Application of Artificial Neural Networks for building Sensors Models of Physical Quantity

The article considers creating the model of operation a mobile robot sensor on the example of an incremental encoder using artificial neural networks. Here is described the usage of such a model for detecting sensor faults and hot-swapping a real sensor with its model to ensure the survivability and safety of a mobile robot. For this research, there was carried out the analysis of the subject area and it was established that the subject of fault detection and fault-tolerant control, taking into account the pace of development of mobile robotics, is relevant to the present technologies.