Mathematical support for monitoring the status and control of operating modes of cryogenic storage systems

This article considers the development of mathematical support for monitoring and controlling operating modes of cryogenic storage systems which increases their safety and prevent losses during storage and transportation. A large number of stationary and transport cryogenic storage systems have strict requirements for safety and the speed of a system’s response to emerging pre-emergency situations which necessitate the automation of managing storage modes using artificial intelligence (AI). The structure of the two-layer neural network for monitoring the state and selecting the storage mode of cryogenic products is presented, including first-layer neurons associated with the monitoring system and second-layer neurons whose outputs are connected to the inputs of logical blocks for selecting the operating mode of the storage system. Options are proposed for processing neural network signals using linear filtering and using a filter based on order statistics – the use of which is advisable in conditions of impulse noise in data transmission channels from the sensors to the monitoring network. A mathematical description of the procedure which initiates the algorithms for correcting the state of the monitoring object is given. The scheme can be used for a wide range of stationary and transport storage systems, including those equipped with refrigeration for recondensing cryogenic vapors.