodal Fusion

The high-speed rail train control system is a large-scale complex system.As the operating speed of high-speed rail trains continues to increase,fault diagnosis for on-board equipment becomes increasingly important.At present,in the field of vehicle equipment fault diagnosis,there is a lack of an efficient and reliable diagnostic method.Therefore,under the theoretical guidance of multi-modal fusion technology,this article combines DS evidence synthesis rules to construct two vehicle equipment fault diagnosis models to provide modalities Based on this,the decision-level integration of fault diagnosis results has been completed to improve the accuracy and reliability of fault diagnosis.In terms of modal construction,according to the characteristics of vehicle equipment fault data,a hidden Markov model was used to segment the text,and an LSTM text classification model was constructed to implement fault diagnosis.Two improvement measures were proposed based on the lack of model classification results.According to the structural characteristics of the failure mode,the fuzzy Petri net(FPN)is applied to fault diagnosis modeling.In order to overcome the shortcomings of Petri net's poor self-learning ability and difficult to determine network parameters,the particle swarm optimization algorithm is used to solve the problem.The parameters in the FPN model are trained,and the results prove that the FPN diagnosis model optimized by the parameters can better realize the fault diagnosis function of the vehicle equipment.According to the basic trust distribution function of the two modalities and their diagnosis results constructed above,the two modalities are merged at the decision level through weighted improved synthesis rules.The fusion result proves that the probability of misjudgment can be reduced by using multimodal fusion technology and it has a good effect in improving the accuracy of diagnosis.After obtaining a reliable fusion fault diagnosis conclusion,in order to quantitatively analyze the impact of the fault on the system,a structure color Petri net model was established by analyzing the information flow of the on-board equipment.Based on this,a suitable quantitative index was selected and a corresponding grade score was given.Combined with examples,achieving the assessment of fault hazard levels.This helps managers judge the urgency of a fault and make scientific maintenance decisions to reduce losses.