Ensuring Observability and Controllability of Complex Technical Systems in Difficult and Irregular Situations When Commands with a Large Distortion Component Are Received

摘要

The article looks at some problems of development of an automated information system. This system carries out monitoring and technical control of complex technical systems at nuclear power plants (NPP) and thermal power plants (TPP). There is substantiated the necessity of using automatic recognition and data synthesis. This improves control when the received command signal is distorted. The control signal can be complicated by the white-noise component. There is offered the development of a software complex, which provides monitoring and analysis of commands, requests, signals for substitution of questionable and unrecognized command signals, received from an unidentified caller via unprotected communication channel. The most similar command is chosen algorithmically from the library of already accumulated commands. In the library the commands are systemized in consideration of the substantiated probability of coincidence and probable damage to the system in case of erroneous command replacement. There are suggested measures to raise the level of software complex safety using neural technologies.