Diagnostics of leaks with unknown amplitudes against the background of interference caused by accidental consumption in the hydraulic system for the forest complex

Abstract

The formation of algorithms for leak detection is considered. It is shown that the solution to the problem of specific leak detection may vary depending on the type of hydraulic system and the technologies used. Hydraulic fire extinguishing systems for buildings and structures are used in the forest complex. It is shown that main pipelines with combustible substances pass through the forest complex, leaks in which can cause fires, explosions and environmental damage. Mathematical models and methods for diagnosing leaks in hydraulic systems are considered, including identifying the facts of leakage based on the use of mathematical models to determine the location and size of such leaks. In this paper the research focuses on the detection of leaks with unknown amplitudes based on the verification of a two-alternative hypothesis for a hydraulic system, taking into account interference from stochastic consumption. A single leak is considered as a significant amplitude jump against the background of noise arising in the hydraulic pipeline system due to the selection of the target product by consumers. It is noted that the transition from the task of verifying a two-alternative hypothesis to the task of verifying multi-alternative hypotheses or pattern recognition provides a deeper and more comprehensive analysis of leaks.