Due to the fact that the equipment of modern electric trains is functionally and technologically complicated, the relevance of creating airborne systems for predictive monitoring of the technical condition of trains to identify their actual and predicted technical condition is increasing. At present, it has not been possible to build automatic on-board systems for predictive monitoring of the technical condition of trains. One of the possible solutions to this problem can be considered the creation of on-board systems, the identification of the technical condition of equipment in which is carried out using neural network technologies. The article proposes a methodology for identifying the technical condition of electric train equipment using artificial neural network technologies, which allows real-time detection of the occurrence and development of malfunctions of electric train equipment with the display of information on the display in the driver’s cab. Taking into account the specifics of the problem being solved, the choice of a multilayer architecture of a direct distribution neural network is justified. All layers of the neural network are completely interconnected, while the number of neurons of the input and output layers of the network is determined, equal to the number of controlled parameters of the technical condition of the electric train and the number of its possible technical conditions, respectively. As a function of activation of network neurons, a logistic function was selected. A heuristic approach is used to train an artificial neural network.