The Technical Diagnostics of Electronic Schema on the Base of the Artificial Neural Network

The object of research is the electronic schema, which consists of elements. The subject of research is the reliability of the electronic schema. The tool used for research is artificial neural network. The goal is to reduce expenses on the artificial neural network training. The problems are: the big amount of manual labor on training examples and expenses of time for neural network training. The neural network was used for technical diagnostics of the electronic schema. The artificial neural network must be trained. Training of an artificial neural network requires the training examples, computing resources, time of training. The new switches architecture of the electronic schema was proposed. In article is developed the new way of training of the artificial neural network used the mathematical models of elements in the electronic schema with switchers architecture.